Pesticide effects on earthworms. A tropical perspective

Pesticide effects on earthworms A tropical perspective

ISBN: 978-90-8659-394-1 Cover design: M.M. Medagoda & R. Wickramaratne Lay out: Désirée Hoonhout Printing: Ipskamp Drukkers B.V., Enschede Thesis 2009-04 of the Department of Ecological Science, VU University, Amsterdam, The Netherlands.

VRIJE UNIVERSITEIT Pesticide effects on earthworms A tropical perspective ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. L.M. Bouter, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de faculteit der Aard- en Levenswetenschappen op maandag 23 november 2009 om 15.45 uur in de aula van de universiteit, De Boelelaan 1105 door Pallagae Mangala Chathura Surendra De Silva Geboren te Badulla, Sri Lanka

promotor : copromotoren: prof.dr. N.M. van Straalen dr.ir. C.A.M van Gestel prof.dr. A. Pathiratne

Earthworms are the pulses of the soil, the healthier the pulse the healthier the soil" Prof. dr. Sultan Ahmed Ismail Eco-science research foundation, India This research was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB 84-609) residing under the Netherlands Organization for Scientific Research (NWO). The studies were performed at Department of Animal Ecology, VU University, Amsterdam, The Netherlands and Department of Zoology, University of Ruhuna, Matara, Sri Lanka.

Contents Page Chapter 1. General Introduction 9 Chapter 2. Influence of temperature and soil type on the 19 toxicity of three pesticides to Eisenia andrei. Chapter 3. Toxicity of chlorpyrifos, carbofuran, 33 mancozeb and their formulations to the earthworm Perionyx excavatus under tropical conditions. Chapter 4. Comparative sensitivity of Eisenia andrei 45 and Perionyx excavatus in earthworm avoidance tests using two soil types in the tropics. Chapter 5. Development of an alternative artificial soil 57 for earthworm toxicity testing in tropical countries. Chapter 6. Chlorpyrifos causes decreased organic matter 69 decomposition by suppressing earthworm and termite communities in tropical soil Chapter 7. General Discussion 83 References 91 Summary 107 Samenvatting 111 Acknowledgements 115

Chapter 1 General Introduction 9

10

General Introduction General Introduction The world population is expected to reach 9.1 billion by the year 2050 with the current annual growth rate of 1.2 % and the tropical regions of the world will be mostly accounted for this trend (UN, 2001; Cohen, 2005; UN, 2005). Increasing food productivity for food security is very essential in many tropical countries (FAO, 2005) and one direct consequence of the booming food production is the threat to natural ecosystems. The situation is especially alarming in tropical ecosystems because biodiversity is higher than in temperate regions (Lacher and Goldstein, 1997). Tropical ecosystems, although a relatively small fraction of the land surface of the earth (26%), hold a very high primary productivity (Deshmukh, 1986). Sustainability of such systems is directly linked to attainment of equilibrium between human needs such as agricultural production and environmental damage. Some decades ago Chang (1968) already stressed the importance of deliberate planning in maintaining such a balance, which is still lacking in tropical ecosystems. A significant proportion of the world s biodiversity is recorded in agroecosystems (Pimentel et al, 1992). The increasing use of agrochemicals in such systems affects non-target organisms in soil and water (Abdullah et al, 1997; Reinecke and Reinecke, 2007) and is often a major factor contributing to declining biodiversity (Barr et al, 1993; Chamberlain et al, 2000; Robinson and Sutherland 2002; Benton et al, 2003; Bianchi et al, 2006). Nevertheless agricultural production is directly linked with the number and quantity of agrochemicals used (Klassen, 1995; Matson et al, 1997; Carvalho, 2006), illustrating the trade-off between food production and biodiversity. Pesticide use and trends in Sri Lanka The increasing use of agrochemicals is well illustrated by world crop protection product sales. Table 1 shows the growth of the global crop protection market by product sector during 2006-2007. The global market value for crop protection products rose by 9.7% in 2007 compared to 2006. Fungicides highly contributed to this trend followed by herbicides and insecticides and market performance of tropical regions like Latin America, Middle East/Africa and Asia is very strong (Table 2). This general trend of increased pesticide use in the world also holds for Sri Lanka, where the use of agrochemicals has been steadily increasing in spite of more or less constant agricultural production. The development of the volume index of agriculture production in Sri Lanka during 1995-2005 compared to the base period of 1977-1979 is given in Figure 1 together with the annual consumption of pesticides in the same period. The volume index 11

Chapter 1 Table 1. Global crop protection market growth by product sector during 2006-2007. Values are given in nominal US dollars. 2006 (M$) 2007 (M$) Growth (%) Herbicides 14,805 16,115 8.8 Insecticides 7,380 8,016 8.6 Fungicides 7,180 8,105 12.9 Others 1,060 1,154 8.9 Total 30,425 33,390 9.7 Source: Crop Life International (2008) Table 2. Global crop protection market growth by region (2006-2007). Values are given in nominal US dollars. 2006 (M$) 2007 (M$) Growth (%) NAFTA 7,379 7,507 1.7 Latin America 5,203 6,170 18.6 Asia 7,405 7,815 5.5 Europe 9,217 10,568 14.7 Middle East/ Africa 1,221 1,330 8.9 Total 30,425 33,390 9.7 Source: Crop Life International (2008) of production is based on the sum of price-weighted quantities of different agriculture commodities produced after deduction of the quantities used as seed and feed and weighted in a similar manner. It clearly shows that agrochemical consumption is increasing; herbicides are mostly consumed followed by insecticides and fungicides. The pesticide market in Sri Lanka is mainly dependent on imported formulations rather than on technical grade materials with the use of the latter being more or less constant from 1995 to 2005 (Figure 2). In Sri Lanka carbamates, organophosphate insecticides, phenoxy and amide herbicides and inorganic and dithiocarbamate fungicides are the most used groups of pesticides. This is also illustrated by recent data on local pesticide sales from 2000-2005 (Table 3). Pesticides commonly used in the region itself may vary according to the country based on different registration and regulation procedures. On a global scale, Class II insecticides, which are classified as moderately hazardous by the World Health Organization (WHO) based on the human health risks are on top of the list, however, in Sri Lanka, carbofuran (classified as "highly 12

General Introduction Annual consumption (Mt) 9000 8000 7000 6000 5000 4000 3000 2000 1000 160 140 120 100 80 60 40 20 Volume index 0 0 1994 1996 1998 2000 2002 2004 2006 Year Insecticide Herbicides Fungicides Total Volume Index Figure 1. Annual consumption of pesticides by product sector in Sri Lanka and volume index of agriculture production during 1995-2005 (Source: Pesticide registrar s office, Department of Agriculture, Sri Lanka and Department of Census & Statistics, Colombo, Sri Lanka). 8000 7000 6000 Volume (Mt) 5000 4000 3000 2000 1000 0 1994 1996 1998 2000 2002 2004 2006 Year Technical Formulations Total Figure 2. Imports of pesticide technical products and formulations to Sri Lanka during 1995-2005 (Source: Pesticide registrar s office, Department of Agriculture, Sri Lanka). 13

Chapter 1 hazardous") is the most used pesticide. Generally the choice of pesticides in Sri Lanka is still much better compared to other tropical countries like Thailand, India and African counterparts, where chemicals banned in Europe and America, such as organochlorines, are still being produced and used. Table 3. Local sales of commonly used pesticides in Sri Lanka during 2000-2005 (Source: Pesticide registrar s office, Department of Agriculture, Sri Lanka) and hazard classification of pesticides according to WHO (2004). Pesticide Hazard Local sales (Mt) Classification 2000 2001 2002 2003 2004 2005 Insecticides Carbofuran 1b 1068 947 1050 939 756 1057 Diazinon 11 309 119 236 259 360 434 Chlorpyrifos 11 217 214 285 289 258 334 Phenthoate 11 159 205 188 40 207 280 Dimethoate 11 160 131 109 157 173 178 Weedicides Glyphosate U 836 770 963 1271 1525 1827 Propanil 111 797 802 736 817 855 MCPA 111 257 343 625 530 527 Diuron U 29 28 40 47 46 Alachlor 111 11 3 5 7 5 Fungicides Mancozeb U 193 214 258 333 365 725 523 49 8 416 Sulphur U 35 44 70 86 129 111 Cu Oxychloride 111 116 108 80 77 67 58 Maneb U 76 61 97 93 91 103 Cu hydroxide 111 13 11 13 22 27 27 1b: highly hazardous, 11: moderately hazardous, 111: slightly hazardous, U: unlikely to present acute hazard in normal use Tropical ecotoxicology and risk assessment Tropical ecotoxicology is the science that studies adverse effects of xenobiotics on natural organisms under tropical conditions. This approach is unique compared to traditional ecotoxicology since it has to deal with biotic and abiotic factors in a system that is very different from the temperate regions. Tropical ecotoxicology is also incorporating socioeconomic, cultural and political aspects because agricultural management decisions are directly affecting the lives of families in local villages (see also Lacher and Goldstein, 1997). The greatest challenge in tropical ecotoxicology is to assess the ever growing use of agrochemicals 14

General Introduction and their effects on the ecosystem. Nevertheless, assessing such effects should be a major consideration when aiming for sustainable development in tropical regions. Unfortunately assessment systems incorporating sustainability considerations are lacking in tropical regions due to various reasons. A lack of information on the side-effects of pesticides on major ecological processes has been identified as a major obstacle in establishing sustainable agriculture in tropical conditions (Van den Brink et al, 2003). Information on the impacts of agrochemicals on human health (Roberts et al, 2003; Perera et al, 2008; Samarawickrema et al, 2008; Mohamed et al, 2009) and aquatic organisms (Castillo et al, 1997; Guruge and Tanabe, 2001; Satapornvanit et al, 2004, De Silva and Samayawardehena, 2005, Chandrasekara and Pathiratne, 2007; Daam et al, 2009a) is available compared to only limited data on effects on soil organisms. This limited information on the effects of pesticides on soil organisms includes some studies with non-standardized methods (Senapati et al, 1991; Verma and Pillai, 1991; Rao and Kavitha, 2004) and also with standardized methods (Helling et al, 2000; Römbke et al, 2007; Garcia et al, 2008). Nevertheless the use of a proper risk assessment scheme for soil organisms in tropical risk assessment is limited due to scarcity of data. The complexity of ecological processes in soil and their interactions under tropical conditions also demands more data when assessing such risks. Therefore assessing the effects of agrochemicals under these conditions is a prerequisite. Pesticide regulation and risk assessment in tropical areas are still largely dependent on temperate data, which represent temperate species and conditions. In aquatic ecosystems it has been shown that temperate data could also be used in tropical conditions (Castillo et al, 1997; Karlsson, 2004; Maltby et al, 2005; Kowak et al, 2007; Daam et al, 2009b). In soil ecosystems this approach may not be appropriate while assessing pesticide effects on soil ecosystems under tropical conditions is very essential in view of the sustainability aims. Pesticide toxicity in the tropics It is expected that toxicity of a pesticide in tropical regions may not be comparable with other geographically distinct ecosystems in the world. Both the fate and the effects of the pesticide could be different in tropical ecosystems compared to other regions (Bourdeau et al, 1989; Lacher and Goldstein, 1997) and this could also depend on the type of pesticide (Garcia, 2004). Higher temperature and humidity unique to tropical climates are the main environmental differences. In addition, tropical species might have a greater or lesser innate sensitivity to a pesticide, even when conditions are the same. There are two possible hypotheses to explain differences in pesticide toxicity in tropical conditions. On the effects side, higher temperatures and 15

Chapter 1 humidity may increase the overall activity of living organisms. The resulting higher uptake of pesticides may cause more serious adverse effects. Alternatively, exposure conditions may vary in tropical conditions mainly due to degradation and sorption, two key processes that determine pesticide fate in soil. Degradation rates may depend on physico-chemical and biological properties of the soil (Rao et al, 1983) and especially humid tropical climates may enhance the degradation of a chemical due to hydrolysis, volatilization and other degradation processes (Klein, 1989; Viswanathan and Krishnamurti, 1989; Racke et al, 1997; Laabs et al, 2002). On the other hand, sorption may mainly depend on the organic matter content (Spark and Swift, 2002; Coquet, 2003) and higher variability of soil structure in tropical regions may therefore significantly affect bioavailability. In addition a more active microbial community may enhance the biodegradation of a pesticide (Abdullah et al, 1997; Singh et al, 2003). The overall effects of all these processes may result in lower as well as higher availability of pesticides in the tropical ecosystem resulting in lower as well as higher toxicities to soil organisms. Since predictions on the risks of pesticides in tropical ecosystems are difficult to make, more experimental work is necessary. This thesis reports on such work, focusing on earthworms. Earthworm ecotoxicology Earthworms play a vital role in soil ecosystems and their presence to sustain a healthy soil was already highlighted by Charles Darwin (Darwin, 1881). Their contribution to complex processes such as litter decomposition, nutrient cycling and soil formation is very important hence presence of earthworms is beneficial to agro-ecosystems (Edwards and Fletcher, 1988; Lavelle, 1988; Lavelle et al, 1997; Edwards, 1998; Eriksen-Hamel and Whalen, 2007). As a major component of the soil ecosystem they are susceptible to xenobiotics like pesticides and effects could be seen at the species, population and even the community level (Edwards and Bohlen, 1992). In fact it was Kennel (1972, 1990) who first reported that reduced litter decomposition in orchards could indicate pesticide effects on earthworms. Since then, the use of earthworms in ecotoxicological studies is common and a large database on pesticide effects on earthworms exist (Spurgeon et al, 2003; Frampton et al, 2006), varying from biomarkers (Van Gestel and Weeks, 2004; Rodriguez-Castellanos and Sanchez-Hernandez, 2007) to field effects (Förster et al, 2006; Reinecke and Reinecke, 2007; Casabé et al, 2007). Linking effects reported in the laboratory to field conditions was a challenge and nevertheless extrapolation of such data to realistic conditions was rather successful (Heimbach, 1992; Van Gestel, 1992; Kula, 1995; Holmstrup, 2000; Jänsch et al, 2006)). Unfortunately most of these studies were focused on conditions representing North 16

General Introduction America and Europe and therefore they may not be applicable to tropical conditions. Eisenia fetida and Eisenia andrei have been the standard species in earthworm ecotoxicology especially in the temperate regions (see Jänsch et al, 2005). A relatively short life cycle, a high cocoon production, continuous breeding and easy culturing under laboratory conditions have made them suitable model species. One main advantage of using Eisenia sp. in toxicity studies is their cosmopolitan distribution (Blakemore, 2006), which explains why it also has been used in tropical ecotoxicity studies (Garcia, 2004). However, species belong to the genus Eisenia are primarily vermicomposting worms and their ecological role in natural soil ecosystems is rather questionable. Therefore linking effects observed in Eisenia sp. to ecologically relevant species is a concern. This makes it even more vital to use indigenous tropical species for toxicity testing. So far other species like Perionyx sp. from the family Megascolecidae (Callahan et al, 1994; Maboeta et al, 1999; Reinecke et al, 2001; Muthukaruppan et al, 2004), Lampito mauritii (Megascolecidae) (Reddy et al, 1984; Santhanam et al, 1992), Eudrilus eugeniae (Eudrilidae) (Callahan et al, 1994; Fitzgerald et al, 1996; Reinecke et al, 2001) and Pontoscolex corethrurus (Glossoscolecidae) (Garcia, 2004) have been used in ecotoxicity studies. This limited information on tropical species is rather inadequate and the use of local species in earthworm toxicity tests is therefore of high priority. Outline of the thesis The main objective of this thesis is to assess pesticide effects on earthworms under tropical conditions by laboratory studies using different endpoints and linking such effects to earthworm community structure and functioning of soil ecosystems. The second aim was to develop soil ecotoxicological methods for tropical regions. In Chapter 2, effects of three commonly used pesticides on the standard earthworm test species Eisenia andrei at two different temperatures representing temperate (20 0 C) and tropical (26 0 C) conditions were investigated. Different soils representing both tropical and temperate conditions were included to determine the possible effects of soil types. The endpoints measured were survival, reproduction and growth. The standard test species Eisenia andrei and Eisenia fetida have been a popular choice in soil ecotoxicology for decades, nevertheless these species are rarely found in natural ecosystems. Therefore toxicity data generated through these species may not be ecologically relevant and the use of local species should be encouraged in tropical ecotoxicology. In Chapter 3, toxicity of chlorpyrifos, carbofuran and mancozeb on the same endpoints was investigated using the tropical earthworm species Perionyx excavatus and compared with Eisenia andrei (reported in Chapter 2) in the same soil 17

Chapter 1 types under tropical conditions. In realistic field conditions, farmers use formulated products, which are combinations of one or more active ingredients and several additives. Standard toxicity tests, however, always use pure compounds. Comparative studies of the formulated products and pure compounds are rare both under temperate and tropical conditions. Therefore the comparative toxicity of the pure compounds of chlorpyrifos, carbofuran and mancozeb together with their formulated products to Perionyx excavatus were also investigated in Chapter 3. Avoidance tests have been considered as complementary screening tests in the risk assessment of chemicals and are reported to be ecologically relevant and even more sensitive than traditional endpoints such as survival and reproduction. The short duration of avoidance tests and their cost effectiveness seem to make them ideal under tropical conditions. Earthworm avoidance tests were therefore performed using the standard test species Eisenia andrei as well as the tropical species Perionyx excavatus in chlorpyrifos and carbofuran spiked soils (Chapter 4). This has also been linked to the relative sensitivity of the two species reported in Chapter 2. One of the major constraints in ecotoxicology is a lack of tropical perspectives in standard test guidelines and designs. The use of artificial soil, which was developed earlier by OECD (1984) and later adopted by ISO (1993), is very common in soil ecotoxicity tests in temperate regions. It consists of 70 % sand, 20 % kaolin, 10 % sphagnum peat and a small amount of CaCO 3 to adjust ph. Sphagnum peat is not widely available in tropical and subtropical regions, making it necessary to develop a tropical artificial soil with locally available materials. Therefore an attempt was made to replace the peat component with environmentally friendly local materials. A test procedure using various replacements was validated using chlorpyrifos, carbofuran and carbendazim (Chapter 5). A tiered approach involving effect assessment of plant protection products on soil organisms has rarely been reported in both temperate and tropical regions. This approach may be beneficial in risk assessment processes especially in tropical regions, mainly due to the complex biotic and abiotic factors and their interactions. The use of field tests to assess pesticide effects on soil organisms may yield valuable information on soil community structure (diversity, abundance, and biomass) and functions (nutrient cycling, community respiration and organic matter decomposition). Hence in Chapter 6, effects on organic matter breakdown of chlorpyrifos doses relevant to normal agricultural practice were investigated, linking the observed effects to earthworm and termite abundance in the field. Chapter 7 gives a general discussion of the results (Chapters 2-6) of this thesis, focusing again on the special challenges facing tropical ecotoxicology of soil ecosystems. 18

Chapter 2 Influence of Temperature and Soil Type on the Toxicity of Three Pesticides to Eisenia andrei P. Mangala C. S. De Silva, Asoka Pathiratne, Cornelis A.M. van Gestel Chemosphere 76 (2009) 1410-1415 19

Chapter 2 Abstract Expansion of agriculture in the tropics has increased the use of pesticides that may affect the soil ecosystems. Few studies so far determined the effects of pesticides in the tropics and tropical risk assessment therefore often relies on data from temperate conditions. Hence the toxicity of chlorpyrifos, carbofuran and carbendazim to the earthworm Eisenia andrei was compared at two different temperatures reflecting temperate and tropical conditions. The toxicity of the three pesticides in both conditions decreased in the order carbendazim > carbofuran > chlorpyrifos. For chlorpyrifos and carbofuran, but not for carbendazim, survival was more sensitive at the higher temperature, probably due to increased earthworm activity. Sub-lethal effects (reproduction and growth) however, varied inconsistently with temperature and soil types. We conclude that toxicity of pesticides in tropics may not be predicted from data generated under temperate conditions, even within the same species. Key words: Earthworms, Chlorpyrifos, Carbofuran, Carbendazim, Tropical and temperate conditions, Sri Lanka. 20

Influence of temperature and soil type on the toxicity of three pesticides Introduction The ever-growing human population, which is expected to be 9.1 billion by 2050 (UN, 2005) places high demands on agricultural production, especially in tropical regions. Increasing agricultural production in these regions relies heavily on pesticide use as shown by the growth of the pesticide market (Matson et al., 1997; Ecobichon, 2001). The world agrochemical use has shown a continuous growth with an increase of 4.7 % in annual sales from 2002 to 2003. Tropical regions in Latin America, Asia and Africa have recorded a dramatic growth of pesticide use over the last few years. Also in Sri Lanka annual imports of pesticides have shown a steady increase over the years. The amount of imported technical grade materials has increased from 365.95 metric tons (MT) in 2001 to 489.72 MT in 2005. This trend is also more obvious when considering the amount of imported pesticide formulations that increased from 3457.60 MT in 2001 to 6214.77 MT in 2005 (Unpublished data, Pesticide Registrar s office, Department of Agriculture, Sri Lanka). Increased use of pesticides may affect non-target organisms in soil and water and can cause serious damage to ecosystems (see e.g. Reinecke and Reinecke, 2007). Information on the side effects of pesticides in the tropics is scarce or lacking (Van den Brink et al., 2003), and environmental effects often are totally ignored. Lacher and Goldstein (1997) reported that most of the studies on pesticide toxicity in tropical areas are dealing with water quality and aquatic toxicology, with only little data available on effects in soil. Due to the different environmental conditions, like higher temperature and humidity, fate and transport of the pesticides (Bourdeau et al., 1989) and route of exposure and toxicity may be quite different under tropical conditions (Viswanathan and Krishnamurti, 1989; Laabs et al., 2002). A risk assessment based on temperate data could therefore be less appropriate for tropical conditions. Soil invertebrates play an important role in decomposition and nutrient cycling processes that are very important for sustaining a healthy and productive soil. For that reason, it is important to properly assess pesticide risks to the soil community, and such an assessment should be part of a system of sustainable agriculture. Although for pesticide registration internationally accepted protocols should be used to determine fate and effects in tropical soils, the few studies available (e.g. Senapati et al., 1991; Verma and Pillai, 1991; Rao and Kavitha, 2004) used non-standardized methods. Only recent studies (Helling et al., 2000; Römbke et al., 2007; Garcia et al., 2008) applied standardized protocols to determine pesticide effects to soil invertebrates under tropical conditions. But these standardized protocols have been developed mainly for the assessment of chemical toxicity under temperate conditions. The toxicity tests currently described by the Organization for Economic Co-operation and 21

Chapter 2 Development (OECD) and the International Standardization Organization (ISO) for soil invertebrates, for example, all employ an ambient temperature of 20 C, while under tropical conditions temperatures of 25->30 C prevail. Earthworms are major components of the soil ecosystem. They play a major role in decomposition, nutrient cycling, and soil structuring, which is vital for the existence of healthy soil (Edwards, 1998) and their presence therefore is beneficial especially in agro-ecosystems. The presence of earthworms in a wide range of soils and their high contribution to the soil biomass also makes them suitable to determine the effects of soil pollutants such as pesticides. Edwards and Bohlen (1993) reported that earthworms are susceptible to pesticides and therefore may act as model organisms in evaluating the effects of pesticides. The aim of this study was to compare the toxicity of pesticides to earthworms at two different temperatures, indicative of tropical and temperate conditions. We therefore determined the effects of chlorpyrifos, carbofuran and carbendazim, three frequently used pesticides, using the artificial soil described in the standardized test guidelines developed by OECD (1984) and ISO (1998). Tests were conducted in The Netherlands (temperate) and in Sri Lanka (tropical); in addition to the standard artificial soil a local natural soil was included. Materials and Methods Test species Eisenia andrei (Lumbricidae) was selected as the test species. The animals were taken from a synchronized culture at the VU University in Amsterdam, The Netherlands, where experiments were carried out at 20 C, simulating temperate conditions, as prescribed in the OECD test protocol. The worm cultures were maintained at 20 ± 2 C in complete darkness in plastic boxes with 50 % potting soil and 50 % peat moss, which was moistened to 50 % of the maximum water holding capacity (WHC max ), and horse manure was used as the food source. The tropical cultures were maintained at the Department of Zoology, University of Ruhuna, Matara, Sri Lanka at 26 ± 2 C. The culture substrate under tropical conditions was a mixture of finely ground cow manure (30 %) and composted coco peat (70 %) with fresh cow manure as the food. Adult worms (400 500 mg wet weight) with well-developed clitellum were selected for the toxicity tests. The selected worms were separated from the culture substrates 24 hours before the experiment and transferred to test substrates for acclimatization. 22

Influence of temperature and soil type on the toxicity of three pesticides Test substrates The standard artificial soil (AS) developed by OECD (1984) was used as the common substrate for the experiments. AS was prepared by mixing 70 % quartz sand, 20 % kaolin clay and 10% finely ground sphagnum peat (< 1 mm). By adding CaCO 3, the ph (0.01 M CaCl 2 ) of the AS was adjusted to approx. 6.0. In addition, two natural soils were selected: LUFA 2.2 (Rhineland-Palatine, Germany), classified as loamy sand with a ph of 5.9-6.1, 3.5-4.5 % organic matter and a WHC max of 50 %, was selected as the temperate natural soil. This soil is commonly used in toxicity tests with soil invertebrates (Løkke and van Gestel, 1998). Due to its unavailability in the tropics, a natural soil defined as sandy clay loam was collected from a site near Dickwella, Matara, Sri Lanka. The Dickwella soil was selected because it usually has high earthworm population densities. The uppermost soil layer (20 cm) was taken, sieved (5 mm) and air-dried at 26 ± 2 C. The soil had a ph (0.01M CaCl 2 ) of 6.2, 9 % organic matter and a WHC max of 45 %. Test chemicals Chlorpyrifos (O,O-diethyl O-3,5,6-trichloro-2-pyridyl phosphorothioate) is an organophosphate insecticide, carbofuran (2,3-dihydro-2,2-dimethyl-7- benzofuranyl-n-methylcarbamate) a carbamate insecticide/ nematicide, and carbendazim (methyl N-(1H-benzoimidazol-2-yl) carbamate) is a fungicide. Chlorpyrifos (98%, Cheminova Ltd, Denmark) and carbofuran (97%, Sigma Aldrich) were tested as pure compounds, while carbendazim was tested as the formulation Derosal (AgrEvo, 360 g l -1 ). Experimental procedure The toxicity tests were based on the guidelines first published by ISO (1998b) and later adopted by OECD (2004a). The first series of experiments were performed using AS and LUFA 2.2 soil as substrates under temperate conditions (20 ± 2 C), the second series with AS and Dickwella natural soil under tropical conditions (26 ± 2 C). Based on the results of range-finding tests, concentrations series of chlorpyrifos (1, 3, 10, 30, 100, 300 and 900 mg a.i. kg -1 dry soil) and carbofuran (0.5, 1, 2, 4, 8, 16 and 32 mg a.i. kg -1 dry soil) were selected to enable assessment of full dose-response relationships. The reference chemical carbendazim was tested at concentrations of 0.1, 0.3, 1, 3, 10, 30 and 90 mg a.i. kg -1 dry soil. All tests included controls and solvent controls without pesticide treatment. The pure pesticides were mixed in with the AS and natural soils using acetone ( 99.5% (Sigma-Aldrich). After mixing in, the soil was left overnight to allow for the acetone to evaporate. Solvent controls received 23

Chapter 2 the same acetone treatment. Carbendazim, which was tested as a formulation, was introduced into the test soils as an aqueous suspension. After treatment, the soils were moistened to 50 % of their respective WHC max. Test containers were glass bottles (750 ml) with loosely closing lids. Test containers were filled with approx. 710 (AS) or 610 g (natural soil) moist soil. Five grams (dry weight) of food (finely ground horse manure for tests under temperate conditions and cow manure for tropical conditions) moistened to 50% dry weight was added in a hole made in the middle of the soil (see Van Gestel et al., 1989). Earthworms (n = 10, 4 replicates) were introduced into the soil, and the lids were not closed tightly to facilitate free exchange of air. Before introduction, the worms were rinsed with water, blotted dry on filter paper and mass were determined (per 10 worms, and individually for a number of worms to check for individual variation). The test containers were weighed and kept at 20 ± 2 C and 26 ± 2 C, respectively for tests under temperate and tropical conditions. At weekly intervals moisture loss from the test soils was checked by weighing the test containers and replenished if needed. Additional food was introduced when all the food added was consumed. After 28 days of exposure, adults were removed by hand sorting, and mortality and biomass were determined. The soil was returned into the test containers and incubated for another 28 days at the respective temperatures to allow for cocoon development. After 56 days, juveniles were extracted from the test soil using a water bath kept at 60 C and counted. The final endpoints studied were adult mortality, change of biomass after 28 days, and number of juveniles produced after 56 days. Statistical analysis Levene s test and Kolmogorov-Smirnov test were performed to assess homogeneity of variance and normal distribution of the data, respectively. LC 50 s, the concentrations causing 50 % mortality of the earthworms, and corresponding 95 % confidence limits were calculated with the Trimmed Spearman Karber method (TSK; Hamilton et al, 1977). The model for logistic response of Haanstra et al. (1985) was used for the calculation of concentrations affecting reproduction by 50 % (EC 50 ) and 10 % (EC 10 ) and the corresponding 95 % confidence limits. Differences in EC 50 values between different soil types and the two test conditions were tested for significance using a generalized likelihood ratio test (Sokal and Rohlf, 1995). LOEC and NOEC values were determined using ANOVA and Dunnett s test by SPSS version 14.0. 24

Influence of temperature and soil type on the toxicity of three pesticides Results Effects on survival No mortality was recorded in controls throughout the 28-day exposure period. Data on effects of chlorpyrifos, carbofuran and carbendazim on survival of E. andrei in different soil types and at both test temperatures are given as LC 50 values in Tables 1-3. The toxicity of the three pesticides in both conditions decreased in the order carbendazim > carbofuran > chlorpyrifos. LC 50 values for the effects of chlorpyrifos on the survival of Eisenia andrei in the test soils at both test temperatures (Table 1) show that chlorpyrifos was less toxic at 20 C, with LC 50 in AS and natural soil being 3.2 and 4.5 times higher, respectively at 20 C compared to 26 C. In addition, at 26 C the LC 50 was 1.5 times lower in the natural soil than in the AS. For carbofuran, the highest LC 50 was recorded in AS, with toxicity being a factor of 1.7 lower at 20 C than at 26 C (Table 2). In natural soil however carbofuran toxicity was 1.4 times lower at 26 C compared to 20 C (Table 2). In both AS and natural soil LC 50 s for the effects of carbendazim on the survival of E. andrei were higher at 26 C than at 20 C with differences being a factor of 5.4 times in AS and a factor of 15.3 in natural soil (Table 3). Table 1: Toxicity of chlorpyrifos to Eisenia andrei in OECD artificial soil (AS) and in a natural soil (NS) at two different temperatures. LC 50 values for the effect on survival and EC x values for the effect on reproduction are given with corresponding 95% CI. Also included are NOEC values for the effect on growth and reproduction. All values are in mg a.i. kg -1 dry soil. * Significantly different from the value for the comparable soil type at 20 C. Soil Type LC 50 EC 50 (Reproduction) EC 10 (Reproduction) NOEC NOEC (Reproduction) (Growth) AS (20 C) 492 (456-531) NS (20 C) AS (26 C) 481 (395-586) 154 * (134-177) 7.49 (5.14-9.84 ) 1.79 (1.23-2.36) 3.86 * (2.87-4.86) 0.48 (0.15 0.82) 0.90 <1 (0.18 1.6) 0.31 (0.13-0.48) <1 1 < 1 100 < 1 NS (26 C) 106 * (87.0-130) 5.87 * (4.18-7.57) 0.29 (0.10 0.47) <1 < 1 25

Chapter 2 Sub-lethal toxicity In the tests with chlorpyrifos and carbofuran, the number of juveniles was significantly different between controls and solvent controls (t-test, p < 0.05), hence the solvent controls were used in further data analysis and interpretation. The solvent controls of the tests with chlorpyrifos at 26 C produced the highest number of juveniles (102 129 juveniles per replicate) compared to 20 C and the lowest number was recorded in the natural soil at 20 C (26 36). Juvenile production in the control experiments with carbofuran was similar in all soil types (73 95 juveniles per replicate) and approximately 85-136 juveniles per replicate were recorded in the controls of the carbendazim except for the Lufa 2.2 natural soil (54 77 juveniles per replicate). Dose response relationships for the effects of three pesticides on the reproduction of E. andrei under different test conditions are presented in Figure 1. The toxicity was higher at 20 C than at 26 C in the natural soil. Chlorpyrifos was most toxic to earthworm reproduction in Lufa 2.2 natural soil and least toxic in AS at 20 C. The EC 50 for effects on the reproduction in AS was 1.9 times higher at 20 C than at 26 C (Table 1). The differences were statistically significant (χ 2 = 10.0, p < 0.001). In 1 natural soil, chlorpyrifos was 3.3 times more toxic at 20 C than at 26 C and EC 50 s were significantly different (χ 2 = 12.2, p < 0.001). At 26 C, 1 reproductive toxicity in natural soil was 1.5 times lower than in AS with Table 2: Toxicity of carbofuran to Eisenia andrei in OECD artificial soil (AS) and in a natural soil (NS) at two different temperatures. LC 50 values for the effect on survival and EC x values for the effect on reproduction are given with corresponding 95% CI. Also included are NOEC values for the effect on growth and reproduction. All values are in mg a.i. kg -1 dry soil. * Significantly different from the value for the comparable soil type at 20 C. Soil Type LC 50 EC 50 (Reproduction) EC 10 NOEC NOEC (Reproduction) (Reproduction) (Growth) AS (20 C) 23.4 (20.0-27.4) 1.57 (1.43-1.71) 0.66 (0.53 0.80) <0.5 < 0.5 NS (20 C) 8.46 (6.32-11.3) 0.74 (0.69-0.79) 0.49 (0.43 0.56) 0.5 0.5 AS (26 C ) 13.5 * (11.6-15.6) 1.25 * (1.17-1.33) 0.53 (0.45-0.60) <0.5 < 0.5 NS (26 C) 11.9 * (10.4-13.7) 1.96 * (1.66-2.26) 0.87 (0.55 1.19) <0.5 <0.5 26

Influence of temperature and soil type on the toxicity of three pesticides significantly different EC 50 values (χ 2 1 = 4.33, p < 0.05). At 20 C the highest toxicity difference was recorded in AS compared to natural soil (χ 2 1 = 14.3, p < 0.001) with toxicity being 4.1 times higher in the latter soill. Significant reduction of juvenile numbers was observed at all chlorpyrifos concentrations in AS at 20 C except for the lowest concentration, so NOEC is 1 mg a.i. kg -1 dry soil (Dunnett s test, p < 0.05). In AS at 26 C and in both natural soils juvenile production was significantly different from respective controls at all concentrations so NOEC is < 1 mg a.i kg -1 dry soil (Table 1). Data on the sublethal toxicity of carbofuran on reproduction of Eisenia andrei are given in Table 2. The highest toxicity was found in natural soil at 20 C and lowest toxicity in natural soil at 26 C. In AS, carbofuran was more toxic at 26 C than at 20 C and significantly different EC 50 values were found (χ 2 = 23.3, p < 0.001). In natural soil carbofuran 1 was more toxic (2.6 times) at 20 C than at 26 C (χ 2 = 46.0, p < 0.001). At 1 26 C, significantly lower toxicity was recorded in natural soil compared to AS (χ 2 = 21.6, p < 0.001). Juvenile production at all carbofuran 1 concentrations was significantly different (p < 0.05) from the control except for natural soil at 20 C. The LOEC and NOEC in the natural soil at 20 C were 1 and 0.5 mg kg -1 dry soil, respectively and 0.5 and < 0.5 mg kg -1 dry soil for the AS at 20 C and the AS and natural soil at 26 C. Table 3: Toxicity of carbendazim to Eisenia andrei in OECD artificial soil (AS) and in a natural soil (NS) at two different temperatures. LC 50 values for the effect on survival and EC x values for the effect on reproduction are given with corresponding 95% CI. Also included are NOEC values for the effect on growth and reproduction. All values are in mg a.i. kg -1 dry soil. * Significantly different from the value for the comparable soil type at 20 C. Soil Type LC 50 EC 50 (Reproduction) EC 10 (Reproduction) NOEC (Reproduction) NOEC (Growth) AS (20 C) 7.10 (5.79-8.79) 0.39 (0.11-0.65) 0.02 (0.0 0.6) <0.1 1 NS (20 C) 2.70 (2.20-3.30) 0.34 (0.20-0.46) 0.07 (0.02 0.13) 0. 1 1 AS (26 C) 38.9 * (33.6-45.1) 1.84 * (1.43-2.25) 0.28 (0.15 0.40) 0.1 < 0.1 NS (26 C) 41.5 * (36.3-47.5) 2.18 * (1.71-2.66) 0.52 (0.30-0.75) 0. 1 0.1 27

Chapter 2 The effect of carbendazim on reproduction of E. andrei is given in Table 3. The EC 50 in AS was 4.7 times higher at 26 C than at 20 C and difference was significant (χ 2 1 = 20.9, p < 0.001). The largest toxicity difference (χ2 1 = 58.1, p < 0.001) was found between natural soils tested at 20 and 26 C, with reproductive toxicity of carbendazim to E. andrei being 6.4 times lower at 26 C than at 20 C. A significant reduction of juvenile numbers (Dunnett s test, p < 0.05) was observed in both soil types with LOEC and NOEC values of 0.3 and 0.1 mg kg -1 dry soil, respectively for all soils except for AS tested at 20 C with a slightly lower NOEC (< 0.1 mg kg -1 dry soil). The biomass of adult earthworms exposed for 28 days showed an increase in all the controls by 5.9 % to 19.2 %. The only exception was the AS at 20 C in the carbendazim experiment, where a slight weight loss of 0.36% was observed. The weight loss in the controls must be considered an artifact, considering the fact that earthworms gained 14.3-16.4% weight at the lowest three carbendazim concentrations in this soil. The highest growth (> 18 %) in all three tests was recorded in the AS control soils kept at 26 C. Biomass increase of the controls at both temperatures did not significantly differ (t test, p > 0.05). Significant weight loss (p < 0.05) was observed with increasing pesticide concentration for all three pesticides irrespective of temperature and soil type. For chlorpyrifos and carbofuran NOEC values for growth were more or less similar except in natural soil kept at 20 C. But for carbendazim NOEC was higher at 20 C than at 26 C. Discussion This study mainly focused on comparison of the toxicity of pesticides to earthworms at two different temperatures, indicative of temperate and tropical conditions, using survival, growth and reproduction as the endpoints. No mortality was recorded in controls throughout the study period and control worms did not loose > 20 % of their initial biomass. In all experiments, each control replicate produced more than 30 juveniles except for natural soil at 20 C where only 26 juveniles were found. The coefficient of variation for reproduction was 30 %. This indicates the validity of the tests performed at both temperatures. For effects on survival, toxicity of chlorpyrifos in both artificial and natural soils was higher at 26 C than at 20 C. Ma and Bodt (1993) studied chlorpyrifos toxicity for E. fetida, a species closely related to E. andrei in AS and reported a 2-wk LC 50 of 1077 mg kg -1 dry soil, which is 2 and 10 times higher than the values found in our study at 20 and 26 C, respectively. The shorter test duration may explain this difference. 28

140 NS ( 26 0 C) NS ( 20 0 C) AS ( 20 0 C) AS (26 0 C) Number of juveniles 120 100 80 60 40 20 0-8 -6-4 -2 0 2 4 6-6 -4-2 0 2-8 -6-4 -2 0 2 4 6 Ln concentration (mg a.i kg -1 dry soil) Figure 1: Dose-response relationships for the effects of chlorpyrifos (left), carbofuran (middle) and carbendazim (right) on the reproduction of Eisenia andrei in artificial (AS) and natural soil (NS) at 20 C and 26 C.

Chapter 2 A lower LC 50 (118.5 mg kg -1 dry soil) was reported by Zhou et al. (2007) after 14 days exposure of E. fetida. The LC 50 of chlorpyrifos was independent of soil type at 20 C but a slightly higher toxicity was found in tropical natural soil. From the figures presented by Zhou et al. (2007), we derived an EC 50 for the effects of chlorpyrifos on juvenile numbers in AS of 9.75 mg kg-1 at 20 C, which is in agreement with our data for reproduction (Table 1). Toxicity of carbofuran did not differ much (less than a factor of 3 in LC 50 ) for the two temperatures tested. The LC 50 s are within the range of the values of 4 64 mg kg -1 dry soil reported by Heimbach (1985), Panda and Sahu (2000) and Haque and Ebing (1983). Gordon (2003) suggested that temperature could be one of the most critical factors in animal life. With limited information, Viswanathan and Krishnamurti (1989) suggested an increased toxicity of xenobiotics under tropical conditions. This may be supported by the effects on earthworm survival in AS spiked with chlorpyrifos and carbofuran. The uptake of pesticides by earthworms can be either through passive diffusion from pore water over the skin or by ingestion together with soil particles (Lord et al., 1980; Belfroid et al., 1994). It is expected that short-term toxicity endpoints, such as survival with organophosphates and carbamates depend on dermal uptake by the earthworms, which can be clearly seen as increased activity right after initial exposure. Tropical conditions, with higher temperatures, may enhance the activity of earthworms resulting in higher uptake of pesticides. This higher uptake could explain the higher toxicity of chlorpyrifos and carbofuran under tropical conditions when using artificial soil as the substrate. The only exception was the higher toxicity of carbofuran in natural soil at 20 C, which is more difficult to explain. This could be due to the complex nature of the natural soils with low organic matter and clay contents which can cause higher bioavailability of the pesticides. Sub-lethal endpoints like reproduction and growth did not give a clear image of the differences in toxicities at the different temperatures. Reproductive toxicity of chlorpyrifos and carbofuran in AS was higher at 26 C, which is in agreement with survival data. But toxicity was higher at 20 C than at 26 C in the natural soil (Figure 1). Our results therefore suggest that soil type and nature of the pesticide were more important for long-term toxicity than temperature. Similar inconsistent trends in toxicity were also observed for growth. A lower sub-lethal toxicity might be related to degradation kinetics. The degradation of the two pesticides could mainly be due to chemical and microbial processes and could be favored by the higher temperature in tropical conditions (Racke et al., 1997). Getzin (1981), Racke and Coats (1988) and Singh et al. (2003) reported a significant contribution of soil microorganisms to degradation of organophosphorus insecticides in natural soil. A higher microbial activity in natural soils under tropical than under temperate conditions therefore could 30

Influence of temperature and soil type on the toxicity of three pesticides result in faster biodegradation resulting in lower sub-lethal toxicities. It is also noted that for all three pesticides toxicity in the natural soil was higher at 20 C than at 26 C. The organic matter content of the Lufa 2.2 natural soil (~ 4 %) was lower than that of the tropical natural soil (~ 9 %), which might result in lower adsorption and increased bioavailability of the pesticide to the organisms resulting in higher toxicity. This was e.g. illustrated by Van Gestel and Ma (1988), who have also reported that earthworm toxicity is higher in soils with low organic matter content. A deviation from the higher toxicities at 26 C was observed for carbendazim with high LC 50 and EC 50 values indicating lower toxicity at higher temperatures irrespective of the soil type used. Adema et al. (1985) using E. fetida, Van Gestel (1992) and Federschmidt (1994) using E. andrei in AS at 20 C reported 2-week LC 50 values similar to this study. Data of the effects of carbendazim on earthworm survival in soils tested at higher temperatures are scarce. Garcia (2004) reported a very high LC 50 (> 1000 mg a.i. kg -1 dry soil, 14 days) in AS at 28 C compared 20 C (5.8 mg a.i. kg -1 dry soil), resulting a factor of 172 difference in toxicity between the two temperatures. We found a lower 4-week LC 50 (38.9 mg a.i. kg -1 dry soil) in AS at 26 C compared 20 C (7.1 mg a.i. kg -1 dry soil) and the difference was only a factor of 5.48. Garcia (2004) also determined effects of carbendazim on the reproduction of E. fetida at 20 and 28 C in AS and Lufa 2.2 natural soil. Toxicity was much higher at 20 C, with EC 50 values of 2.7 and 0.6 mg kg -1 in AS and natural soil compared to 14.2 and 9.6 mg kg -1, respectively at 28 C. Although Garcia (2004) used E. fetida as the test species it is difficult to explain the toxicity difference between two studies due to the similarities of both species and similar toxicity values were found with natural soil tested at the higher temperature. It should be noted that Garcia (2004) used two different strains of E. fetida, while we used the same strain but adapted to different temperatures. The general observation of lower toxicity of carbendazim at 26 C could be due to temperature effects on the stability of the chemical, which could have reduced the exposure concentration with time. Conclusions The effects of chlorpyrifos and carbofuran on earthworm survival, growth and reproduction in artificial soil may be higher at higher temperature (indicative of tropical conditions) than at the standard temperature of 20 C, even when tested in similar soils and within the same species. Survival, reproduction and growth may, however, differ in natural soil at both temperatures tested. Carbendazim toxicity was lower at high temperatures irrespective of the soil types. Overall there is no clear trend of increased toxicity for sub-lethal endpoints with temperature. Toxicity may vary with the pesticide, endpoint, soil type and temperature. We therefore conclude 31

Chapter 2 that earthworm toxicity data obtained at 20 C should only be used with caution for the assessment of ecological risks of chemicals under tropical conditions. Acknowledgements This work was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB84-609) residing under the Netherlands Organization for Scientific Research (NWO). We thank Rudo Verweij and H.G. Hector for technical assistance. 32

Chapter 3 Toxicity of Chlorpyrifos, Carbofuran, Mancozeb and Their Formulations to the Earthworm Perionyx excavatus P. Mangala C. S. De Silva, Asoka Pathiratne, Cornelis A.M. van Gestel Applied Soil Ecology (Accepted) 33

Chapter 3 Abstract Effects of chlorpyrifos, carbofuran, mancozeb and their formulated products on survival, growth and reproduction of the tropical earthworm Perionyx excavatus were investigated in standard artificial soil under tropical conditions. Toxicity of the three chemicals decreased in the order carbofuran > chlorpyrifos > mancozeb. In general, formulations were more toxic than the active ingredients, but differences in LC 50 and EC x values were significant only in two cases and not larger than a factor of 2.0. This could mainly be due to masking of the effects of additives in the soil. Comparison with available literature data revealed that P. excavatus is more sensitive than the standard species Eisenia sp. The use of tropical species in the risk assessment of pesticides in tropical regions should therefore be encouraged. Key words: Earthworms, Pesticide formulations, Perionyx excavatus, Additives, Active ingredients 34

Toxicity of Pesticides to Perionyx excavatus Introduction Tropical ecosystems cover a relatively small fraction (26 %) of the total land area of the world. But the primary productivity is as high as 60 % and indicates the great importance of the existence of these systems in the world (Deshmukh, 1986). Increasing productivity in the tropics and resulting human activities have caused serious threats to tropical ecosystems. The degradation of terrestrial and aquatic ecosystems due to xenobiotics is a major concern and is a direct result of the increased use of synthetic substances such as pesticides to boom the productivity of the region. Nevertheless, risk assessment of pesticides in the tropics is hardly performed due to scarcity of data. And a limited number of tropical risk assessments currently available mainly rely on data generated in temperate regions. But species sensitivity, fate of xenobiotics and complex interactions within the system could be different in the tropical environment. As a consequence, risks may be over or under estimated when using additional extrapolation factors. Sensitivity comparisons of tropical and temperate species have so far shown contradictory results. Maltby et al. (2005) and Kwok et al. (2007) have reported no large differences in the sensitivity of species belonging to the same taxonomic groups in aquatic ecosystems, except for a few pesticides like chlorpyrifos. But studies by Garcia (2004) and De Silva et al. (2009) have shown that the situation in soil ecosystems could be different with more factors like nature of the pesticide, temperature, soil type, endpoints measured and their interactions causing different effects. Species diversity and abundance of soil invertebrates like earthworms could also vary under tropical conditions; hence prediction of field effects will be rather difficult. Toxicity assessment of pesticides to earthworms mainly uses Eisenia sp as the standard test species (OECD, 1984; OECD, 2004). As compost worms, Eisenia species have a limited ecological role compared to local mineral-dwelling species. In addition, they are mainly occurring in temperate regions. Therefore the use of tropical species in toxicity tests could contribute to a more relevant and reliable risk assessment of chemicals in these areas. Perionyx excavatus (Perrier, 1872), Family Megascolecidae, is primarily a compost worm that can be found in tropical regions, especially in Asia (Blakemore, 2006) and also is present in Europe and North America (Edwards et al., 1998). Generally they are found in the top soil layer (0-15 cm) at temperatures ranging between 20 C and 28 C and ph of 6.4-7.4 (Bhattacharjee and Chaudhuri, 2002). The life cycle and biology of this species have been extensively studied (Reinecke and Hallatt, 1989; Hallatt et al., 1990; Edwards et al., 1998; Bhattacharjee and Chaudhuri, 2002; Chaudhuri and Bhattacharjee, 2002; Joshi and Dabral, 2008). Previously P. 35

Chapter 3 excavatus has been used in ecotoxicity studies including metals (Maboeta et al., 1999; Reinecke et al., 2001), pesticides (Callahan et al., 1994; Reinecke et al., 2002) and gasoline products (An, 2005; An and Lee 2008). Pesticide testing has usually been carried out by generating data on the active ingredients. However farmers apply formulated products, which can be a combination of active ingredients and additives. Some additives are added to enhance toxicity of the pesticide e.g. by affecting uptake or metabolism of the active ingredient. Specific assessment of additives of the formulated products has been hampered due to trade secrets (Cox and Surgan, 2006). Alternatively, toxicity of such ingredients could be derived from studying formulated products together with their active ingredients. Comparative studies of formulated and pure pesticides in birds (Arias, 2003), frogs (Swann et al., 1996; Mann and Bidwell, 1999) and fish (Kiparissis et al., 2003) have reported higher toxicities of formulations compared to their respective active ingredients. Only few studies have so far been performed to investigate the comparative toxicity of formulations and pure chemicals in the soil ecosystem (Salminen et al., 1996; Marques et al., 2009; Pereira et al., 2009). The aim of this study was to determine the toxicity of three commonly used pesticides namely chlorpyrifos, carbofuran and mancozeb together with their formulated products to the tropical earthworm species P. excavatus under tropical conditions and to compare the relative toxicities of pure and formulated products. In addition, the generated toxicity data for the tropical species were compared with the data available in the literature for the standard test species Eisenia andrei and E. fetida. Materials and Methods Age-synchronized adult earthworms (250-275 mg, wet weight) of the tropical species P. excavatus (Family: Megascolecidae) were obtained from the cultures at the Department of Zoology, University of Ruhuna, Matara, Sri Lanka. These cultures were maintained at a temperature representative of tropical conditions (26 ± 2 C) in a substrate of 30 % cow dung and 70 % composted coco peat with 12:12 h photoperiod. The selected worms were transferred to untreated test substrate 24 hours before the experiment, to acclimatize. The standard OECD artificial soil (OECD, 1984) was used as the test soil substrate. The OECD soil was composed of 70 % fine sand, 20 % kaolin and 10 % sphagnum peat with a small amount of CaCO 3 added for ph adjustment. This soil, which acts as the reference substrate for toxicity studies, had a ph (0.01M CaCl 2 ) of 6.5 6.8 and a maximum water holding capacity (WHC max) of 80%. Toxicity tests were performed with chlorpyrifos (98 %, Cheminova Ltd, Denmark), carbofuran (97 %, Sigma Aldrich, The Netherlands) and mancozeb (85 % Limin Chemical Co Ltd, China) together with their 36

Toxicity of Pesticides to Perionyx excavatus commercial formulations Judo 40 EC (40 % a.i. EC, Lankem Ceylon Ltd, Sri Lanka), Curater (3 % a.i. G, Hayleys Agro Products Ltd, Sri Lanka) and Dithane M 45 (80 % a.i. WP, Chemical Industries Colombo Ltd, Sri Lanka), respectively. Concentrations tested were 1, 3, 10, 30, 100, 300 and 900 mg a.i. kg -1 dry soil of chlorpyrifos, 0.5, 1, 2, 4, 8, 16 and 32 mg a.i. kg - 1 dry soil of carbofuran and for mancozeb 1, 3, 10, 30, 100, 300, 900 and 1200 mg a.i. kg -1 dry soil. For the pure compounds, solutions of the test compounds in acetone were added to 50 g soil. These soil samples were kept overnight ensuring evaporation of the solvent and the remainder of the soil (450 g) for each treatment was added and mixed intensively with the spiked soil. Control soils were treated with acetone in a similar way. Pesticide formulations of chlorpyrifos and mancozeb were directly added to the substrate as aqueous solutions, whereas carbofuran 3% G was added as finely ground powder. Soil moisture content was adjusted to 50 % WHC max. The control soils were treated only with water. Earthworm toxicity tests were performed according to OECD (2004a) with slight modifications. Glass bottles (750 ml) were filled with approximately 710 g (wet weight) of the test soils. Earthworms (n = 10, 4 replicates) were introduced into the soil, and the lids were not closed tightly to facilitate free exchange of air. Before introduction, the worms were rinsed with water, blotted dry on filter paper, and mass were determined (per ten worms or individually for a number of worms). The test containers were kept at 26 ± 2 C and 12:12 h photoperiod. Five grams (dry weight) of food (finely ground cow manure) moistened to 50% w/w was added in a hole made in the middle of the soil. Additional food was given when all the food added was consumed. Moisture loss from the test soils was checked by weighing the test containers at weekly intervals and replenished if needed. After 28 days of exposure, adults were removed by hand sorting, and mortality and biomass were determined. The soil was returned into the test containers and incubated for another 28 days for cocoon development. After 56 days, juveniles were extracted from the test soil using a water bath kept at 60 C and counted. The final endpoints studied were adult mortality, change of biomass after 28 days, and number of juveniles produced after 56 days. Levene s test and Kolmogorov-Smirnov test were performed to assess homogeneity of variance and normal distribution of the data, respectively. LC 50 values, the concentrations causing 50% mortality of the earthworms and corresponding 95% confidence limits were calculated with the Trimmed Spearman Karber method (Hamilton et al., 1977). The model for logistic response of Haanstra et al. (1985) was used for the calculation of concentrations affecting reproduction by 50% (EC 50 ) and 10% (EC 10 ) and the corresponding 95% confidence limits. Lowest Observed Effect Concentrations (LOEC) and No Observed Effect Concentrations (NOEC) values were determined using ANOVA and Dunnett s test by SPSS version 37

Chapter 3 14.0. Differences in EC 50 values between pure compounds and the formulations were tested for significance using a generalized likelihood ratio test (Sokal and Rohlf, 1995). Results No mortality was observed in the controls except for a few individuals (one each in three replicates) missing in the mancozeb series. The LC 50 values for the effects of chlorpyrifos, carbofuran, mancozeb and their formulations on survival of P. excavatus are given in Table 1. Toxicity decreased in the order of carbofuran > chlorpyrifos > mancozeb for both the pure compounds and the formulations. The highest LC 50 values were recorded for pure mancozeb and its formulation and the lowest ones for carbofuran. The toxicities of the pure and formulated pesticides were similar with overlapping 95% confidence intervals and differences in LC 50 values being less than a factor of 1.2. Mean number of juveniles produced in the controls varied for the different tests and ranged between 80 and 92 per test container. Juvenile production in solvent controls (112-130 juveniles per replicate) was significantly higher (p < 0.05) than of their respective controls and therefore solvent controls were used in further data analysis for the pure compounds. The dose-response relationships for the effects on reproduction of P. excavatus are given in Figure 1 and the estimated values for EC 50, EC 10 and NOEC for reproduction are presented in the Table 1. Formulated carbofuran was more toxic to reproduction of P. excavatus than its pure compound and EC 50 values were significantly different (χ 2 1 = 37.2, p < 0.001) from each other. The EC 50 value for pure chlorpyrifos was 1.3 times higher than for its formulation but the difference was not significant (χ 2 1 = 0.63, p > 0.05). No significant difference of EC 50 values was found for mancozeb (χ 2 1 = 1.96, p > 0.05) although the EC 50 was 1.7 times higher for the formulation than for the pure compound. Similar trends for the pure and formulated products were found for EC 10 values (Table 1). However EC 10 values for chlorpyrifos were lower than those for carbofuran even though carbofuran was the most toxic pesticide in terms of LC 50 and EC 50 values. The NOEC was always lower than the lowest concentration tested for all pure and formulated pesticides (Table 1). The adult P. excavatus in the controls gained 16 % - 22% of weight during the 4- week test period. In general an increasing and significant weight loss was recorded with increasing concentrations for all pesticides and their formulations (Figure 2) and NOEC values for biomass change were similar to those for reproduction (Table 1). 38

Toxicity of Pesticides to Perionyx excavatus Table 1: Toxicity of pure and formulated products of chlorpyrifos, carbofuran and mancozeb to the earthworm Perionyx excavatus in OECD artificial soil under tropical conditions. LC 50 values (28 days) for the effect on survival and EC x values for the effect on reproduction are given with corresponding 95% CI. Also included are NOEC values for effects on the reproduction (Rp) and growth (Gr). All values expressed are in mg a.i. kg -1 dry soil. * indicates significant difference from the respective value for the pure chemical. Pesticide Type LC 50 EC 50 EC 10 NOEC (Rp / Gr) Chlorpyrifos Pure 122 (104-145) 4.0 (2.2-5.7) 0.27 (0.02-0.54) < 1 Formulation 100 3.0 0.16 < 1 (84-120) (2.4-3.7) (0.08-0.24) Carbofuran Pure 9 (8-10) 1.7 (1.6-2.0) 0.60 (0.47-0.72) < 0.5 Formulation 8 1.1* 0.30 < 0.5 (7-9) (1.0-1.2) (0.24-0.35) Mancozeb Pure 541 (497-590) 37.4 (18.6-56.2) 0.80 (0.17-1.7) < 1 Formulation 500 (460-544) 22.0 (11.6-32.0) 0.68 (0.25-1.38) < 1 Discussion This study focused on the toxicity of chlorpyrifos, carbofuran and mancozeb, both pure and formulated products, on the survival, reproduction and growth of the tropical earthworm species P. excavatus under tropical conditions. Mortality in the controls ( 10 %), loss of weight ( 20 %), juvenile production per replicate ( 30) and coefficient of variation of reproduction ( 30 %) reported in the tests indicates the validity of the tests performed with P. excavatus. These validity criteria were mainly set for E. fetida and E. andrei in standard guidelines by OECD and ISO. Nevertheless these criteria can also be applied to the tropical species P. excavatus. In this study, ph of the OECD artificial soil was slightly higher than the range described by the test guidelines (ph-cacl 2 6.5-6.8 instead of 6.0±0.5). Considering the ecological preferences of P. excavatus (abundant in soils with ph values up to 7.4) and the nature of the pesticides tested (nondissociating), this small deviation of soil ph is not expected to have any effects on the outcome of the tests performed. For all three pesticides, formulations were more toxic than the active ingredients although the differences were small. Chlorpyrifos was moderately toxic with 4-week LC 50 values of 100 and 122 mg a.i. kg -1 dry soil. The toxicity of chlorpyrifos to P. excavatus was 8.8 times higher than 39

Chapter 3 for Eisenia fetida (14-days LC 50 = 1077 mg a.i. kg -1 dry soil artificial soil; Ma and Bodt, 1993) and only 1.2 times higher than for E. andrei under similar conditions (28-days LC 50 = 154 mg a.i. kg -1 dry soil artificial soil; De Silva et al., 2009). Toxicity of chlorpyrifos to earthworm species varies greatly as reported, for example, by Ma and Bodt (1993), with 14-day LC 50 values ranging between 129 and 1174 mg a.i. kg -1 dry soil and ecologically relevant species like Lumbricus rubellus and L. terrestris being more sensitive than Eisenia sp. This is in accordance with this study reporting higher sensitivity of P. excavatus under tropical conditions than Eisenia species. For carbofuran, LC 50 values found in this study are within the range of values reported for E. andrei (5-10 mg a.i. kg -1 dry soil; Heimbach, 1985) and toxicity is 1.6 times higher for P. excavatus than for E. andrei under similar conditions (De Silva et al., 2009). Vermeulen et al (2001) found relatively low toxicity of mancozeb to E. andrei with a 14 - day LC 50 of 1262 mg a.i. kg -1 dry soil. This study reports a 2.3 times higher toxicity of mancozeb to P. excavatus than the E. andrei. Reinecke et al. (2002) showed that P. excavatus could avoid low concentrations of mancozeb (8 mg a.i. kg -1 dry soil), which also confirms that P. excavatus is more sensitive to mancozeb than Eisenia species. Higher sensitivity of P. excavatus has also also been reported with methyl tert-butyl ether (MTBE) in tests on filter paper and in two different soils (An, 2005). In contradiction to the survival data, available EC 50 values for chlorpyrifos and carbofuran under similar conditions show that effects on reproduction were slightly higher for E. andrei than for P. excavatus (De Silva et al., 2009). In addition, toxicity assessment of formulations of the pesticides, which include additives together with the active ingredient, will be essential. Components of the formulations are not to be considered inert ingredients, as they may have chemical and biological activity and could be toxic. This is further confirmed by the fact that more than 500 additives are currently also being used as active ingredients (Cox and Surgan, 2006). Results obtained for effects on survival, reproduction and growth of P. excavatus indicated no difference in toxicity except for two cases. This observation contradicts with most of the aquatic studies that compared toxicity of formulations and their active ingredient alone (Swann et al., 1996; Mann and Bidwell, 1999; Kiparissis, 2003). In aquatic environments increased toxicity of formulations could be linked with higher solubility of the active ingredients in the medium. Aquatic organisms could be exposed to chemicals through the body surface or gills and additives might affect the uptake sites e.g due to their effects on the surface tension or structure of membranes. Such an affect might also occur in case of soil organisms like earthworms that also take up chemicals from pore water via their skin (Belfroid et al., 1994). 40

Pure Formulation 140 Number of juveniles 120 100 80 60 40 20 0-8 -6-4 -2 0 2 4 6 8-6 -4-2 0 2-8 -6-4 -2 0 2 4 6 Ln concentration (mg a.i kg -1 dry soil) Figure 1: Dose-response relationships for the effects of chlorpyrifos (left), carbofuran (middle) and mancozeb (right) and their formulated products on the reproduction of Perionyx excavatus in OECD artificial soil under tropical conditions.

Pure pestiside Formulation 40 20 Mean biomass change (%) 0-20 -40-60 -80-100 -120 0+ 1 3 10 30 100 300 900 0+ 0.5 1 2 4 8 16 32 0+ 1 3 10 30 100 300 900 1200 Pesticide concentration (mg a.i kg -1 dry soil) Figure 2: Mean biomass change (%) of Perionyx excavatus after 28-day exposure to chlorpyrifos (left), carbofuran (middle) and mancozeb (right) and their formulated products in OECD artificial soil under tropical conditions.

Toxicity of Pesticides to Perionyx excavatus In soil, additives may however, easily be bound to soil particles, like clay and organic matter, reducing their availability. This masking of additive effects through sorption probably explains the small differences in the toxicity of formulations and pure compounds. Recent studies by Marques et al. (2009) and Pereira et al. (2009) with herbicide formulations reported either underestimated or overestimated toxicity of the active ingredients. The lack of studies on additives and their interactions with soil hampers our understanding of their behaviour and the way in which they can affect toxicity. Additional research in soils with different pesticides and their additives is needed to enhance further interpretation. Species sensitivity is a critical factor in the risk assessment of pesticides in soil, but in particular for tropical soils with their higher diversity and abundance of earthworm species compared to temperate soils. Nevertheless, current tropical risk assessment mainly uses data generated in tests with the temperate compost worm Eisenia sp, which is less ecologically relevant because it is not a soil-dwelling species. Actually, the same limitations concerning its ecological role are also valid for the tropical compost worm P. excavatus. However, this study as well as literature data (Maboeta et al., 1999; Reinecke et al., 2001; An and Lee, 2008) show that P. excavatus may be used in earthworm toxicity tests as a standard tests species for tropical soils. This recommendation is based on the same reasons as in the case of Eisenia sp., which was successfully identified as temperate standard test species about 25 years ago: The biology and ecology of this species is well known and therefore it can easily be cultured under laboratory conditions. Although the life cycle of P. excavatus is shorter than that of Eisenia sp, current test guidelines can easily be applied. The preference of P. excavatus for optimum temperatures around 25 C makes it more suitable for tropical regions than Eisenia sp Conclusions Chlorpyrifos, carbofuran and mancozeb were more toxic to P. excavatus than to the standard species E. andrei under tropical conditions. Tropical risk assessment will therefore be more realistic when it is done with ecologically relevant earthworm species. Toxicity of formulations and their respective active ingredients was more or less similar probably due to masking of the effects of additives by sorption to the soil. Additional research on both aspects is needed to arrive at more definite general conclusions. P. excavatus may be used as an alternative test species under tropical conditions. Future studies with this species, including different endpoints, temperature regimes and soil types, are encouraged. 43

Chapter 3 Acknowledgements This work was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB 84-609) residing under the Netherlands Organization for Scientific Research (NWO). We thank H.G. Hector and C.H. Priyantha for technical assistance. 44

Chapter 4 Comparative Sensitivity of Eisenia andrei and Perionyx excavatus in Earthworm Avoidance Tests Using Two Soil Types in the Tropics P. Mangala C. S De Silva, Cornelis A.M. van Gestel Chemosphere (Accepted) 45

Chapter 4 Abstract Terrestrial avoidance behaviour is proposed as a fast and cost-effective method for assessing effects of pesticides on earthworms. Tropical species however, have rarely been used in avoidance tests. Avoidance tests were performed with Perionyx excavatus, a tropical species, and Eisenia andrei as the standard species, using chlorpyrifos and carbofuran in artificial and natural soil. Earthworms were exposed to concentrations of 1-900 (chlorpyrifos) and 1-32 (carbofuran) mg a.i. kg -1 dry soil in a two-chamber system under tropical conditions (26 ± 2 C, 48 hrs). No significant difference was found in the control tests comparing the two soils used, suggesting soil type did not affect the distribution of the worms. The results suggest a higher sensitivity of E. andrei, with EC 50 s for the effect on avoidance behaviour for both pesticides being factors of 2-3 lower than for P. excavatus. Earthworm avoidance tests with local species should therefore be used with caution when applied as a tool for pesticide risk assessment in the tropics. Endpoints generated through avoidance tests in this study are shown to be less sensitive than reproduction and more sensitive than survival. This was further confirmed by literature data available. Earthworm avoidance tests therefore can only replace survival tests as an initial screening tool for risk assessment. Key words: Risk assessment, Tropical, Chlorpyrifos, Carbofuran, Behaviour, Sri Lanka 46

Comparative sensitivity of earthworms in avoidance tests Introduction The ever growing demand for food in the world has been closely linked with intensifying the agriculture practices to facilitate higher crop production. Increased use of pesticides in agro-ecosystems is one of the most critical consequences of this enhanced food production. This becomes most obvious especially in the tropics, where the pesticide market has dramatically increased over the last decade. This also has resulted in an increase of pesticide residue levels in the soil. Soil supports a wide range of beneficial flora and fauna, which may be at risk due to this practice. Although the risk assessment of pesticides should be part of sustainable agriculture, pesticide regulations are hard to implement in the tropics due to various problems, like lack of data or data applicable to local conditions, political climate, absence of regulatory framework etc. The potential risk of chemicals for the habitat function of soils is often investigated applying acute (OECD, 1984; ISO, 1993) and reproduction tests (ISO, 1998b; OECD, 2004a) using earthworms as representatives of the soil biocenosis. These tests are very popular but mainly designed for temperate regions. Nevertheless, these methods are now also being used to asses the effects of pollutants in tropical areas. Constraints such as the inability to assess population effects with acute toxicity tests and also the long duration (56 days) and the labor intensive nature of the reproduction test called for rapid assessment methods with shorter duration retaining high sensitivity and ecological relevance. The fact that animals are able to move is important for their potential to colonize new habitats, to find food and to find a conspecific for mating. Several indications exist that chemicals may affect animal behaviour. This may also trigger avoidance of contaminated parts of the habitat, therefore providing the animal with the possibility to reduce its exposure to pollutants. Behaviour therefore is suggested as a sensitive and relevant alternative endpoint for toxicity tests with soil organisms, like earthworms, enchytraeids and Collembola (Hund-Rinke and Wiechering, 2001; Heupel, 2002; Natal da Luz et al., 2004; Schaefer, 2004; Loureiro et al., 2005; Lukkari and Haimi, 2005; Aldaya et al., 2006; Barateiro Diogo et al., 2007). The International Standardization Organization (ISO, 2007) and Environment Canada (2004) therefore drafted guidelines to determine the habitat function of the soil through an earthworm avoidance test. The avoidance test is a complementary screening test in soil risk assessment (Slimak, 1997; Stephenson et al., 1998; Hund-Rinke et al., 2002). It has a very simple test design and relatively short test period, which makes it an ideal rapid assessment of pollutants in the soil (Yeardley et al., 1996). It has been suggested that its sensitivity is comparable to that of the earthworm reproduction test (Hund-Rinke et al., 2003). Avoidance tests use the presence of chemoreceptors on the anterior segments and sensory tubercles 47

Chapter 4 on the earthworm s body surface, which can detect a wide range of contaminants (Reinecke et al., 2002). The short duration and the low costs associated with the earthworm avoidance test make it suitable for tropical risk assessment of pesticides. Most of the studies so far carried out use Eisenia andrei as the test species in temperate conditions as recommended by ISO (2007). Eisenia spp. However, is a compost or manure worm, rather difficult to find in natural ecosystems and therefore may not be ecologically relevant for natural earthworm species. Several alternative species with higher ecological relevance, such as Lumbricus terrestris (Schaefer, 2001), Aporrectodea calignosa (Hodge et al., 2000), L. rubellus (Lukkari & Haimi, 2005) and Dendrobaena octaedra (Lukkari et al., 2005), have been shown to be suitable for use in avoidance tests. In spite of the potential suitability of earthworm avoidance tests in the tropics, so far the only study performed was by Garcia et al. (2008) who used a tropical variant of E. fetida as the test species. Therefore tests with alternative species in different exposure conditions will be essential in further risk assessment. This will also contribute to risk assessment of pesticides in the tropics, which in general is severely hampered by a lack of data. The aim of this study was to compare the response of the standard test species E. andrei with that of an ecologically relevant tropical species, Perionyx excavatus, in avoidance tests using the standard artificial soil and a natural soil as test substrates under tropical conditions. Materials and Methods Adult Eisenia andrei (420-440 mg wet weight) and Perionyx excavatus (250-270 mg wet weight) with well developed clitellum were obtained from synchronized cultures at the Department of Zoology, University of Ruhuna, Matara, Sri Lanka. E. andrei originated from cultures at the Vrije Universiteit Amsterdam and had been cultured in Sri Lanka for about 1 year before starting these experiments. The selected worms were transferred to the two test substrates, 24 hours before the experiment, to acclimatize. Artificial soil (OECD soil) and natural soil were used as the test soil substrates. OECD soil was composed of 70 % fine sand, 20 % kaolin, 10 % sphagnum peat with a small amount of CaCO 3 for the adjustment of the ph. This soil, which acts as the reference substrate for toxicity studies, had a ph (0.01M CaCl 2 ) of 6.5-7.2 and a maximum water holding capacity (WHC max) of 80%. The natural soil was collected from a site near Matara where the uppermost soil layer (5 cm) was excavated. The soil was air dried at 26 ± 2 0 C in the laboratory and sieved (2 mm) to obtain a homogeneous soil mixture. The ph (0.01M CaCl 2 ) of this soil was 6.9-7.2; it contained 9% organic carbon, 85 % sand, 5.5% clay, 4.5% silt and had a WHC max of 45%. 48

Comparative sensitivity of earthworms in avoidance tests Chlorpyrifos (98%, Cheminova Ltd, Denmark) and carbofuran (97%, Sigma Aldrich), two pesticides widely used in tropical countries like Sri Lanka, were used in the experiments. Nominal chlorpyrifos concentrations of 1, 3, 10, 30, 100, 300, 900 mg a.i. kg -1 dry soil and carbofuran concentrations of 1, 2, 4, 8, 16, 32, 64 mg a.i. kg -1 dry soil were prepared using solutions in acetone. These acetone solutions were added to 50 g soil and kept overnight ensuring evaporation of the solvent. After evaporation of the acetone, the remainder of the soil for each treatment was added and mixed intensively with the spiked soil. Next, water was added to 50% of WHC max. Control soils were treated with acetone in a similar way, and also moistened to 50% of WHC max. The experimental procedure of the avoidance tests was based on ISO (2007). The test containers were plastic trays (30 22 6 cm depth) each with two sections. One side of each container was filled with control soil (500 g dry soil) and the other side with pesticide-treated soil. The two sides were separated by a thin plastic sheet to prevent mixing of the two soils. The random distribution of the worms was confirmed by performing dual control tests, which consisted of pesticide-free soil at both sides. After the preparation of the containers, the separator (thin plastic sheet) was removed and 10 adult earthworms were introduced into the separation line. This procedure was followed for all tested concentrations, each with four replicates. Dual control tests were performed with the same control soil on both sides, but also to compare the two tested soils (OECD versus natural soil). Initially the containers were kept in the light for no more than one hour to force the earthworms into the soil. Then the containers were covered with perforated transparent lids to facilitate gaseous exchange between the medium and the atmosphere and to facilitate the access of light. After that the test containers were incubated in darkness to avoid the lateral effects of light. The initial and final soil moisture contents were determined for each soil in all test containers. The containers were incubated for 48 hours at 26 ± 2 0 C, after which period the presence of worms at both sides was determined by hand sorting. The worms found on the separating line were counted by locating the direction of the head. For each replicate the avoidance response was calculated using the equation given below. Where, NR = [ (C - T) / N] 100 NR = Net avoidance response (%) C = Number of worms in control soil T = Number of worms in pesticide-amended soil N = Total number of worms exposed 49

Chapter 4 Avoidance is indicated by a positive NR value, non-response by zero and attraction by a negative NR. Validity of the avoidance test was checked by the random distribution of the worms in dual control tests, where avoidance behaviour should be absent, judged by the proportion of the earthworms in both sides being not significantly different (student t - test, p > 0.05). The tested soils were considered to have limited or reduced habitat function when 80% of the worms stayed in the control soil compared to the pesticide-treated soil. The NR value of each exposure concentration was used to determine the median effect concentration (EC 50 ) using a logistic model according to Haanstra et al. (1985). In case of attraction at lower concentrations, a modified logistic model with correction for hormesis was used (Van Ewijk and Hoekstra, 1993). The Lowest Observed Effect Concentration (LOEC) and the No Observed (Adverse) Effect Concentration (NO(A)EC) were estimated using ANOVA and Dunnett s test. Results No dead or missing worms were found in the test, except for 100 % mortality at the highest concentrations of chlorpyrifos (900 mg a.i. kg -1 dry soil) and carbofuran (64 mg a.i. kg -1 dry soil) in both OECD and natural soils. Hence these concentrations were excluded from the statistical analysis of avoidance behaviour. The dual control tests, which used control OECD and natural soils in alternate sides, did not show any significant preference or aggregation to one side (p > 0.05) and mean distribution was 50 ± 10 % in all replicates. The dual control tests with OECD soil and natural soil in the same trays also reported no significant preference (p > 0.05) to a specific soil type. For both chemicals, Perionyx excavatus was significantly attracted at the lowest test concentrations. For that reason we had to use a modified logistic model to calculate EC 50 values for avoidance behaviour. This resulted in rather wide confidence intervals for the EC 50 values (Table 1), even though fit generally was good with R 2 > 0.87. The effects of chlorpyrifos on the avoidance behavior of both species in OECD and natural soil are given in Tables 1 and 2. P. excavatus was significantly attracted by chlorpyrifos at the lowest three concentrations (1-10 mg a.i. kg -1 dry soil); such an effect was not seen for Eisenia andrei (Table 2). When neglecting this negative avoidance behaviour and focusing on the avoidance of chlorpyrifos in OECD artificial and natural soil, E. andrei was 2.5 and 2.0 times more sensitive, respectively, than P.excavatus. For both species, the avoidance behaviour was similar in both OECD and natural soil. NOAEC values for E. andrei and P. excavatus in both soil types were 1 and 10 mg a.i. kg -1 dry soil, respectively. Habitat function ( 80% of the worms in the control soil) was reduced at chlorpyrifos 50

Comparative sensitivity of earthworms in avoidance tests Table 1: Median effect concentrations (EC 50 ) with corresponding 95% confidence intervals and No-Observed Adverse Effect concentrations (NO(A)EC) for the effects of chlorpyrifos and carbofuran on the avoidance behaviour of Eisenia andrei and Perionyx excavatus in OECD artificial soil and a natural soil (effect concentrations are based on nominal concentrations expressed in mg a.i. kg -1 dry soil). Tests were performed in a two-compartment system. Species Soil type Chlorpyrifos Carbofuran Eisenia andrei Perionyx excavatus OECD Natural OECD Natural EC 50 NO(A)EC EC 50 NO(A)EC 9.31 2.15 1.0 (2.57-16.0) (2.01-2.29) 1.0 12.1 2.44 1.0 (8.37-15.9) (2.15-2.74) 1.0 23.5 4.39 10 (2.82-44.0) (0.70-8.05) 2.0 23.9 5.47 10 (-) (-) 2.0 Table 2: Avoidance response of Eisenia andrei and Perionyx excavatus to chlorpyrifos in OECD artificial soil and a natural soil. * and *** significant avoidance compared to control with p < 0.05 and p < 0.001, respectively. indicates reduced habitat function (>60% avoidance). Mean net response (% +/- SD) Concentration (mg a.i.kg -1 dry soil) Eisenia andrei Perionyx excavatus OECD Natural OECD Natural 0 0 (16.3) 5 (19.1) -5 (10.0) 5 (19.1) 1 20 (16.3) 20 (32.6) -70 (11.5) -70 (11.5) 3 30 (11.5) *** 20 (23.0) * -55 (25.1) -55 (34.1) 10 35 (10.0) *** 45 (19.1) *** -50 (11.5) -45 (19.1) 30 85 (19.1) *** 95 (10.0) *** 55 (30.0) *** 60 (1 100 100 *** 100 *** 100 *** 95 (1 300 100 *** 100 *** 100 *** 100 *** 6.3) *** 0.0) *** concentrations of 30 mg a.i. kg -1 dry soil and 100 mg a.i. kg -1 dry soil for E. andrei and P. excavatus, respectively. Effects of carbofuran on the avoidance behavior of E. andrei and P. excavatus in both soil types are summarized in Tables 1 and 3. P. excavatus was significantly attracted by carbofuran at 1 mg a.i. kg -1 dry soil in OECD soil and at 2 mg a.i. kg -1 dry soil in both soils (Table 3). When focusing on 51

Chapter 4 EC 50 values for the avoidance of carbofuran, E. andrei was 2.0-2.2 times more sensitive than P. excavatus in OECD and natural soil, respectively. Avoidance response did not differ between the two soil types, which were confirmed by the NOAEC values of 1 and 2 mg a.i. kg -1 dry soil for E. andrei and P. excavatus, respectively. Habitat function was reduced by carbofuran (>60% avoidance or 80% of the worms in the control soil) at 4 mg a.i. kg -1 dry soil for E. andrei and 8 mg a.i. kg -1 dry soil for P. excavatus. Table 3: Avoidance response of Eisenia andrei and Perionyx excavatus to carbofuran in OECD artificial soil and a natural soil. *** Significant avoidance compared to control with p < 0.001. indicates reduced habitat function (>60% avoidance). Mean net response (% +/- SD) Concentration (mg a.i.kg -1 dry soil) Eisenia andrei Perionyx excavatus OECD Natural OECD Natural 0 0 (16.3) 5 (10.0) 0 (16.3) 0 (16.3) 1 10 (20.0) 15 (30.0) -40 (23.1) -5 (34.1) 2 45 (19.1) *** 40 (16.3) *** -35 (10.0) -30 (11.5) 4 85 (10.0) *** 80 (16.3) *** 55 (19.1) *** 50 (11.5) 8 100 *** 100 *** 65 (19.1) *** 50 (11.5) 16 100 *** 100 *** 100 *** 100 *** Discussion The avoidance test is considered a sensitive tool in risk assessment as earthworms detect a wide range of contaminants including polycyclic aromatic hydrocarbons, heavy metals, explosives, crude oil and pesticides (ISO, 2007). The studies with pesticides so far reported different sensitivities. Low concentrations of mancozeb (Reineke et al., 2002), benomyl and carbendazim (Loureiro et al., 2005; Garcia et al., 2008) and lambda-cyhalothrin (Garcia et al., 2008) were avoided by earthworms but Hodge et al. (2000) reported no avoidance of the organophosphates diazinon and chlorpyrifos. The present study reports avoidance of chlorpyrifos and carbofuran by both Eisenia andrei and Perionyx excavatus. The validity of the test design was determined by counting the number of dead or missing worms in the test containers. The design is considered to be invalid if the number of dead or missing worms is > 10 % per treatment (ISO, 2007). This condition was met in all the concentrations except the 52

Comparative sensitivity of earthworms in avoidance tests highest concentrations used and therefore the latter were excluded from the analysis. Carbofuran was more toxic to both species than chlorpyrifos with low EC 50 and NOEC values for both soil types tested. This study shows that avoidance data generated through using E. andrei and P. excavatus in pesticide treated soils could vary. E. andrei was considered more sensitive than P. excavatus because the EC 50 s were approximately two times higher. This observation is mainly due to the attraction of the latter species to low concentrations of both pesticides. Reinecke et al. (2002) also reported the inability of P. excavatus to detect low concentrations of Pb. Therefore with the limited data available it is expected that the avoidance response of P. excavatus might be a poor indicator of soil contamination and its potential risk in the field conditions. Species sensitivity could also be a vital factor in the avoidance test. Different sensitivities of earthworm species like Aporrectodea caliginosa, Dendrobeana octaedra and Lumbricus rubellus in soil spiked with Cu and Zn mixtures have been reported (Lukkari and Haimi, 2005). A study by Owojori and Reinecke (2009) has shown that A. calignosa is more sensitive than E. fetida in two different substrates spiked with NaCl. The speciesspecific sensitivity could mainly be due to characteristic differences in chemoreceptors (Stephenson et al., 1998), physiological and morphological (Edwards and Bohlen, 1996) and ecological differences (Lukkari and Haimi, 2005). But ecological properties may not be good indicators in comparing E. andrei and P. excavatus as both species are epigeic in nature and have preferences for high organic matter contents. Therefore the different avoidance responses of the two species reported here could mainly be due to physiological and morphological differences. The unavailability of detailed comparative studies of morphological and physiological characters of the two species hampers further explanations. This study also reports similar sensitivity of both species in different soil types. Also no significant attraction or avoidance was observed in tests comparing control OECD and control natural soil. This confirms that avoidance or attraction responses were induced directly by the pesticides and not by the soil types. It could be expected that endpoints generated through short-term exposure (48 h) might not vary much for substrates having different physico-chemical parameters. Hund-Rinke and Wiechering (2001) tested six soils. They found no specific movement towards any of the soils and concluded that earthworm avoidance behaviour was mainly driven by pollutants rather than by soil type. But studies by Amorim et al. (2005) with Enchytraeus albidus and Garcia (2004) with E. fetida gave different outcomes and did show the importance of soil properties for avoidance behaviour. 53

Chapter 4 Table 4. Comparison of the effects of different chemicals on earthworm survival (LC 50 ), reproduction (EC 50RP ) and avoidance behaviour (EC 50AV ) in tests performed under similar conditions in OECD artificial soil (OECD), tropical artificial soil (TAS), standard European natural soil (LUFA 2.2) and field soils (FS). Chemical Species Soil Type LC 50 (mg/kg dry soil) EC 50RP (mg/kg dry soil) EC 50AV (mg/ kg dry soil) NaCl E. fetida OECD 5436 a 2020 a 1164 b FS 0.75 c 0.29 c b 0.56 A. caliginosa FS 1.15 c 0.29 c b 0.26 Benomyl E. fetida TAS 633 d 3.8 d 54.9 e OECD 22 d 1.6 d 28.2 LUFA 2.2 14.6 d 1.0 d e 1.6 Carbendazim E. fetida TAS >1000 d 4.6 d 33.3 OECD 5.8 d 2.7 d e 127.4 LUFA 2.2 4.1 d 0.6 d e 7.1 E. andrei LUFA 2.2 2.7 g 0.34 g f <1 OECD 38.9 g 1.8 g 126.4 Chlorpyrifos E. andrei OECD 154 g 3.8 g i 9.31 FS 106 g 5.8 g i 12.1 Carbofuran E. andrei OECD 13.5 g 1.2 g i 2.15 FS 11.9 g 1.9 g 2.44 Cadmium E. fetida FS - 157 j k 183 PCP E. fetida FS - 60 j k 24 FS - 13 j k 8.2 TBT E. fetida FS - 2.7 j k 2.6 FS - 1.3 j k 2.3 TNT E. fetida FS - 167 j k 27 FS - 68 j <32 Lambda E. fetida TAS 23.9 d 7.7 d e 0.2 cyhalothrin OECD 99.8 d 37.4 d e 3.3 LUFA 2.2 140 d 44.5 d e 0.5 a Owojori et al. (2008), b Owojori & Reinecke (2009), c Owojori et al. (2009), d Garcia (2004), e Garcia et al. (2008), f Loureiro et al. (2005), g De Silva et al. (2009), h De Silva and Senewirathne (2008), i This study, j Hund- Rinke and Simon (2005), k Hund-Rinke et al. (2005) The possibilities of replacing acute and chronic tests by avoidance tests to investigate the effects of pollutants on soil invertebrates are still rather questionable. A study by De Silva et al. (2009) performed under similar conditions found LC 50 values for the effects of chlorpyrifos and carbofuran on the survival of the same species after 28 days exposure of 154 and 13.5 mg a.i. kg -1 dry soil and EC 50 values for the effects on reproduction of 3.86 and 1.25 mg a.i. kg -1 dry soil, respectively. This shows that the avoidance response is more sensitive than survival but less sensitive than reproduction. 54

Comparative sensitivity of earthworms in avoidance tests This observation is in accordance with the earthworm avoidance responses reported by Hund-Rinke et al., (2005), Garcia et al. (2008) and Owojori & Reinecke (2009). Table 4 summarizes the studies performed under similar conditions with the same chemicals and comparing different endpoints. Eighteen studies were found for the comparison of survival and reproduction with avoidance response and another seven studies to compare reproduction and avoidance. In 67 % of these studies avoidance was more sensitive than survival. Reproduction generally was more sensitive than avoidance behaviour with the latter being most sensitive only in 8 out of 25 cases. In nine cases (36 %) EC 50 values for the effects on avoidance and reproduction did not differ by more than a factor of 2. These results demonstrates that earthworm avoidance tests could be an ideal replacement for earthworm acute tests and might be used as an initial screening method to assess the risk of pesticide pollution in the tropics. They may however, not replace the ecologically more relevant reproduction test. Conclusions Eisenia andrei was more sensitive in avoidance tests with two pesticides such as chlorpyrifos and carbofuran than Perionyx excavatus under tropical conditions. This difference in sensitivity might lead to a wrong estimation of potential effects in the tropics. It is therefore recommended to consider avoidance tests with tropical species like Perionyx excavatus with extra caution in tropical risk assessment. The generally lower sensitivity of avoidance tests compared to reproduction tests and their higher sensitivity than survival indicates they could only be used as an initial screening of pesticide toxicity to earthworms. Acknowledgements This work was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB 84-609) residing under the Netherlands Organization for Scientific Research (NWO). We thank H.G. Hector for technical assistance.. 55

56

Chapter 5 Development of An Alternative Artificial Soil for Earthworm Toxicity Testing in Tropical Countries P. Mangala C. S De Silva, Cornelis A.M. van Gestel Applied Soil Ecology (in press) 57

Chapter 5 Abstract The standard soil invertebrate toxicity tests developed by OECD and ISO use an artificial soil as the test substrate, which contains sphagnum peat as a component. This type of peat is not widely available. Investigation of possible alternative substrates using locally available materials therefore is vital for performing such ecotoxicity tests, particularly in the tropics. We studied the suitability of paddy husk (PH), saw-dust (SD), non composted (NCCP) and composted coco peat (CCP) as a replacement for sphagnum peat. Artificial soil (AS) was prepared by mixing 70% sand and 20% kaolin clay with 10% PH, SD, NCCP or CCP. First, the reproduction potential of the earthworm Eisenia andrei was investigated in modified artificial soil (MAS) using the original OECD AS as the control. The number of juveniles produced in OECD AS, MAS PH and MAS CCP was not significantly different but it was significantly reduced (p<0.05) in MAS SD and MAS NCCP. The toxicity of chlorpyrifos, carbendazim and carbofuran for E. andrei was determined to validate the substrates. The 28 day LC 50 s for the three pesticides in original AS, MAS CCP and MAS PH were not significantly different, but the EC 50 for effects on reproduction in the MAS PH was significantly lower (p < 0.05) compared to OECD AS and MAS CCP. We conclude that composted coco peat might be a suitable replacement for sphagnum peat in AS for soil ecotoxicity studies. Key words: OECD soil, Coco peat, Saw-dust, Paddy husk, Artificial soil 58

Development of an alternative artificial soil Introduction Soil properties may have a large impact on the availability and therefore on the toxicity of chemicals to soil invertebrates (Van Gestel, 1997). The risk from chemicals in soil is strongly dependent on soil type and the assessment of such risks has to take into account soil properties. This consideration led to the conclusion that it would be preferable to standardize the soils used for determining the toxicity of chemicals to soil organisms. Standardization might be achieved by selecting and prescribing a natural soil, such as the German standard soils commercially available from LUFA (Agricultural Research Center, Speyer, Germany) that are used for routinely assessing sorption, mobility and biodegradation of pesticides in soil. Alternatively, artificial soils that can be prepared from well-standardized and easy-toobtain materials may be used. OECD (1984) described such an artificial soil for determining the acute toxicity of chemicals to earthworms. Since that time, the OECD artificial soil has been used in many other toxicity tests with earthworms, enchytraeids and Collembola as described by OECD (2004a,b) and ISO (1993, 1998b, 1999b). OECD artificial soil consists of 10 % finely ground sphagnum peat, 20 % kaolin clay, approx. 69 % quartz sand and approximately 1 % of CaCO 3 to adjust ph to approximately 6.0. Although extrapolation of toxicity data generated through artificial soil to field conditions is sometimes difficult, it remains a popular substrate in soil ecotoxicology due to the existence of a large data base for a wide range of xenobiotics, such as pesticides, metals and industrial chemicals, and the easy interpretation and easy comparison of the results. This artificial soil uses sphagnum peat as a component to represent soil organic matter. Sphagnum peat is also a popular potting medium in floriculture and horticulture. Purchase costs of sphagnum peat are however increasing (Meerow, 1994) and there is a growing concern that it is mined from endangered bogs and fen ecosystems which are declining rapidly due to environmental constraints (Barkham, 1993; Robertson, 1993; Frolking et al., 2001). Increasing knowledge of peat-land conservation has also created a need to find possible alternatives. More environmentally acceptable substrates, such as coir dust (Noguera et al., 2000; Abad et al., 2002), vermicompost (Zaller, 2007), organic waste (Abad et al., 2001), poultry feather (Evans and Karcher, 2004) and dredged sediments from rivers (Marchese et al., 2006; Di Benedetto, 2007) have successfully replaced sphagnum peat in horticulture practices. So far little efforts have been made to replace the sphagnum peat in artificial soil. Recently, however, it has become evident that sphagnum peat is scarce and completely unavailable in many regions, including the tropics. Garcia (2004) and Römbke et al. (2007) also considered this problem and suggested using locally available materials, like coir dust, for performing soil ecotoxicity tests especially in these regions. After some preliminary 59

Chapter 5 assessments we hypothesized that locally available coco peat (composted and non-composted), paddy husk and saw dust might be suitable substrates for replacing sphagnum peat in artificial soil. Coco peat is dried and compressed coir dust extracted from the mesocarp of the coconut (Cocos nucifera) fruit, which is common in Asia, tropical America and Africa (FAO, 2000) and produced at fairly constant quality. It has been used as a soil-less growing medium for more than three decades in tropical regions (Chweya et al., 1978) and has become commercially very popular recently in both composted and non-composted forms (Offord et al., 1998; Noguera et al., 2000). In addition, it has also been used in road and infrastructure construction. Saw-dust is the heterogeneous fibrous material originating from timber processing mills. Mostly sawdust is incinerated or dumped and occasionally it may be used as an energy source. Paddy husk is also a very common material in the tropics, where rice is the main agricultural commodity. It is a biodegradable material resulting from processing paddy into rice. Currently it has no commercial value but it has been used in agriculture as a fertilizer and as an energy source. The aim of this study was to determine the suitability of coco peat (composted and non-composted), paddy husk and saw-dust as alternatives to sphagnum peat in the artificial soil used in soil ecotoxicity studies. We aimed at validating the potential substrates by performing toxicity tests with the common compost earthworm Eisenia andrei (Lumbricidae), using chlorpyrifos, carbofuran and carbendazim as test chemicals. Materials and Methods Modified artificial soils (MAS) were prepared according to the standard guidelines of OECD (1984) and ISO (1998b), but replacing the (10 %) sphagnum peat with a similar amount of finely ground non-composted coco peat (NCCP), composted coco peat (CCP), paddy husk (PH) or saw-dust (SD). The resulting MAS were moistened to 50 % of their maximum water holding capacities and physico-chemical properties of the modified artificial soils were determined by the Industrial Technology Institute (ITI), Colombo, Sri Lanka according to the guidelines prepared by the Sri Lanka Standard Institution (SLS, 2003), Association of Analytical Communities International (AOAC, 2000) and International Standardization Organization (ISO, 1993; ISO, 1998a). In the first series of tests survival, growth and reproduction of earthworms were measured in the modified artificial soils at two different temperatures (20 and 26 ºC), representing temperate and tropical conditions. Age-synchronized adult Eisenia andrei (450-500 mg wet weight), obtained from a culture kept at the Department of Zoology, University of Ruhuna, Matara, Sri Lanka, were added to 500 g dry weight equivalent of moist 60

Development of an alternative artificial soil MAS in glass containers (750 ml) with loosely closing lids (n = 10, 4 replicates for all treatments and controls). During the study, 5 g (dry weight) of food (finely ground dried cow manure) moistened to 50 %, was added to each container in a hole made in the middle of the soil (Van Gestel et al., 1989). Additional food was introduced when the added food was consumed. Test containers were kept for 28 days in the dark at 20 ± 2 C or 26 ± 2 C after which the adult worms were removed by hand sorting and adult mortality and biomass change were determined. The soil was returned to the test containers and they were incubated for another 28 days to allow for cocoon development. After 56 days the juveniles were expelled from the soil by placing the test containers in a water bath kept at 60 C and counted. In addition, changes in the modified artificial soils and behavior of the worms (ability to burrow into and distribution in the test soil) were also observed. The standard artificial soil described by OECD and ISO was used as the control substrate. In the second series of tests, the selected substrates were tested for their suitability in toxicity experiments using chlorpyrifos (98 % pure, Cheminova Ltd, Denmark), carbofuran (97 % pure, Sigma Aldrich, The Netherlands) and carbendazim (98 % pure, Sigma Aldrich, The Netherlands). Earthworm reproduction tests (ISO, 1998b; OECD, 2004a) were performed at 26 ± 2 C. All tests included controls and solvent controls without pesticide treatment. The pure pesticides were mixed in AS and MAS using acetone ( 99.5% (Sigma-Aldrich). After mixing in, test soils were left overnight to allow the acetone to evaporate. Solvent controls received the same acetone treatment. Final end points of the tests were survival, behavioral changes, and number of juveniles after 56 days. Levene s test and the Kolmogorov-Smirnov test were performed to assess the homogeneity of variance and normal distribution of the data, respectively. Effects of the modified artificial soils on earthworm reproduction and growth were analyzed by ANOVA and Dunnett s test. LC 50 values for the effect of the test chemicals on earthworm survival and corresponding 95% confidence limits were calculated using the Trimmed Spearman Karber method (Hamilton et al., 1977). The model for logistic response (Haanstra et al., 1985) was used for calculation of EC 50, EC 10 and corresponding 95% confidence limits for effects on reproduction. Differences in EC 50 values between different soil types were tested for significance using a generalized likelihood ratio test (Sokal and Rohlf, 1995). No observed effect concentrations (NOEC) values were determined using ANOVA and Dunnett s test. All statistical analyses were run in SPSS version 14.0. 61

Chapter 5 Results Physico chemical characteristics standard and modified artificial soils Selected physico-chemical properties of modified artificial soils and the standard OECD artificial soil are given in Table 1. The soils differed markedly in certain properties, especially in ph. MAS NCCP was acidic followed by the slightly acidic MAS PH and MAS SD, while the ph of the MAS CCP and the OECD soil was closer to neutral. Higher Ca and Mg contents were found in MAS CCP and OECD soils than in the other MAS. The concentration of essential elements such as NPK did not differ between soil types, and the same applied for organic carbon content, organic matter content and C/N ratio. The highest maximum water holding capacity was recorded in the OECD soil and the lowest in MAS PH and MAS SD. Performance of earthworms in modified artificial soils There was no adult mortality in the modified artificial soils except for MAS NCCP with 40 % mortality. Formation of a thick fungal layer was noticed in MAS SD and MAS PH test containers. Homogeneous distribution of the introduced worms was observed in OECD AS, MAS PH and MAS CCP, while earthworm aggregation in the food was seen in MAS SD and MAS NCCP. Effects of the different types of soil on weight change of the earthworms are shown in Figure 1. The overall soil effect was highly significant (F = 373.73, p < 0.001). Growth of adults in MAS PH and MAS CCP was not significantly different from the control OECD soil (p > Table 1. Selected physico-chemical parameters of the modified artificial soils and standard OECD soil used in this study. Parameter Method OECD MAS CCP MAS NCCP MAS PH MAS SD ph (0.01M CaCl2) ISO (1994) 6.2 6.4 4.5 5.6 5.2 N Total(% ) SLS (2003) 0.1 0.1 0.1 0. 1 0.03 P ( % ) SLS (2003) 0.08 0.1 0.1 0. 1 0.1 K ( % ) SLS (2003) 0.1 0.08 0.09 0. 1 0.1 Ca (mg kg -1 ) AOAC (2000) 240 8 7 4 7 Mg (mg kg -1 ) AOAC (2000) 100 144 188 100 78 C Org( % ) SLS (2003) 6.4 6.8 8.6 6. 7 7.0 Organic Matter (% ) Storer (1984) 8.5 8.8 10.6 8. 7 9.2 C/N ratio - 64 68 86 67 233 Density (g cm -3 ) ISO (1998a) 1.1 1.2 1.1 0. 9 0.8 WHCmax ( % ) ISO (1998b) 82.5 70 66 42 46 62

Development of an alternative artificial soil Mean growth ± SD (%, 28 days) 30 20 10 0-10 -20-30 -40 20 0 C 26 0 C OECD MAS SD MAS PH MAS CCP MAS NCCP Soil Type Figure 1. The growth of Eisenia andrei (mean ± SD, n = 4) after 28 days of incubation in different artificial soil types, see text for explanation. Mean number of juveniles ± SD 160 140 120 100 80 60 40 20 0 20 0 C 26 0 C OECD MAS SD MAS PH MAS CCP MAS NCCP Soil Type Figure 2. Reproduction of Eisenia andrei, expressed as the number of juveniles produced (mean ± SD, n = 4) after 28 days of incubation in different artificial soil types, see text for explanation. 63

Chapter 5 0.05). In MAS SD a significant earthworm weight loss was recorded at 20 ºC but a slight weight gain was observed at 26 ºC (p < 0.05). In MAS NCCP the earthworms lost over 25% of their initial body weight, and growth differed significantly from that in the control OECD AS (p < 0.05) at both temperatures. The mean number of juveniles produced after 56 days also varied in the different test soils (F = 494.44, p < 0.001, Figure 2). The highest numbers of juveniles were recorded in MAS CCP, MAS PH and the control OECD AS and the differences between these three treatments were not significant (p > 0.05). In MAS NCCP E. andrei produced the lowest number of juveniles, followed by MAS SD and the deviation from the OECD control soil was significant (p < 0.05). Toxicity of chemicals in modified artificial soils Based on the performance of the earthworms in the modified artificial soils, MAS PH and MAS CCP were selected for toxicity testing. The LC 50 and EC 50 values for the effects of the three pesticides on the survival and reproduction of E. andrei in the different soil types are given in Table 2. For all three chemicals LC 50 values in the modified artificial soils were similar to the value recorded in the OECD control soil. The numbers of juveniles were significantly higher in solvent treated controls than in untreated controls (t-test, p < 0.05), hence the solvent controls were used in further data analysis and interpretation of effects on reproduction. The doseresponse relationships for the effects of the three pesticides on the reproduction of E. andrei in MAS CCP overlapped with those for OECD AS but no overlap was found for MAS PH (Figure 3). The EC 50 was similar in MAS CCP and OECD AS for all three chemicals, and no significant difference was observed (Generalized likelihood ratio test, p > 0.05). A significantly lower EC 50 was recorded in MAS PH for all three pesticides (p < 0.05). A similar trend was also observed for EC 10 values except for carbofuran with no significant differences between the three soil types. For all three pesticides the NOEC values were below the lowest concentration tested. 64

140 OECD MASccp MASph Number of juveniles 120 100 80 60 40 20 0-8 -6-4 -2 0 2 4 6 8-6 -4-2 0 2-8 -6-4 -2 0 2 4 Ln concentration (mg a.i kg -1 dry soil ) Figure 3: Dose-response relationships for the effects of chlorpyrifos (left), carbofuran (middle) and carbendazim (right) on the reproduction of Eisenia andrei in OECD artificial soil and modified artificial soils using composted coco peat (MAS CCP ) and paddy husk (MAS PH ).

Chapter 5 Table 2. Effects of three selected chemicals on the survival and reproduction of Eisenia andrei after 28 days exposure in standard OECD artificial soil (OECD) and manipulated artificial soils using composted coco peat (MS CCP ) and paddy husk (MAS PH ). LC 50, EC x values and corresponding 95% confidence intervals are given in mg a.i. kg -1 dry soil and * denotes significant differences as tested by a generalized likelihood ratio test (p < 0.05). Soil type Chlorpyrifos Carbofuran Carbendazim LC 50 EC 50 EC 10 LC 50 EC 50 EC 10 LC 50 EC 50 EC 10 OECD 150 (132 171) 2.0 (1.4-2.6) 0.09 (0.02-0.15) 12.9 (11.1-15.1) 1.1 (0.9 1.3) 0.26 (0.21 0.32) 39.0 (33.2-45.7) 1.0 (0.8-1.2) 0.14 (0.08-0.26) MS CCP 154 (134 177) 2.0 (1.3 2.7) 0.08 (0.01-0.15) 12.5 (10.7 14.6) 1.2 (0.9 1.4) 0.18 (0.11 0.24) 40.9 (35.1-47.6) 1.2 (0.9-1.3) 0.16 (0.09-0.22) MS PH 142 (122. 164) 0.5 * (0.4 0.6) 0.03 (0.01-0.05) 12.1 (10.3 14.1) 0.6 * (0.5 0.7) 0.21 (0.17 0.25) 37.4 (31.4-44.5) 0.5 * (0.3-0.6) 0.04 (0.02-0.63) Discussion Replacing the organic matter component resulted in varying physicochemical properties of the resulting substrates although differences in general were small (Table 1). Noguera et al. (2000) have earlier reported similar physico-chemical parameters of peat and coir waste but no comparative studies with paddy husk and saw dust could be found in the literature. The most pronounced difference between the modified artificial soil types was the ph, which is important when performing standard toxicity tests. The modified artificial soils containing NCCP, PH and SD were slightly acidic in nature, which may have contributed to the 40 % earthworm mortality in MAS NCCP and aggregation behavior towards food particles in MAS SD and MAS NCCP. The similar ph of OECD soil and MAS CCP indicated that CCP could be a good alternative, which is also supported by homogeneous distribution of the worms in these substrates. Due to the acidic nature of the sphagnum peat the ph of the OECD soil usually has to be adjusted by adding CaCO 3, whereas CCP is neutral by itself and does not require ph adjustment. However, Tapia et al. (2008) indicated that the physico-chemical properties of coco peat may vary with location and processing type and therefore the source of the coir dust will be a key factor when considering replacing standard sphagnum peat in artificial soil. The performance of the earthworms in the modified artificial soils is a good indicator of the suitability of the alternative substrates for the OECD artificial soil. The ISO and OECD guidelines require 10% mortality in the 66

Development of an alternative artificial soil untreated controls in both acute and chronic tests as validity criteria. Furthermore they also require less than 20% weight loss compared to the initial biomass and 30 juveniles per replicate with a coefficient of variation of 30% (ISO, 1993; ISO, 1998b; OECD, 2004a). The higher mortality in MAS NCCP at both test temperatures shows the unsuitability of NCCP as an alternative for sphagnum peat. This is also supported by the weight loss and lower juvenile production in MAS NCCP (Figures. 1, 2). This could mainly be due to the acidic nature of the substrate that could trigger mortality, weight loss and lower reproduction. Lee (1985) reported that most earthworm species dislike ph values around 4.0-4.5 and Van Gestel et al. (1992) suggested an optimum ph range of 5.0-6.0 for E. andrei (see also Jänsch et al., 2005). Our study with acidic substrates also confirms the fact that a low ph is not suitable for this earthworm species. Although no earthworm mortality was recorded in MAS SD, the significant weight loss and lower juvenile production at 20 ºC could be critical. The performance of earthworms in untreated MAS CCP and MAS PH controls was similar to that in OECD soil making them suitable alternative test substrates. The LC 50 values recorded for the effects of all three chemicals on earthworm survival were similar, indicating the suitability of CCP and PH as alternatives to sphagnum peat. The recorded LC 50 values were within the range of a previous study by De Silva et al. (2009) for 26 ºC. These values are also in the range of LC 50 values reported for carbofuran (Heimbach, 1985; Haque and Ebing, 1983) but not for chlorpyrifos and carbendazim. Available 14-day LC 50 values for chlorpyrifos (1077 mg kg -1 dry soil) and carbendazim (> 1000 mg kg -1 dry soil) are higher under different and similar test conditions, respectively (Ma and Bodt, 1993; Garcia, 2004). Lack of data under similar conditions hampers further comparison of the recorded LC 50 values obtained for the different substrates. For the effects on reproduction, results for the three pesticides varied with the substrates used. The dose response curves for the effects on reproduction in MAS CCP and OECD AS were similar for all three pesticides (Figure. 3), suggesting that similar results could be obtained by using MAS CCP instead of OECD AS. The EC 50 values reported for chlorpyrifos and carbendazim were only slightly lower than the values reported in our previous study (De Silva et al., 2009). However Garcia (2004) reported 14 times higher EC 50 for the effects of carbendazim on reproduction of Eisenia fetida at 28 ºC. No other data could be found for chlorpyrifos and carbofuran. In MAS PH, however dose-response relationships and resulting EC 50 values for all three chemicals were significantly different from those for MAS CPP and OECD AS. As LC 50 values were similar for MAS CCP, MAS PH and OECD AS, the lower EC 50 in MAS PH could be due to effects on cocoon development, which resulted in lower numbers of juveniles in MAS PH. It is expected that microbial activity is greater in MAS CCP than in MAS PH. An active microbial community could cause faster degradation 67

Chapter 5 (Singh and Walker, 2006) and lower the bioavailability of the chemicals in the soil. This could explain the lower EC 50 observed for MAS PH, but it could also be due to unexplained interactions between the pesticides tested and the physico-chemical properties of MAS PH. Further studies with additional stress factors using MAS PH as the substrate may explain the observed higher toxicity in MAS PH. Conclusions The results of this study indicate that composted coco peat may be a good alternative to sphagnum peat in artificial soils for performing earthworm toxicity studies, which may be more practical especially in the tropics. Additional tests with metals, other organic chemical compounds and other organisms will be necessary to confirm the general suitability of composted coco peat as an alternative in soil invertebrate toxicity studies. Acknowledgements We thank Agro Soil Private Limited, Sri Lanka for providing coco peat and H.G. Hector for technical assistance. This work was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB 84-609) residing under the Netherlands Organization for Scientific Research (NWO) and National Science Foundation, Sri Lanka (RG/2006/EB/06). 68

Chapter 6 Chlorpyrifos Causes Decreased Organic Matter Decomposition by Suppressing Earthworm and Termite Communities in Tropical Soil P. Mangala C. S. De Silva, Asoka Pathiratne, Nico M. van Straalen, Cornelis A.M. van Gestel (Submitted) 69

Chapter 6 Abstract Effects of pesticides on structural and functional properties of ecosystems are hardly being studied under tropical conditions. In this study litter bag tests and earthworm field tests were performed simultaneously at the same field site sprayed with chlorpyrifos (CPF). The recommended dose of CPF (0.6 kg a.i ha -1 ) and two higher doses (4.4-8.8 kg a.i ha -1 ) significantly decreased litter decomposition during the first 3 months after application. Decreased organic matter breakdown could be explained from lower earthworm and termite abundances recorded during this period. Speciesspecific effects of CPF on earthworm abundance and biomass were observed for Perionyx excavatus being mostly affected and Megascolex sp. least affected. Decomposition was completed already within 4 months after CPF application even in the treated plots, which suggests the need for modification of standard test guidelines to comply with faster litter degradation under tropical conditions. Key words: Litter bag test, Earthworm field test, Termites, Decomposition, Sri Lanka 70

Chlorpyrifos causes decreased organic matter decomposition Introduction Soil contamination due to increased use of agrochemicals and subsequent effects on soil fauna diversity and functions is a major concern for tropical ecosystems. Risk assessment and follow-up regulation of agrochemicals is an urgent need and also should be a part of sustainable agriculture. But such processes are being hampered by scarcity of data. Studies by Garcia (2004), Römbke et al. (2007), Garcia et al. (2008) and De Silva et al. (2009) have attempted to fulfill this gap of knowledge mainly by using basic and adapted laboratory toxicity tests. This classical approach of toxicity assessment using single species laboratory studies does not take into account ecologically relevant factors like the structure of the soil community and its role in the functioning of the soil ecosystem. The applicability of basic ecotoxicity tests to tropical ecosystems is also rather questionable, where complex environmental conditions and their interactions co-exist. A tiered approach has therefore been suggested (Römbke et al., 1996), which consists of laboratory tests together with semi-field and field studies. Such an approach seems more relevant in addressing the effects of soil pollutants in tropical systems. The importance of linking laboratory studies and field studies has also been highlighted by Van Gestel and Weeks (2004). Stone (1995) and Ekschmitt et al. (2001) highlighted the importance of species diversity and their functions in soil ecosystems. Soil pollutants such as pesticides may cause serious damage to soil organisms and can cause population decline and community shifts. Such effects on structural properties of communities subsequently may affect soil functions, such as decomposition and nutrient cycling (Muthukaruppan et al., 2004; Römbke et al., 2005). This has been insufficiently studied under tropical conditions except for a few cases (Förster et al., 2006; Casabé et al., 2007; Reinecke and Reinecke, 2007). Tropical soil ecosystems obviously differ from temperate ones in both biotic and abiotic components, but this may both increase and decrease the adverse effects of pesticides. Generally pesticides will be degraded faster under tropical conditions causing lower toxicity (Viswanathan and Krishanamurti, 1989; Laabs et al., 2000). Alternatively, activity of soil organisms may be higher in tropical soils compared to temperate climates due to warm and humid conditions and this may trigger higher uptake of pesticides and higher toxicity. De Silva et al. (2009) reported a higher toxicity of chlorpyrifos under tropical conditions due to higher activity of the earthworms but a lower toxicity of carbendazim due to faster degradation of the pesticide (see also Garcia, 2004). These contrasting results show the importance of semi-field and field studies under tropical conditions, which can give more realistic assessments and allow for validation of laboratory data. 71

Chapter 6 Organic matter decomposition is an important function of soil ecosystems (Lavelle et al., 1997). Ecological, microbiological, chemical, and physical endpoints can be used in assessing such functions in soil. Mass loss is the most common end point and it is also linked with biological, chemical and physical properties of the soil ecosystem. Effects of pesticides on decomposition rates are often assessed by litter bag tests. This is primarily targeting the effects of such products on the initial phase of the decomposition expressed as litter mass loss. This test is both simple and realistic for determining effects on soil functions and can be applied in semi-field and field conditions (Knacker et al., 2003; Römbke et al., 2003). In turn proper functioning of the soil may mainly depend on the structure of the soil community such as diversity, abundance and biomass of soil organisms. These structural changes may be investigated by the earthworm field test described by ISO (1999a). Earthworms represent a major component of the soil community and they play a key role in soil functions such as decomposition, nutrient cycling, and soil structuring, which are all vital activities for the maintenance of healthy soil (Edwards, 1998). Nevertheless earthworms are susceptible to many pesticides and are considered good indicators of pesticide pollution. The litter bag test and the earthworm field test are rarely being combined simultaneously in the same field experiment, despite the recommendations of Knacker et al. (2003). The aim of this study was to determine the effects of chlorpyrifos (CPF), a frequently used pesticide in tropical countries, on litter decomposition and invertebrate community structure (earthworms, termites). Materials and Methods Study site & design The study site was situated in Walasgala (5 0 59 N, 80 0 41 E) in the Southern province of Sri Lanka. The site was used for cinnamon (Cinnamomum verum) cultivation a decade back. After that it has not been used for any other agriculture purposes and no prior application of any pesticides has been reported. Two months before starting the experiment the study site was subjected to mechanical weeding ensuring no or little disturbance of soil and soil organisms. Soil analysis prior to the experiments classified soil texture as sandy loamy with ph (0.01 M CaCl 2 ) of 6.5 7.2, 11% organic carbon, bulk density 1.4 g/cm 3, 0.2 % N, 0.2% P and 0.05 % K. An earthworm field test and a litterbag test were performed in the same study site. Plots (n = 16) were laid out for earthworm field tests (10 x 10 m) and randomly assigned to one of three treatments and controls each with four replicates. Each plot was separated from its adjacent plots by 1 m to avoid possible cross contamination. Litter bag tests were performed in 72

Chlorpyrifos causes decreased organic matter decomposition randomly selected subplots (5 x 5 m, n = 16), residing inside the same plots assigned to earthworm field test.. Pesticide application and analysis Formulated chlorpyrifos (Judo 40 EC, 400 g l -1, Lankem Ceylon Limited, Sri Lanka) was applied on 26 August 2007, as an aqueous solution in three treatment levels (T 1, T 2 and T 3 ) using a knapsack sprayer (16 l). T 1 was the highest recommended dose (0.6 kg a.i./ ha) based on the pesticide recommendation guidelines published by the Department of Agriculture, Sri Lanka. T 2 (4.4 kg a.i./ ha) was the dose equivalent to the EC 50 value for the effects of chlorpyrifos on earthworm reproduction under tropical conditions (De Silva et al., 2009), which was seven times the recommended dose and also equivalent to the cumulated dose that could be expected after the average number of applications for one crop cycle under a worst case scenario. T 3 (8.8 kg a.i./ha) was two times the EC 50 value. The control plots received only water. Soil samples (0-5 cm) were taken from each plot for pesticide analysis during the study period, were thoroughly mixed to obtain a pooled sample for each treatment level and kept refrigerated until further analysis. Chlorpyrifos concentration in the soil samples was later determined at the Industrial Technology Institute, Colombo, Sri Lanka using GC techniques applying proper quality control measures. Litterbag test The test methodology was mainly based on the guidelines for assessment of organic matter breakdown in litter bags described by Römbke et al. (2003) and OECD (2006), with slight modifications. Litter bags were prepared from nylon nets with 5 mm mesh size primarily used for manufacturing fishing nets. Each litterbag (10 x 20 cm) was filled with 5 6 g of air dried (30-32 0 C) paddy straw (Oryza sativa, > 90 % stalk), ensuring a thin and evenly distributed layer inside the litterbag. Litterbags (n = 30 for each replicate plot) were buried in the top soil layer at a depth of about 1-2 cm before CPF application. Litterbags (n = 6) were randomly sampled at monthly intervals from each plot after CPF application. Soil, soil organisms and other debris were separated from the litter by dry sieving followed by wet sieving. Resulting litter material was air died for 12 hrs at 30-32 0 C and weighed (W1). Litter material was then subjected to combustion in porcelain crucibles at 550 0 C for 45 minutes. Crucibles were left to cool down in a dessicator till a constant weight was achieved and final weight was determined (W2). Loss on ignition (LOI) was calculated as W1 W2 (g). Additional soil and straw (n = 10 each) were taken to determine the soil correction factor (SCF) and straw correction factor (StCF). SCF was calculated as (soil input - ash 73

Chapter 6 residue) / soil input and StCF was determined as ash residue / straw input. Corrected loss on ignition (CLOI) and non-degraded straw (NDS) values were calculated as: CLOI = LOI (SCF x (W2 / (1 SCF))) NDS = CLOI + (StCF x CLOI x SCF / (1-SCF)) Organic matter breakdown was then determined by subtracting NDS from the initial weight of the straw sample. Soil fauna sampling An earthworm field test was performed in accordance with the ISO guideline 11268-3 (ISO, 1999a) with slight modifications. Prior to CPF application and 1, 2, 3, 6 and 12 months after pesticide application earthworm sampling was done by hand-sorting in randomly selected areas (50 cm x 50 cm, 5-10 cm depth) representing 4 replicates for each plot. Species identification was based on the keys by Stephenson (1923), Gates (1945) and Blakemore (2006). Total adult and juvenile earthworm abundance and biomass (wet weight) were determined. Because of their role in organic matter break down in tropical conditions, abundance of termites was also determined in the same soil samples extracted for earthworm sampling. Statistical analysis Prior to analysis Levene s test and Kolmogorov-Smirnov test were performed to test homogeneity of variance and normal distribution of the data. The effects of CPF on earthworm abundance, biomass and mass loss at each sampling time were compared to control plots using one-way ANOVA. In case of significance difference in the treatments, multiple comparisons were done using Dunnett s test. All statistical analyses were done using SPSS version 14.0. Results Rainfall data Rainfall data obtained from the Department of Meteorology, Colombo, Sri Lanka during the study period (August 2007- August 2008) is given in Table 1. Highest rain fall was recorded prior to application and during 1-2 months after application then decreased with time and later increased in May 2008. 74

Chlorpyrifos causes decreased organic matter decomposition Table 1. Rainfall data obtained from the Department of Meteorology, Colombo, Sri Lanka for a station near the study area. Month Rainfall (mm) Month Rainfall (mm) August 2007 456.0 February 2008 45.6 September 568.5 March 112.4 October 369.5 April 73.9 November 79.1 May 22.4 December 35.7 June 303.4 January 2008 6.1 July 2008 152.5 Table 2. Chlorpyrifos levels in pooled soil samples immediately after application and after 30, 60 and 90 days. Treatment levels are expressed as kg a.i. ha -1 and all the other values are expressed as mg a.i. kg -1 dry soil. Treatment (kg a.i ha -1 ) 0 days (mg a.i kg -1 ) 30 days (mg a.i kg -1 ) 60 days (mg a.i kg -1 ) 90 days (mg a.i kg -1 ) 0.6 0.81 0.05 ND ND 4.4 4.97 3.64 2.90 ND 8.8 8.96 7.50 2.86 ND ND: not detected Chlorpyrifos analysis Chlorpyrifos concentrations measured in pooled soil samples representing the three treatments are given in Table 2. After 1 month of application, at the lowest dose only 6% of the initial CPF concentration was detected. But at two higher doses 73 % and 84 % of the initial CPF concentration was detected after 1 month. CPF was no longer detected after 3 months of application at all three doses. From these data, and assuming first order 75

Chapter 6 degradation kinetics, half lives of 76 and 36 days may be calculated for the highest two doses during the first 60 days of the CPF application. Litter bag test Mean mass loss (%) in litter bags as a function of sampling time is presented in Figure 1. Mass loss in the control plots increased from 44% after 1 month to 100% after 4 months of CPF application. Litter decomposition was significantly reduced at the two highest doses after one month and at all treatments after two and three months (p < 0.05, Dunnett s test). After 4 months, litter decomposition was completed in the controls and in the lowest two doses but still delayed at the highest dose (Figure 1). 120 Mean mass loss ±SD (%) 100 80 60 40 20 0 1 M 2 M 3 M 4 M Sampling time ( Control T 1 T 2 T 3 ) Figure 1: Effects of chlorpyrifos on the development with time of the mean mass loss (± SE, %) of litterbags incubated in a tropical agricultural soil (n = 6). * Indicates a significant difference from the control plots (p < 0.05, Dunnett s test) at each sampling time (in months). C = control, T 1, T 2, and T 3 : chlorpyrifos dosages of 0.6, 4.4 and 8.8 kg a.i. ha -1, respectively. Earthworm abundance and biomass A total of ten earthworm species were recorded throughout the study period. Nearly 60 % of them belonged to the Family Megascolecidae. Three species were identified to be Dichogaster bolaui (Michaelsen), Perionyx excavatus (Perrier), and Lampito mauritii (Kinberg); the biggest earthworm species 76

Chlorpyrifos causes decreased organic matter decomposition found (80-90 cm long) belonged to the genus Megascolex. This species has never been documented in the earthworm fauna of Sri Lanka and possibly is a new species (pers. com., J.M. Julka). Other species could not be identified due to their low numbers and lack of good samples. The mean adult and juvenile earthworm abundance is given in Figure 2. The total number of earthworms prior to the experiment showed no differences between plots assigned and amounted 25-34 adult worms and 28-33 juveniles per 0.25 m 2. The number of adult earthworms in all field plots gradually decreased with time during the first 6 months but increased after 12 months. Juvenile numbers were more constant throughout the year. CPF at the two highest doses significantly reduced both adult and juvenile numbers after 1 month (p < 0.05, Dunnett s test). Juvenile numbers were still reduced at both doses after 2 months but the number of adults was only significantly lower at the highest dose. After 3 months both juveniles and adult numbers were still significantly lower at the highest dose compared to the controls (Figure 2). Figure 3 illustrates the mean abundance of each commonly found earthworm species with sampling time. During the 2-6 month posttreatment period, the mean abundance of Megascolex species was not affected by CPF as the number of individuals found was already rather low or zero. The lowest abundance of Lampito mauritti was found at T 3 after one month of CPF application and in case of Dichogaster bolaui the lowest number was observed two months after CPF application. Lower numbers of Perionyx excavatus were recorded at the highest two doses and after 2-6 months of application and differences were significant when compared to the control plots (p < 0.05, Dunnett s test). Similar trends were also observed for the mean biomass of the individual earthworm species (Figure 4). Termite abundance No significant differences in the abundance of termites were observed between the different plots prior to CPF application. A total number of 96-128 individuals per 0.25 m 2 were found. In the control plots termite abundance slightly varied with time and the highest abundance was recorded after 12 months. Termite abundance was significantly reduced at the two highest CPF doses after 1 month and all three doses after 2 and 3 months. No significant differences in termite abundance were found after 6 months (Figure 2). 77

Chapter 6 Mean abundance ± SD (0.25 m 2 ) 45 40 35 30 25 20 15 10 5 0 50 45 40 35 30 25 20 15 10 5 0 160 140 120 100 80 60 40 20 0 Juvenile earthworms Adult earthworms Termites PA 1 M 2 M 3 M 6 M 12 M Sampling time ( control T 1 T 2 T 3 ) Figure 2: Effects of chlorpyrifos on the development with time of the total abundance (mean ± SE, 0.25 m 2 ) of termites, adult and juvenile earthworms in experimental field plots. * Indicates a significant difference from the control plots (p < 0.05, Dunnett s test) at each sampling time (in months; PA = prior to pesticide application). C = control, T 1, T 2, and T 3 : chlorpyrifos dosages of 0.6, 4.4 and 8.8 kg a.i. ha -1,respectively. 78

Chlorpyrifos causes decreased organic matter decomposition Megascolex sp Perionyx excavatus 12 8 10 7 6 8 5 Mean abundance ±SD (0.25 m 2 ) 6 4 2 0 8 7 6 5 4 3 Lampito mauritii 4 3 2 1 0 25 20 15 10 Dichogaster bolaui 2 1 5 0 PA 1 M 2 M 3 M 6 M 12 M 0 PA 1 M 2 M 3 M 6 M 12 M Sampling time ( control T 1 T 2 T 3 ) Figure 3: Effects of chlorpyrifos on the development with time of the mean abundance (± SE, 0.25 m 2 ) of different earthworm species in experimental field plots. * Indicates a significant difference from the control plots (p < 0.05, Dunnett s test) at each sampling time (in months; PA = prior to pesticide application). C = control, T 1, T 2, and T 3 : chlorpyrifos dosages of 0.6, 4.4 and 8.8 kg a.i. ha -1, respectively. Discussion Organic matter breakdown is an essential function of the soil and litter bag tests are appropriate in evaluating pesticide effects on it (Lavelle et al., 1997; Dinter et al., 2008). Generally plant protection products were shown to cause effects on organic matter breakdown (Knacker et al., 2003; Dinter et al., 2008). The present study also reports significant effects of CPF on organic matter break down during 2-3 months after application at the recommended dose (0.6 kg a.i./ha). In tropical environments, abiotic factors like higher temperature and moisture could play a more important role in decomposition rates than in temperate climates. In turn decomposition rate may depend on soil structure and litter quality (Bargali et al., 1993). Therefore observed effects could be linked to effects of CPF on soil organisms. 79

Chapter 6 1,8 Perionyx excavatus 3000 Megascolex sp 1,6 2500 Mean biomass ±SD (g / 0.25 m 2 ) 1,4 1,2 1 0,8 0,6 0,4 0,2 0 3,5 3 2,5 2 1,5 Lampito mauritii 2000 1500 1000 500 0 6 5 4 3 Dichogaster bolaui 1 2 0,5 1 0 PA 1 M 2 M 3 M 6 M 12 M 0 PA 1 M 2 M 3 M 6 M 12 M Sampling time ( Control T 1 T 2 T 3 ) Figure 4: Effects of chlorpyrifos on the development with time of the mean biomass (± SE, g / 0.25 m 2 ) of different earthworm species in experimental field plots. * Indicates a significant difference from the control plots (p < 0.05, Dunnett s test) at each sampling time (in months; PA = prior to pesticide application). C = control, T 1, T 2, and T 3 : chlorpyrifos dosages of 0.6, 4.4 and 8.8 kg a.i. ha -1, respectively. Effects of CPF on organic matter breakdown have hardly been reported except for Siedentop (1995) and Casabé et el. (2007). Their results are contradictory with the results reported here. Siedentop (1995) found effects of CPF on organic matter breakdown at a dose of 1.92 kg ha -1 using rye straw but the effects were not statistically significant. Application of formulated CPF (0.62 kg a.i /ha), which is similar to the T 1 dose used in the study under tropical conditions, did not have any effects on mass loss of dried lucerne leaves (Casabé et al., 2007). Influence of pesticides on organic matter breakdown under tropical conditions could be masked by other abiotic and biotic factors of the system (Garcia, 2004; Förster et al., 2006; Casabé et al., 2007). The descriptive nature of these limited studies makes them rather difficult to compare due to the many factors that can affect the process (see Knacker et al., 2003). Earthworm abundance and biomass were significantly reduced for 1-3 months after CPF application. This effect coincided with the presence of CPF residues in the soil. At the two higher doses, CPF was degraded with half-lives between 36 and 76 days and completely gone after 3 months. 80

Chlorpyrifos causes decreased organic matter decomposition These half lives are much lower than previously recorded values of 60 and 120 days (Wauchope et al., 1992). But half life in soil may depend on soil type, climate and other factors hence comparing half lives under different field conditions is hardly possible. Still it was clear that Megascolex sp. was not affected by CPF application even at the highest dose. Indeed this species is an endogeic species living in deep burrows; it is present in the litter layer only for feeding. Therefore we expect that the presence of CPF in the soil may trigger avoidance and escape to deep burrows with no noticeable biological effects. In contrast epigeic species Perionyx excavatus was highly affected by CPF during 1-3 months after application and could not be found during 2-3 months in plots receiving the highest dose. De Silva and Van Gestel (Chapter 4) indicated that P. exvavatus could only avoided CPF concentrations > 30 mg a.i kg -1 dry soil in both artificial and natural soils under tropical conditions, explaining the fact that it was affected by the doses applied in this study. Lampito mauritii and Dichogaster bolaui were only slightly affected in the first 2 months at the highest dose applied. Reinecke and Reinecke (2007) reported no significant differences in density of earthworms in orchards and adjacent areas after CPF application (a.i 480 g/l, 1500 ml/ha) but predicted adverse chronic effects. O Halloran et al. (1999) and Booth et al. (2001) could not find any effects of CPF at recommended dose on earthworm communities in New Zealand, but species composition and abiotic factors were very different from those in our study. This present study also reports large effects of CPF on termite abundance (Figure 2) compared to their presence in control plots throughout the study period. The number of termites per 0.25 m 2 was higher than the total number of earthworms though their biomass may be lower. Generally termites are susceptible to CPF as it is widely used as termiticide in the tropics. The colonization of litter by termites may greatly accelerate decomposition (Lavelle et al., 1993; Didham., 1998; Martius et al., 1998). The reported effects of CPF on litter decomposition might be mediated through its toxicity to termites and therefore termite abundance should also be incorporated into future studies under tropical conditions. This study highlights the challenges facing tropical risk assessment using guidelines developed for temperate regions. Garcia (2004) reported that there should be no major set back in implementing guidelines for litterbag tests under tropical conditions except for site selection. However, the present study shows the importance of test duration and litter quality in performing such tests in the tropics. The technical guidelines aim for a duration of 6 months, equivalent to 60% mass loss in the control. This duration is based on the studies carried out so far in temperate regions, where 17 out of 29 studies recorded 60 % mass loss in the controls after 6 months (Dinter et al., 2008). Föster et al. (2006) have also reported a similar mass loss upon the same test duration in Brazil. But Casabé et al. (2007) 81

Chapter 6 reported 60 % mass loss of lucerne leaves already 10 days after application of CPF in an Argentinean soya field. This study also indicated faster mass loss in controls after 2 months. It is obvious that litter quality is a major factor in this faster decomposition, nevertheless it could be predicted that the standard test duration for temperate conditions may not be applicable to tropical conditions. A shorter duration with more frequent sampling may be more appropriate for tropical conditions. Conclusions Chlorpyrifos application may cause adverse effects on organic matter breakdown under tropical conditions. The effects were obvious for 2-3 months after application even at the recommended dose. Decreased decomposition could be linked with effects of CPF on the abundance and biomass of soil organisms such as termites and earthworms. Effects of CPF were greater for the epigeic species Perionyx excavatus than in for the endogeic Megascolex sp. To fit tropical conditions, shorter test duration and a more frequent sampling schedule should be used when performing standardized litter bag tests. Simultaneous studies of earthworm field tests with litter bag tests may give a good impression of pesticide effects on both functional and structural end points in soil and therefore future studies combining both tests should be encouraged. Acknowledgements This work was supported by the Netherlands Foundation for the Advancement of Tropical Research (WB 84-609) residing under the Netherlands Organization for Scientific Research (NWO). We thank H.G. Hector, C.H. Priyantha, K.G. Rayaston, S. Niranjala, T. Premaratne, S. Seniwiratne, D.L.A.R Kumara and A. Galhena for technical assistance in the field. 82

Chapter 7 General Discussion 83

Chapter 7 General Discussion Pesticides and their effects on non-target organisms in tropical regions have drawn relatively little attention in spite of increasing knowledge in temperate regions like North America and Europe. The aim of this thesis therefore was to attempt to fill this gap of knowledge to a certain and significant extent. To achieve this aim, effects of commonly used pesticides in Sri Lanka, as a representative of tropical regions, were studied using earthworms as model organisms. Temperature, soil types and toxicity Climate in tropical temperature regimes is mainly warmer and more uniform (yearly average) than in temperate regions. It has been well identified that temperature is a key element that affects toxicity of chemicals and knowledge on temperature-effect relationships is essential for a proper risk assessment of chemicals in tropical regions. Therefore comparative studies of temperature effects on pesticide toxicity are key issue addressed in this thesis. Garcia (2004) first attempted to compare the toxicity of selected pesticides in temperate and tropical conditions. His results gave more dimensions to current tropical ecotoxicology but the use of a tropical strain of Eisenia fetida hampered direct comparison of temperature effects under both conditions. In Chapter 2 of this thesis, the effect of temperature on the toxicity of chlorpyrifos, carbofuran and carbendazim was described using the same species (E. andrei), soil types and endpoints hence avoiding the use of different ecotypes of the test species. The only differences were the environmental temperatures selected to be 20 0 C (temperate) and 26 0 C (tropical). In general higher temperatures may exacerbate toxicity (Gordon, 2003). This is also linked with additional stress factors that may be enhanced by higher temperatures together with toxicity of the chemical itself. Indeed, for effects on survival, toxicity of chlorpyrifos in both artificial and natural soils was higher at 26 0 C than at 20 0 C (Chapter 2). Earthworms may take chemicals up through skin or by ingestion (Lord et al, 1980; Belfroid et al, 1994). Initial sudden exposure of earthworms to chlorpyrifos may increase their activity as observed in our experiments. As a result they may have higher food intake and more intense contact with soil and pore water. This may trigger higher uptake of pesticides and possibly lead to a higher toxicity. On the other hand degradation kinetics of chemicals is largely depending on temperature; higher temperatures may stimulate microbial activity and enhance degradation, resulting in a reduced toxicity of the parent compound. Carbendazim was more toxic under representative 84

General Discussion temperate conditions than under tropical conditions irrespective of soil types and endpoints measured. This was also supported by Garcia (2004) who reported a factor of 172 differences in carbendazim toxicity to E. fetida under temperate and tropical conditions, which could mainly be explained from the faster degradation of carbendazim at higher temperatures. This faster degradation and resulting lower toxicitiy at 26 0 C were also observed for sub-lethal endpoints with all three pesticides tested. From the results of Chapter 2, it is clear that soil type may also affect toxicity of the pesticides, which could be linked with sorption. Sorption is a key process that governs bioavailability of chemicals and largely depends on soil organic matter content (Spark and Swift, 2002; Coquet, 2003). Therefore natural soils, which have lower organic matter content than artificial soils, as used in this study, may exhibit higher toxicity. Observed higher effects of the three pesticides on reproduction and growth in natural soils irrespective of temperatures could be explained from the lower organic matter content of the used soils especially for LUFA 2.2. In conclusion, temperature could be a critical factor affecting survival of earthworms upon pesticide exposure. For the effects on sub-lethal endpoints, such as reproduction and growth, soil type and the nature of the pesticide seems more important than temperature. The complexity of toxicity under tropical conditions, which ultimately depends on a variety of interacting factors, illustrates the need for caution using temperate data in tropical risk assessments. Species specific sensitivity Species selection in earthworm toxicity tests mainly depends on life cycle and the potential for laboratory culturing. E. fetida and E. andrei have been widely used as they are very easy to breed and maintain under laboratory conditions. Ecological relevance of these species has largely been ignored and remains a concern in ecotoxicity tests. Species specific sensitivity of earthworms to pollutants is well documented and depends mainly on physiological, morphological and ecological characters (Edwards and Bohlen, 1996; Stephenson et al, 1998; Lukkari and Haimi, 2005). In Chapters 3 and 4, it was shown that the sensitivity of E. andrei and the tropical earthworm species Perionyx excavatus greatly varied with endpoints measured. In case of survival, P. excavatus was more sensitive for pesticides than E. andrei. In general P. excavatus is smaller than E. andrei resulting in a larger surface to volume ratio that may lead to faster uptake kinetics. In addition AChE inhibitors used in this study, such as chlorpyrifos and carbofuran, may trigger uptake within an even shorter time. This higher sensitivity may also be linked with higher activity, lower fat content, difference in metabolic capacity and other physiological 85

Chapter 7 characters. But lack of comparative knowledge on all these factors for the two species hampers further conclusions. Contradictory, sub-lethal endpoints such as reproduction and avoidance suggest that the standard species E. andrei is more sensitive than the local species P. excavatus. The standard earthworm reproduction test runs through 56 days including 28 days exposure of the adult worms and an additional 28 days to allow for hatching of the cocoons. Effects on the number of juveniles after pesticide exposure could be due to effects on cocoon production during the first 28 days and as well as effects on cocoon hatchability or juvenile survival. In the long run, especially for effects on hatching and juvenile survival, degradation of pesticides might be more important. Therefore species sensitivity differences to sub-lethal endpoints like reproduction may mostly be masked by the fate of pesticides in soil. This argument may not be hold for the avoidance tests (Chapter 4), which clearly show that E. andrei is more sensitive than P. excavatus. This higher sensitivity could be a result of the choice of the pesticides used. Both chlorpyrifos and carbofuran are AChE inhibitors and may cause paralysis in organisms. The observed attraction of P. excavatus to lower pesticide concentrations remains unsolved, but it is clear that they were less capable of detecting low concentrations of these pesticides than E. andrei. This is not in agreement with work reported by Reinecke et al. (2002) where P. excavatus was more sensitive to mancozeb than Eisenia sp. From these results it can be concluded that species sensitivity is highly dependent on the type of pesticide and the endpoints measured. Linking laboratory data to field Extrapolation of laboratory studies to realistic field conditions is rather difficult mainly due to the masking of the effects in field by other environmental variables. Nevertheless Chapter 6 of this thesis indicates that such linking could be successful. Chlorpyrifos application to soil reduced earthworm and termite abundance and organic matter breakdown even at the recommended dose of 0.6 kg a.i. ha -1. In Chapter 2, it has been shown that NOEC for reproduction and growth is < 1 mg a.i. kg -1 under similar conditions but with E. andrei. The second highest dose (4.4 kg a.i. ha -1 ), which is equivalent to the EC 50 value for the eefects on the reproduction of E andrei (Chapter 2), also coincidently equals the cumulative dose applied during one crop cycle. Due to degradation and crop interception, direct exposure to this dose may not be realistic. Nevertheless such doses may be found considering the unnecessary repeated applications, over-application and improper mixing and handling of spray equipment common to tropical agriculture practices. Van Gestel (1992) also found fairly good agreement between lab and field data but did not include functional endpoints. The observed effect on species like P. excavatus indicates the importance of 86

General Discussion using naturally occurring and ecologically relevant species in toxicity studies. This thesis could only contemplate on a single occasion of linking laboratory data and field conditions and therefore more research under realistic conditions will be necessary. New developments in tropical ecotoxicology This thesis illustrates the need for new tools and considerations for tropical regions. One main obstacle was the lack of suitable tropical substrates to perform toxicity tests. Artificial soil typically representative of temperate regions has been used in toxicity studies for decades and enhanced the uniformity of comparative studies. Further development of such artificial substrates for tropical areas, first initiated by Garcia (2004), was performed and successfully validated with a few selected pesticides. In general, replacing sphagnum peat in tropical countries is not an uneasy task due to the greater availability of materials. Coco peat, saw-dust and paddy husk all proved to be useful in early investigations and finally composted coco peat was selected as suitable replacement for tropical artificial soils (Chapter 5). Further use of the new developed tropical artificial soil needs more research especially with other pesticides, metals and the other organic chemicals with different modes of action. In addition physico-chemical properties of the coco peat may vary with the location and type of processing (Tapia et al., 2008). Therefore standardization of this substrate by ring testing in tropical countries will become a priority in future studies. The results described in Chapter 6 demonstrate the need for modification of standard higher tier test guidelines according to tropical conditions. Organic matter break down was much faster in the tropics than in the temperate regions; therefore time duration should be carefully evaluated after future studies in the region. Litter type, quality and mainly contribution of different soil organisms to organic matter decomposition may also change with different abiotic and biotic factors of the region and should be taken into account in future studies. The biggest obstacle is poor understanding of the role of different organisms in decomposition in tropical ecosystems. It may hamper linking effects on structure and functioning of ecosystems and predicting effects on ecosystem functioning. This was clearly shown in Chapter 6. Decreased abundance of both termite and earthworm populations could be linked to delay in organic matter decomposition but the proportional contribution of each group to soil functions could not be quantified. Nevertheless the use of available test guidelines proved to offer an excellent starting point although slight modifications will be needed to make them applicable to tropical regions. 87

Chapter 7 Avoidance test: obligatory for tropical regions? Pesticide regulation in tropical countries like Sri Lanka is primarily based on risks to human health. The well being of non-target organisms generally is neglected due to the lack of knowledge and proper assessment tools specific for the region. The earthworm avoidance test, which is a complementary screening tool in soil risk assessments (Slimak, 1997; Hund-Rinke et al., 2002), has a short duration and lower costs than the standard acute and chronic tests. In Chapter 4 it is illustrated that the sensitivity of the avoidance test is always higher than that of the earthworm survival test and sometimes even higher than the reproduction test. An increasing number of new or modified pesticides in the form of various formulations are introduced to the tropical markets every year. Regulatory authorities however, still depend on details given by the manufacturers. These details are more relevant to temperate regions where the pesticides have been developed and manufactured. It is accepted that most of the developing countries have no capacity to test each and every chemical that enters to the market. Therefore fast and cheap tests are urgently needed. One good alternative is the inclusion of the earthworm avoidance test as an obligatory in regulation and registration of pesticides to provide a fast assessment of potential pesticide effects on soil organisms. Implications for Sri Lankan scenario In Sri Lanka, the main pesticides registered are organophosphates, carbamates, amides, dithiocarbamates and inorganic compounds. The selection of the chemicals for this thesis was primarily based on the pesticide usage in Sri Lanka and the availability of comparative toxicity data in other regions. Therefore the pesticides most frequently used carbofuran, chlorpyrifos, mancozeb and carbendazim (also as reference chemical) were selected for this study. Based on the endpoints investigated in Chapters 2, 3 and 4, the general order of toxicity of these pesticides was carbofuran > carbendazim > chlorpyrifos > mancozeb. No previous record of the toxicity of these pesticides to earthworms (standard tests developed by OECD and ISO) under tropical conditions existed except for carbendazim (Garcia, 2004). As stated in the general introduction, carbofuran is the most commonly used insecticide / nematicide in Sri Lanka and widely used in variety of crops including rice, vegetables and friuits. The results shown here indicate that it is the pesticide most toxic to earthworms under tropical conditions. Carbofuran, which has been classified as highly hazardous by WHO (2004), has been banned in temperate regions. The current use of the 3% formulation of this chemical in Sri Lanka may vastly reduce its risk to humans but its effects on beneficial soil organisms such as earthworms may still be very high, which 88

General Discussion is well supported by data presented in this thesis. Therefore restricted use or even banning the use of carbofuran in Sri Lanka should be considered. Future Perspectives Tropical countries like Sri Lanka often lack specific country/regional information on pesticides and their effects on soil organisms. This thesis contributes to fulfill this lack of knowledge and information to some extent. It is expected that this work will further contribute to improvement of the scientific evaluation of pesticide side effects on soil organisms. This work also indicates the necessity for more information on pesticide effects as they may depend on different factors like types of pesticide and mode of action, temperature, species, endpoints, soil types and their complex interactions. Future investigations on pesticide effects in tropical regions should be performed also linking with pesticide fate and behaviour in tropical environments. Knowledge obtained through these studies should be helpful for respective authorities in taking decisions on further reducing pesticide use and increasing sustainable crop protection. In addition incorporation of locally available data in future risk assessments is necessary and further collaborative studies within the tropical regions may provide better understanding of the regional aspects. 89

90

References Abad, M., Noguera, P., Bures, S., 2001. National inventory of organic wastes for use as growing media for ornamental potted plant production: case study in Spain. Bioresour. Technol. 77, 197-200. Abad, M., Noguera, P., Puchades, R., Maquieira, A., Noguera, V., 2002 Physico-chemical and chemical properties of some coconut coir dusts for use as a peat substitute for containerized ornamental plants. Bioresour. Technol. 82, 241 245. Abdullah, A.R., Bajet, C.M., Matin, M.A., Nhan, D.C., Sulaiman, A.H., 1997. Ecotoxicology of pesticides in the tropical paddy field ecosystem. Environ. Toxicol. Chem. 16, 59 70. Adema, D.M., Barug, D., Vonk, J.W., 1985. Comparison of the effects of several chemicals on micro-organisms, higher plants and earthworms. In: Actes Symposium Intl. Ecotoxicologie Terrestre 199-214. Aldaya, M.M., Lors, C., Salmon, S, Ponge, J.F., 2006. Avoidance bioassays may help to test the ecological significance of soil pollution. Environ. Pollut. 140,173-180. Amorim, M.J.B., Römbke, J., Soares, A.M.V.M., 2005. Avoidance behaviour of Enchytraeus albidus: Effects of Benomyl, Carbendazim, Phenmedipham and different soil types. Chemosphere 59, 501 510. An, Y. J., 2005. Assessing soil ecotoxicity of methyl tert-butyl ether using earthworm bioassay; closed soil microcosms test for volatile organic compounds. Environ. Pollut. 134, 181 186. An, Y.J., Lee, W.M., 2008. Comparative and combined toxicities of toluene and methyl tert-butyl ether to an Asian earthworm Perionyx excavatus. Chemosphere 71, 407-411. AOAC, 2000. Official Methods of Analysis. Association of Official Analytical Chemist. Gaithersburg, Maryland, USA. Arias, E. 2003. Sister chromatid exchange induction by the herbicide 2, 4 dichlorophenoxyacetic acid in chick embryos. Ecotoxicol. Environ. Safety 55, 338 343. Barateiro Diogo, J., Natal-da-Luz, T., Sousa, J.P., Vogt, C., Nowak, C., 2007. Tolerance of genetically characterized Folsomia candida strains to phenmedipham exposure - A comparison between reproduction and avoidance tests. J. Soils Sediments 7, 388-392. Bargali, S.S., Singh, S.P., Singh, R. P., 1993. Patterns of weight loss and nutrient release from decomposing leaf litter in an age series of Eucalypt plantations. Soil Biol. Biochem. 25, 1731 1738. Barkham, J.P., 1993. For peat's sake: conservation or exploitation? Biodiv. Conserv. 2, 556 566. Barr, C.J., 1993. Countryside survey 1990: main report. Department of Environment. Eastcote, UK. 91

Belfroid, A., Meiling, J., Sijm, D., Hermens, J., Seinen, W., Van Gestel K., 1994. Uptake of hydrophobic halogenated aromatic hydrocarbons from food by earthworms (Eisenia andrei). Arch. Environ. Contam. Toxicol. 27, 260-265. Benton, T.G., Vickery, J.A., Wilson, J.D., 2003. Farmland biodiversity - Is habitat heterogeneity the key? Trends Ecol. Evolut. 18, 182 188. Bhattacharjee, G., Chaudhuri, P.S., 2002. Cocoon production, morphology, hatching pattern and fecundity in seven tropical earthworm species a laboratory-bases investigation. J. Biosci. 27, 283 294. Bianchi, F. J. J. A., Booij, C. J. H., Tscharntke, T., 2006. Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proc. Royal Soc. B 273, 1715-1727. Blakemore, R.J., 2006. A Series of Searchable Texts on Earthworm Biodiversity, Ecology and Systematics from Various Regions of the World 2nd Edition and Supplement (2006). General Eds.: N. Kaneko & M.T. Ito. COE Soil Ecology Research Group, Yokohama National University, Japan. Booth, L. H., Hodge, S., O Halloran, K., 2001. Use of biomarkers in earthworms to detect use and abuse of field applications of a model organophosphate pesticide. Bull. Environ. Contam. Toxicol. 67, 633 640. Bourdeau, P., Haines, J.A., Klein, W., Krishnamurti, C.R., 1989. Ecotoxicology and Climate. SCOPE 38 IPCS Joint Symposia 9, Chichester, UK. 392. Callahan, C. A., Shirazi, M.A., Neuhauser, E.F., 1994. Comparative toxicity of chemicals to earthworms. Environ. Toxicol. Chem. 13, 291-298. Carvalho, F., 2006. Agriculture, pesticides, food security and food safety. Environ. Sci. Policy 9, 685 692. Casabé, N., Piola, L., Fuchs, J., Oneto, M. L., Pamparato, L., Basack, S., Giménez, R., Massaro, R., Papa, J. C., Kesten, E., 2007. Ecotoxicological assessment of the effects of glyphosate and chlorpyrifos in an Argentine soya field. J. Soils Sediments 7, 232 239. Castillo, L.E., De la Cruz, E., Ruepert, C., 1997. Ecotoxicology and pesticides in tropical aquatic ecosystems of Central America. Environ. Toxicol. Chem. 16, 41 51. Chamberlain, D. E., Fuller, R.J., Bunce, R.G.M., Duckworth, J.C. Shrubb, M., 2000. Changes in abundance of farmland birds in relation to the timing of agricultural intensification in England and Wales. J Appl. Ecol. 37, 771 788. Chandrasekara, L.W.H.U., Pathiratne, A., 2007. Body size-related differences in the inhibition of brain acetylcholinesterase activity in juvenile Nile Tilapia (Oreochromis niloticus) by chlorpyrifos and carbosulfan. Ecotoxicol. Environ. Safety 67, 109 119. 92

References Chang, J., 1968. The agricultural potential of the humid tropics. Geog Rev. 58, 333 361. Chaudhuri, P.S., Bhattacharjee, G., 2002. Capacity of various experimental diets to support biomass and reproduction of Perionyx excavatus. Bioresour. Technol. 82, 147 150. Chweya, J.A., Gurnah, A.M., Fisher, N.M., 1978. Preliminary studies on some local materials for propagation media. 2. Trials with mixtures containing local materials. East Afr. Agric. For. J. 43, 334 342. Cohen, J.E.., 2005. Human population grows up. Sci. Am. 293, 48-55. Coquet, Y., 2003. Sorption of pesticide atrazine, isoproturon, and metamitron in the Vadose Zone. Vadose Zone J. 2, 40 51. Cox, C., Surgan, M. 2006. Unidentified inert ingredients in pesticides: Implications for human and environmental health. Environ. Health Perspect. 114, 1803-1806. Crop Life International, 2008. Facts and figures - The status of global agriculture. Crop Life International, Brussels, Belgium. Daam, M. A., Satapornvanit, K., Van den Brink, P.J., Nogueira, A.I.A., 2009a. Sensitivity of macroinvertebrates to carbendazim under semifield conditions in Thailand: Implications for the use of temperate toxicity data in a tropical risk assessment of fungicides. Chemosphere 74, 1187-1194. Daam, M.A., Van den Brink, P.J., Nogueira, A.J.A., 2009b. Comparison of fate and ecological effects of the herbicide linuron in freshwater model ecosystems between tropical and temperate regions. Ecotoxicol. Environ. Safety 72, 424-433. Darwin, C. R., 1881. The formation of vegetable mould, through the action of worms, with observations on their habits. John Murray, London. Deshmukh, I., 1986. Ecology and Tropical Biology. Blackwell, Palo Alto, CA, USA. De Silva, P.M.C.S., Samayawardhena, L.A., 2005. Effects of chlorpyrifos on reproductive performances of guppy (Poecilia reticulata). Chemosphere 58, 1293-1299. De Silva, P.M.C.S., Senewirathne, S.H.S., 2008. Turn of the worms: Assessment of Bavistin FL (active ingredient Carbendazim) toxicity in Earthworm, Eisenia andrei under tropical conditions. Proceedings of the fifth academic sessions, University of Ruhuna, Matara, Sri Lanka. 13-17. De Silva, P.M.C.S., Pathiratne, A., Van Gestel, C.A.M., 2009. Influence of temperature and soil type on the toxicity of three pesticides to Eisenia andrei. Chemosphere 76, 1410-1415. Di Benedetto, A., 2007. Alternative substrates for potted ornamental plants based on Argentinean peat and Argentinean river waste: a review. Floric. Plant Ornamental Biotechnol. 1, 90 101. 93

Didham, R.K., 1998. Altered leaf-litter decomposition rates in tropical forest fragments. Oecologia 116, 397-406. Dinter, A., Coulson, M., Heimbach, F., Keppler, J., Krieg, W., Kolzer U., 2008. Technical experiences made with the litter bag test as required for the risk assessment of plant protection products in soil. J. Soils Sediments 8, 333-339. Ecobichon, D.J., 2001. Pesticide use in developing countries. Toxicology.160, 27-33. Edwards, C.A., Fletcher, K.E., 1988. Interactions between earthworms and microorganisms in organic-matter breakdown. Agric. Ecosyst. Environ. 24, 235 247. Edwards, C.A., Bohlen, P.J., 1993. The effects of toxic chemicals on earthworms. Rev. Environ. Contam. Toxicol. 125, 23-99. Edwards, C.A., Bohlen P.J., 1996. Biology and Ecology of Earthworms, Chapman and Hall, London. Edwards, C.A. 1998. Earthworm Ecology. St. Lucie Press, New York. Edwards, C.A., Dominguez, J., Neuhauser, E.F., 1998. Growth and reproduction of Perionyx excavatus (Perr.) (Megascolecidae) as factors in organic waste management. Biol. Fertil. Soils 27, 155 161. Ekschmitt, K., Klein, A., Pieper, B., Wolters, V., 2001. Biodiversity and functioning of ecological communities - why is diversity important in some cases and unimportant in others? J. Plant Nutr. Soil Sci. 164, 239-246. Environment Canada, 2004. Biological test method: Tests for toxicity of contaminated soil to earthworms (Eisenia andrei, Eisenia fetida, or Lumbricus terrestris). Report EPS 1/RM/43, Environment Canada, Ottawa, Ontario, Canada. Eriksen-Hamel, N.S., Whalen, J.K., 2007. Impacts of earthworms on soil nutrients and plant growth in soybean and maize agroecosystems. Agric. Ecosyst. Environ. 120, 442 448. Evans, M. R., Karcher, D., 2004. Properties of plastic, peat, and processed poultry feather fiber growing containers. Hort. Science 39, 1008-1011. FAO, 2000. Production Yearbook 1999. FAO Statistics series No. 53. Food and Agriculture Organization, Rome, Italy. FAO, 2005. Proceedings of the Asia regional workshop. Regional office for Asia and the Pacific, Bangkok, Thailand. Federschmidt, A., 1994. Die Oligochätenfauna zweier Ökosysteme auf Lößlehm unter Berücksichtigung der Auswirkungen von Chemikalienstress. Ph.D. thesis. Johann Wolfgang Goethe Universität, Frankfurt, Germany. Fitzgerald, D.G., Warner, K.A., Lanno, R.P., Dixon, D.D., 1996. Assessing the effects of modifying factors on pentachlorophenol toxicity to earthworms: applications of body residues. Environ. Toxicol. Chem. 15, 2299-2304. 94

References Förster, B., Garcia, M., Francimari, O., Römbke, J., 2006. Effects of carbendazim and lambda-cyhalothrin on soil invertebrates and leaf litter decomposition in semi-field and field tests under tropical conditions (Amazonia, Brazil). Eur. J. Soil Biol. 42, 171-179. Frampton, G.K., Jänsch, S., Scott-Fordsman, J.J., Römbke, J., Van den Brink, P.J.,2006. Effects of pesticides on soil invertebrates in laboratory studies: a review and analysis using species sensitivity distributions. Environ. Toxicol. Chem. 25, 2480 2489. Frolking, S., Roulet, N.T., Moore, T.R., Richard, P.J.H., Lavoie M., Muller, S.D., 2001 Modeling northern peat-land decomposition and peat accumulation. Ecosystems 4, 479 498. Garcia, M.V.B., 2004. Effects of pesticides on soil fauna: development of ecotoxicological test methods for tropical regions. Ecology and Development Series, vol. 19. University of Bonn, Germany. Garcia, M., Römbke, J., de Brito, M.T., Scheffczyk, A., 2008. Effects of three pesticides on the avoidance behavior of earthworms in laboratory tests performed under temperate and tropical conditions. Environ. Pollut. 153, 450-456. Gates, G.E., 1945. On some earthworms from Ceylon. II. Spolia zeylan. 24, 69-90. Getzin, L.W., 1981. Degradation of chlorpyrifos in soil: influence of autoclaving, soil moisture, and temperature. J. Econ. Entomol. 74, 158-162. Gordon, C.J., 2003. Role of environmental stress in the physiological response to chemical toxicants. Environ. Res. 92, 1-7. Guruge, K.S., Tanabe, S., 2001. Contamination by persistent organochlorines and butyltin compounds in the west coast of Sri Lanka. Mar. Pollut. Bull. 42, 179-186. Haanstra, L., Doelman, P., Oude Voshaar, J.H., 1985. The use of sigmoidal dose response curves in soil ecotoxicological research. Plant and Soil 84, 293-297. Hallatt, L., Reinecke, A.J., Viljoen, S.A., 1990. Life-cycle of the oriental compost worm Perionyx excavatus (Oligochaeta). S. Afr. J. Zool. 25, 41 45. Hamilton, M.A., Russo, R.C., Thurston, R.V., 1977. Trimmed Spearman- Karber method for estimating median lethal concentrations in toxicity bioassays. Environ. Sci. Technol. 11, 714-719. Haque, A., Ebing, W., 1983. Toxicity determination of pesticides to earthworms in the soil substrate. Z. Pflanzenkrankh Pflanzenschutz. 90, 395-398. Heimbach, F., 1985. Comparison of laboratory methods using Eisenia fetida and Lumbricus terrestris for the assessment of hazard of chemicals to earthworms. Z. Pflanzenkrank. Pflazenschutz. 92, 186-93. 95

Heimbach, F., 1992. Correlation between data from laboratory and field tests for investigating the toxicity of pesticides to earthworms. Soil Biol. Biochem. 24, 1749-1753. Helling, B., Reinecke, S.A., Reinecke, A.J., 2000. Effects of the fungicide copper oxychloride on the growth and reproduction of Eisenia fetida (Oligochaeta). Ecotoxicol. Environ. Saftey 46, 108-116. Heupel, K., 2002. Avoidance response of different collembolan species to Betanal. Eur. J. Soil Biol. 38, 273-276. Hodge, S., Webster, K.M., Booth, L., Hepplethwaite, V., O Halloran, K., 2000. Non-avoidance of organophosphate insecticides by the earthworm Aporrectodea calignosa (Lumbricidae). Soil Biol. Biochem. 32, 425-428. Holmstrup, M., 2000. Field assessment of toxic effects on reproduction in the earthworms Aporrectodea longa and Aporrectodea rosea. Environ. Toxicol. Chem. 19, 1781-1787. Hund-Rinke, K., Wiechering, H., 2001. Earthworm avoidance test for soil assessment. J. Soils Sediments 1,15 20. Hund-Rinke, K., Kordel, W., Hennecke, D., Eisentraeger A., Heiden, S., 2002. Bioassays for the ecotoxicological and genotoxicological assessment of contaminated soils (Results of a round robin test): Part I. Assessment of a possible groundwater contamination: ecotoxicological and genotoxicological tests with aqueous soil extracts. J. Soils Sediments 2, 43-50. Hund-Rinke, K., Achazi, R., Römbke, J., Warnecke D., 2003. Avoidance test with Eisenia fetida as indicator for the habitat function of soils: results of a laboratory comparison test. J. Soils Sediments 3, 1-6. Hund-Rinke, K., Simon, M., 2005. Terrestrial ecotoxicity of eight chemicals in a systematic approach. J. Soils Sediments 5, 59 65. Hund-Rinke, K., Lindemann, M., Simon, M., 2005. Experiences with novel approaches in earthworm testing alternatives. J. Soils Sediments 5, 233 239. ISO, 1993. Soil quality- Effects of pollutants on earthworms (Eisenia fetida) Part 1: Determination of acute toxicity using artificial soil substrate. ISO 11268-1. International Organization for Standardization, Geneva, Switzerland. ISO, 1994. Soil quality- Determination of ph. ISO 11465. International Organization for Standardization, Geneva, Switzerland. ISO, 1998a. Soil quality- Determination of dry bulk density. ISO 11272. International Organization for Standardization Geneva, Switzerland. ISO, 1998b. Soil quality-effects of pollutants on earthworms (Eisenia fetida) part 2: Determination of effects on reproduction. ISO 11268-2. International Organization for Standardization, Geneva, Switzerland. 96

References ISO, 1999a. Soil quality - effects of pollutants on earthworms, Part 3: Guidance on the determination of effects in field situations. ISO 11268-3: International Organization for Standardization, Geneva, Switzerland. ISO, 1999b. Soil Quality Inhibition of reproduction of Collembola (Folsomia candida) by soil pollutants. ISO 11267. International Organization for Standardization, Geneva, Switzerland. ISO, 2007. Soil quality: Avoidance test for testing the quality of soils and effects of chemicals on behavior- Part 1: Test with earthworms (Eisenia fetida and Eisenia andrei). ISO/DIS 17512-1.2, International Organization for Standardization, Geneva, Switzerland. Jänsch, S., Amorim, M.J., Römbke, J., 2005. Identification of the ecological requirements of important terrestrial ecotoxicological test species. Environ. Rev. 13, 51 83. Jänsch, S., Frampton, G., Römbke, J., Van den Brink, P., Scott-Fordsmand, J.,2006. Effects of pesticides on soil invertebrates in model ecosystem and field studies: A review and comparison with laboratory toxicity data. Environ. Toxicol. Chem. 25, 2490 2501. Joshi, N., Dabral, M., 2008. Life cycle of earthworms Drawida nepalensis, Metaphire houlleti and Perionyx excavatus under laboratory controlled conditions. Life Science Journal. 5, 83-86. Karlsson, S.I., 2004. Institutionalized knowledge challenges in pesticide governance: the end of knowledge and beginning of values in governing globalized environmental issues. Int. Environ. Agreem. Polit. La. E. 4, 195 213. Kennel,W., 1972 Schadpilze als Objekte integrierter Pflanzenschutzmassnahmen im Obstbau. Z. Pflanzenkrankh. Pflanzenpath. Pflanzenschutz 79, 400-406. Kennel, W., 1990. The role of the earthworm Lumbricus terrestris in integrated fruit production. Acta Horticulturae 285, 149-156. Kiparissis, Y., Metcalfe, T.L., Balch, G.C., Metcalfe, C.D. 2003. Effects of the antiandrogens, vinclozolin and cyproterone acetate on gonadal development in the Japanese medaka (Oryzias latipes). Aquat. Toxicol. 63, 391 403. Klassen, W., 1995. World food security up to 2010 and the global pesticide situation. Proceedings of the eighth international congress on pesticide chemistry. American Chemical Society, Washington DC. USA. Klein, W., 1989. Mobility of environment chemicals including abiotic degradation. In: Bourdeau P, Haines JA, Klein W, Krishanamurti CR. Editors. Ecotoxicology and Climate with Special Reference to Hot and Cold Climates. John Wiley and Sons. New York. Pp. 65-78. Knacker, T., Förster, B., Römbke, J., Frampton, G.K., 2003. Assessing the effects of plant protection products on organic matter breakdown in arable fields - litter decomposition test system. Soil Biol. Biochem. 35, 1269 1287. 97

Kwok, K.W.H., Leung, K.M.Y., Lui, G.C.S., Chu, V.K.H., Lam, P.K.S., Morritt, D., Maltby, L., Brock T.C.M., Van Den Brink, P.J., Warne, M.St.J., Crane, M. 2007. Comparison of tropical and temperate freshwater animal species acute sensitivity to chemicals: Implication for deriving safe extrapolation factor. Integr. Environ. Assess. Manage. 3, 49 67. Kula, H., 1995. Comparison of laboratory and field testing for the assessment of pesticide side effects on earthworms. Acta Zool. Fennica. 196, 338 341. Laabs, V., Amelung, W., Pinto, A., Zech, W., 2002. Fate of pesticides in tropical soils of Brazil under field conditions. J. Environ. Qual. 31, 256-68. Lacher, T.E., Goldstein, M.I., 1997. Tropical ecotoxicology: Status and needs. Environ. Toxicol. Chem. 16, 100-111. Lavelle, P., 1988. Earthworm activities and the soil system. Biol. Fertil. Soils 6, 237 251. Lavelle, P., Blanchart, E., Martin, A., Martin, S., Spain, A., Toutain, F., Barois, I., Schaefer, R., 1993. A hierarchical model for decomposition in terrestrial ecosystems Application to soils of the humid tropics. Biotropica 25, 130 150. Lavelle, P., Bignell, D., Lepage, M., Wolters, V., Roger, P., Ineson, P., Heal, O.W., Dhillion, S., 1997. Soil function in a changing world: the role of invertebrate ecosystem engineers. Eur. J. Soil Biol. 33, 159 193. Lee, K.E., 1985. Earthworms: Their Ecology and Relationships with Soils and Land Use. Academic Press, Sydney. Lord, K.A., Briggs, G.G., Neale, M.C., Manlove, R., 1980. Uptake of pesticides from water and soil by earthworms. Pestic. Sci. 11, 401-408. Loureiro, S., Soares, A.M.V.M., Nogueira, A.J.A., 2005. Terrestrial avoidance behaviour tests as screening tool to assess soil contamination. Environ. Pollut.138, 121-131. Løkke, H., van Gestel, C.A.M., 1998. Handbook of soil invertebrate toxicity tests. Chichester, Wiley & Sons. Lukkari, T., Haimi, J., 2005. Avoidance of Cu and Zn contaminated soil by three ecologically different earthworm species. Ecotoxicol. Environ. Safety 62, 35-41. Lukkari, T., Aatsinki, M., Vaisanen, A., Haimi, J., 2005. Toxicity of copper and zinc assessed with three different earthworm tests. Appl. Soil Ecol. 30, 133-146. Ma, W.C., Bodt, J., 1993. Differences in toxicity of the insecticide Chlorpyrifos to six species of earthworms (Oligochaeta, Lumbricidae) in standardized soil tests. Bull. Environ. Contam. Toxicol. 50, 864 870. Maboeta, M.S., Reinecke, A.J., Reinecke, S.A., 1999. Effects of low levels of lead on growth and reproduction of the Asian earthworm Perionyx excavatus (Oligochaeta). Ecotox. Environ. Safety 44, 236-240. 98

References Maltby, L., Blake, N.N., Brock, T.C.M., Van den Brink, P.J. 2005. Insecticide species sensitivity distributions: The importance of test species selections and relevance to aquatic ecosystems. Environ. Toxicol. Chem. 24, 379 388. Mann, R.M., Bidwell, J.R. 1999. The toxicity of glyphosate and several glyphosate formulations to four species of Southwestern Australian frogs. Arch. Environ. Contam. Toxicol. 36, 193 199. Marchese, N., Di Benedetto, A., Lavado, R., 2006. The possibilities of river waste and Argentinean peat as a plug growing media for Verbena hybrida. Int. J. Agric. Res. 1, 142 150. Martius, C., Höfer, H., Garcia, M., Römbke, J., Hanagarth, W., 2004. Litterfall, litter stocks and decomposition rates in rainforest and agroforestry sites in central Amazonia. Nutr. Cycl. Agroecosyst. 68, 137 154. Marques, C., Pereira, R., Goncalves, F., 2009. Using earthworm avoidance behaviour to assess the toxicity of formulated herbicides and their active ingredients on natural soils. J. Soils Sediments 9, 137-147. Matson, P.A., Parton, W.J., Power, A.G., Swift, M.J., 1997. Agricultural intensification and ecosystem properties. Science 277, 504 509. Meerow, A.W., 1994. Growth of two subtropical ornamentals using coir (coconut mesocarp pith) as a peat substitute. Hort. Science 29, 1484 1486. Mohamed, F., Gawarammana, I., Robertson, T.A., Roberts, M.S., Palangasinghe, C., Zawahir, S., Jayamanne, S., Kandasamy, J., Eddleston, M., Buckley, N.A., Dawson, A.H., Roberts, D.M., 2009. Acute human self-poisoning with Imidacloprid compound: A neonicotinoid insecticide. PLoS ONE 4, e5127. Muthukaruppan, G., Janardhanan, S., Vijayalakshmi, G., 2004. Sublethal toxicity of the herbicide butachlor on the earthworm Perionyx sansibaricus and its histological changes. J. Soils Sediments 5, 82 86. Natal Da Luz, T., Ribeiro, R., Sousa, J.P., 2004. Avoidance tests with Collembola and earthworms as early screening tools for site-specific assessment of polluted soils. Environ. Toxicol. Chem. 23, 2188-2193. Noguera, P., Abad, M., Noguera, V., Puchades, R., Maquieira, A., 2000. Coconut coir waste, a new and viable ecologically-friendly peat substitute. Acta Hortic. 517, 279 286. OECD, 1984. Guideline for testing of chemicals No 207, Earthworm Acute Toxicity Test. Organization for Economic Co-Operation and Development, Paris, France. OECD, 2004a. Guideline for testing of chemicals No 222, Earthworm Reproduction Test (Eisenia fetida/andrei). Organization for Economic Co-Operation and Development, Paris, France. 99

OECD, 2004b. Guideline for testing of Chemicals No 220. Enchytraeid reproduction test. Organization for Economic Co-Operation and Development. Paris, France. OECD, 2006. Guidance document on the breakdown of organic matter in litter bags. OECD series on testing and assessment No 56. Organization for Economic Co-Operation and Development Paris, France. Offord, C.A., Muir, S., Tyler, J.L., 1998. Growth of selected Australian plants in soilless media using coir as a substitute for peat. Aust. J. Exp. Agric. 38, 879 887. O Halloran, K., Booth, L.H., Hodge, S., Thomsen, S., Wratten, S.D., 1999. Biomarker responses of the earthworm Aporrectodea caliginosa to organophosphates: hierarchical tests. Pedobiologia 43, 646 651. Owojori, O.J., Reinecke, A.J., Rozanov, A.B., 2008. Effects of salinity on partitioning, uptake and toxicity of zinc in the earthworm Eisenia fetida. Soil Biol. Biochem. 40, 2385 2393. Owojori, O.J., Reinecke, A.J. 2009. Avoidance behaviour of two ecophysiologically diffrent earthworms. Chemosphere. 75,279-283. Owojori O.J., Reinecke, A.J., Voua-Otomo, P., Reinecke, S.A., 2009. Comparative study of the effects of salinity on life-cycle parameters of four soil dwelling species (Folsomia candida, Enchytraeus doerjesi, Eisenia fetida and Aporrectodea caliginosa). Pedobiologia 52, 351-360. Panda, S., Sahu, S.K., 2002. Acute toxicity assessment of three pesticides to the earthworm Drawida willsi. J. Ecotox. Environ. Monitoring 12, 215-223. Pereira, J.L., Antunes, S.C., Castro, B.B., Marques, C.R., Goncalves, A,M.M., Goncalves, F., Pereira, R., 2009. Toxicity evaluation of three pesticides on non-target aquatic and soil organisms: commercial formulation versus active ingredient. Ecotoxicology 18, 455-463. Perera, P.M., Shahmy, S., Gawarammana, I., Dawson, A.H., 2008. Comparison of two commonly practiced atropinisation regimens in acute organophosphorus and carbamate poisoning, doubling doses vs ad hoc - a prospective observational study. Hum. Exp. Toxicol. 27, 513-518. Pimentel, D., Stachow, U., Takacs, D.A., Brubaker, H.W., Dumas, A.R., Meaney, J.J., O'Neil, J.A.S., Onsi, D.E., Corzilius, D.B., 1992. Conserving biological diversity in agricultural / forestry systems. BioScience 42, 354 362. Racke, K.D., Coats, J.R., 1988. Comparative degradation of organophosphorus insecticides in soil: specificity of enhanced microbial degradation. J. Agric. Food Chem. 36, 193-199. Racke, K.D., Skidmore, M.W., Hamilton, D.J, Unsworth, J.B., Miyamoto, J., Cohen, S.Z., Holland, P.T., 1997. Pesticide fate in tropical soils. Pure Appl. Chem. 69, 1349-1371. 100

References Rao, P.S.C., Mansell, R.S., Baldwin, L.B., Laurent, M.F., 1983. Pesticides and their behavior in soil and water. soil science fact sheet. Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. USA. Rao, J.V., Kavitha, P., 2004. Toxicity of azodrin on the morphology and acetyl cholinesterase activity of the earthworm Eisenia fetida. Environ. Res. 96, 323 327. Reddy, R.D., Purushotham, K.R., Ramamurthi, R., 1984. Sublethal toxicity of fenitrothion to the South Indian earthworm, Lampito mauritii (Kinberg). Nat. Acad. Sci. Lett. India 7, 227-229. Reinecke, A.J., Hallatt, L., 1989. Growth and cocoon production of Perionyx excavatus (Oligochaeta). Biol. Fertil. Soils 8, 303 306. Reinecke, A.J., Reinecke, S.A., Maboeta, M.S., 2001. Cocoon production and viability as endpoints in toxicity testing of heavy metals with three earthworm species. Pedobiologia 45, 61-68. Reinecke, A.J., Maboeta, M.S., Vermeulen L.A., Reinecke, S.A., 2002. Assessment of lead nitrate and mancozeb toxicity in earthworms using the avoidance response. Bull. Environ. Contam. Toxicol. 68, 779-786. Reinecke, S.A., Reinecke, A.J., 2007. The impact of organophosphate pesticides in orchards on earthworms in the Western Cape, South Africa. Ecotoxicol. Environ. Safety 66: 244-251. Roberts, D.M., Karunarathna, A., Buckley, N.A., Manuweera, G., Sheriff, M.H.R., Eddleston, M., 2003. Influence of pesticide regulation on acute poisoning deaths in Sri Lanka. Bull. World Health Org. 81, 789-798. Robertson, A., 1993. Peat, horticulture and environment. Biodiv. Conserv. 2, 541 47. Robinson, R.A., Sutherland, W.J., 2002. Post war changes in arable farming and biodiversity in Great Britain. J. Appl. Ecol. 39, 157-176. Rodriguez-Castellanos, L., Sanchez-Hernandez, J.C., 2007. Earthworm biomarkers of pesticide contamination: current status and perspectives. J. Pestic. Sci. 32 360 371. Römbke,J., Bauer, C., Marschner, A., 1996. Hazard assessment of chemicals in soil. Proposed ecotoxicological test strategy. Environ. Sci. Pollut. Res. Int. 3, 78-82. Römbke, J., Heimbach, F., Hoy, S., Kula, C., Scott-Fordsmand, J., Sousa, P., Stephenson, G., Weeks, J., 2003. Effects of plant protection products on functional endpoints in soil (EPFES). SETAC, Lisbon, 24 26 April 2002. Pensacola FL, USA. Römbke, J., Förster, B., Jänsch, S., Scheffczyk, A., Garcia, M., 2005. Terrestrische ökotoxikologische Testmethoden für die Tropen Teil 2: Halbfreiland- und Freilandtests sowie Risikobeurteilung. UWSF 17, 85 93. 101

Römbke, J., Garcia, M.V., Scheffczyk, A., 2007. Effects of the fungicide benomyl on earthworms in laboratory tests under tropical and temperate conditions. Arch. Environ. Contam. Toxicol. 53, 590-598. Salminen, J., Eriksson, I., Haimi, J., 1996. Effects of terbuthylazine on soil fauna and decomposition processes. Ecotoxicol. Environ. Safety 34, 184 189. Samarawickrema, N., Pathmeswaran, A., Wickremasinghe, R., Peiris-John, R., Karunaratna, M., Buckley, N., Dawson, A., De Silva, J., 2008. Fetal effects of environmental exposure of pregnant women to organophosphorus compounds in a rural farming community in Sri Lanka. Clin. Toxicol. 46, 489-495. Santhanam, K.S.V., Radha, P.K., Limaye, N.M., 1992. Pb 2+ uptake by Lampito mauritii by in vivo bioelectrochemistry. Bioelectrochem. Bioenergetics 27, 99-104. Satapornvanit, K., Baird, D.J., Little, D.C., Milwain, G.K., Van den Brink, P.J., Beltman, W.H.J., Nogueira, A.J.A., Daam, M.A., Domingues, I., Kodithuwakku, S.S., Perera, M.W.P., Yakupitiyage, A., Sureshkumar, S.N., Taylor, G.J., 2004. Risks of pesticide use in aquatic ecosystems adjacent to mixed vegetable and monocrop fruit growing areas in Thailand. Austral. J. Ecotoxicol. 10, 85 95. Senapati, B.K., Biswal, J., Sahu, S.K., Pani, S.C., 1991. Impact of malathion on Drawida willsi Michaelsen, a dominant earthworm in Indian rice fields. I. Survival, population dynamics, reproductive strategy and secondary production. Pedobiologia 35, 117-128. Schaefer, M., 2001. Earthworms in crude oil contaminated soils. Toxicity tests and effects on crude oil degradation. AEHS Contaminated Soil, Sediment and Water, August 2001, 35-37. Schaefer, M, 2004. Assessing 2,4,6-trinitrotoluene (TNT)-contaminated soil using three different earthworm test methods. Ecotoxicol. Environ. Safety 57, 74-80. Siedentop, S., 1995. A litterbag-test for the assessment of side effects of pesticides on soil mesofauna. Acta Zoologica Fennica 196, 357 360. Slimak, K.M., 1997. Avoidance response as a sub lethal effect of pesticides on Lumbricus terrestris (Oligochaeta). Soil Biol. Biochem. 29, 713 715. Singh, B.K., Walker, A., Morgan, J.A.W., Wright, D.J., 2003. Effect of soil ph on the biodegradation of chlorpyrifos and isolation of a chlorpyrifos-degrading bacterium. Appl. Environ. Microbiol. 69, 198 5206. Singh B.K., Walker, A., 2006. Microbial degradation of organophosphorus compounds. FEMS Microbiol. Rev. 30, 428 471. SLS, 2003. Specification for compost from municipal solid wastes and agricultural wastes. Sri Lanka Standard Institution, Colombo, Sri Lanka. SLS 1246. 102

References Sokal, R.R., Rohlf, F.J., 1995. Biometry. W. H. Freeman and Company, New York. Spark, K.M., Swift, R.S., 2002. Effect of soil composition and dissolved organic matter on pesticide sorption. Sci. Total Environ. 298, 147 161. Spurgeon, D.J., Weeks, J.M., van Gestel, C.A.M., 2003. A summary of eleven years progress in earthworm ecotoxicology. Pedobiologia 47, 588-606. Stephenson, J., 1923. Oligochaeta. The Fauna of British India, Including Ceylon and Burma. Taylor and Francis, London. 518 pp. Stephenson, G., Kaushik, A., Kaushik, N.K., Solomon, K.R., Steele, T. and Scroggins, R.P., 1998. Use of an avoidance-response test to assess the toxicity of contaminated soils to earthworms. In: Sheppard, S.C., Holmstrup J.D, Posthuma, L. (Eds): Advances in earthworm ecotoxicology, SETAC Press, Pensacola, FL, pp 67 81. Stone, R., 1995. Ecosystem research: Taking a new look at life through a functional lense. Science 269, 316-317. Storer, D. A., 1984. A simple high sample volume ashing procedure for determining soil organic matter. Commun. Soil Sci. Plant Anal. 15, 759-772. Swann, J.M., Schultz, T.W., Kennedy, J.R., 1996. The effects of the organophosphorus insecticides Dursban and Lorsban on the ciliated epithelium of the frog palate in vitro. Arch. Environ. Contam. Toxicol. 30, 188 194. Tapia, P.V., Ramos, J.Z.C., Garcia, P.S., Chavez, L.T., Romero, R.M.L., Arredondo, J.L.O., 2008. Physical, chemical and biological characterization of coir dust. Revista Fitotecnia Mexicana 31, 375-381. UN, 2001. World Population Prospects. The 2000 Revision. Population Division, Department of Economic and Social Affairs. United Nations, NY. UN, 2005. World Population Prospects. The 2004 Revision. Highlights. Population Division, Department of Economic and Social Affairs, United Nations, NY. Van den Brink, P.J., Sureshkumar, S.N., Daam, M.A., Domingues, I., Milwain, G.K., Beltman, W.H.J., Perera, M.W.P., Satapornvanit, K., 2003. Environmental and human risk of pesticide use in Thailand and Sri Lanka. Results of a preliminary risk assessment. Alterra - report 789. MAMAS Report Series No. 3/2003. Wageningen: Alterra. The Netherlands. Van Ewijk, P.H., Hoekstra, J.A., 1993. Calculation of the EC 50 and its confidence interval when subtoxic stimulus is present. Ecotoxicol. Environ. Safety 25, 25 32. 103

Van Gestel, C.A.M., Ma, W.C., 1988. Toxicity and bioaccumulation of chlorophenols in earthworms, in relation to bioavailability in soil. Ecotoxicol. Environ. Safety 15, 289-297. Van Gestel, C.A.M., van Dis, W.A., van Breemen, E.M., Sparenburg, P.M., 1989. Development of a standardized reproduction toxicity test with the earthworm species Eisenia fetida andrei using copper, pentachlorophenol, and 2, 4-dichloroaniline. Ecotoxicol. Environ. Safety 18, 305-312. Van Gestel, C.A.M., 1992. Validation of earthworm toxicity tests by comparison with field studies: A review of benomyl, carbendazim, carbofuran, and carbaryl. Ecotoxicol. Environ. Safety 23, 221-36. Van Gestel, C.A.M., Dirven-Van Breemen, E.M., Baerselman, R., 1992. Influence of environmental conditions on the growth and reproduction of the earthworm Eisenia andrei in an artificial soil substrate. Pedobiologia 36, 109 120. Van Gestel, C.A.M., 1997. Scientific basis for extrapolating results from soil ecotoxicity tests to field conditions and the use of bioassays. In: Van Straalen, N.M, Løkke, H. (Eds): Ecological risk assessment of chemicals in soil. Chapman and Hall, London, UK, pp 25 50 Van Gestel, C., Weeks, J., 2004. Recommendations of the 3rd International Workshop on Earthworm Ecotoxicology, Aarhus, Denmark, August 2001. Ecotoxicol. Environ. Safety 57, 100 105. Verma, A., Pillai, M.K.K., 1991. Bioavailability of soil-bound residues of DDT and HCH to earthworms. Curr. Sci. 61, 840-43. Vermeulen, L.A., Reinecke, A.J., Reinecke, S.A., 2001. Evaluation of the fungicide manganese-zinc ethylene bis(dithiocarbamate) (mancozeb) for sublethal and acute toxicity to Eisenia fetida (Oligochaeta). Ecotoxicol. Environ. Safety 48, 183 189. Viswanathan, P.N., Krishnamurti, C.R., 1989. Effects of temperature and humidity on ecotoxicology of chemicals. In: Bourdeau, P., Haines, J.A., Klein, W., Krishanamurti, C.R., (Eds): Ecotoxicology and Climate with Special Reference to Hot and Cold Climates, John Wiley and Sons. New York. Pp. 139-154. Wauchope, R.D., Butler, T.M., Hornsby, A.J., Augusteijn-Beckers, P.W.M., Burt, J.P., 1992. SCS/ARS/CES pesticide properties data base for environmental decision making. Rev. Environ. Contam.Toxicol.123, 1-157. WHO, 2004. The WHO recommended classification of pesticides by hazard and guidelines to classification 2004. World Health Organization. Geneva. Switzerland. Yeardley, R.B, Jr., Lazorchak, J.M., Gast, L.C., 1996. The potential of an earthworm avoidance test for evaluation of hazardous waste sites. Environ. Toxicol. Chem.15, 1532 1537. 104

References Zaller, J.G., 2007. Vermicompost as a substitute for peat in potting media: effects on germination, biomass allocation, yields and fruit quality of three tomato varieties. Sci. Horticult. 112, 191 199. Zhou, S.P., Duan, C.Q., Fu, H., Chen, Y.H., Wang, X.H., Yu, Z.F., 2007. Assessing chlorpyrifos-contaminated soil with three different earthworm test methods. J. Environ. Sci. 19, 854 858. 105

106

Summary Expansion of agriculture in tropical regions is heavily relying on the use of pesticides. Pesticides however, are generally also toxic to non-target soil organisms and as a consequence may hamper proper functioning of the soil. Sustainability of tropical agriculture therefore requires information on the effects of pesticides on beneficial soil organisms like earthworms, which play an important role in the soil ecosystem. Nevertheless pesticide toxicity under tropical conditions is only rarely assessed and understanding the effects in such conditions remains a priority. Tropical risk assessment often is relying on data generated under temperate conditions. This approach is rather questionable mainly due to different climatic conditions. Therefore it is essential to understand pesticide toxicity and the consequent ecological effects under tropical conditions. The main objective of this thesis was to asses the toxicity of some selected pesticides to earthworms under tropical conditions. The effects were investigated using basic and extended laboratory tests and linking them with field studies. In addition, it was focused on the development of new tools for tropical ecotoxicology. In Chapter 2, it was hypothesized that temperature and soil type may act as critical factors that modify pesticide toxicity. The effects of three commonly used pesticides, chlorpyrifos, carbofuran and carbendazim, on the survival, growth and reproduction of the standard test species Eisenia andrei were investigated in different soil types at representative temperate and tropical conditions. In case of survival, toxicity of chlorpyrifos and carbofuran in artificial soil was higher at 26 C than at 20 C. But in natural soils, carbofuran was most toxic at 20 C while chlorpyrifos remained more toxic at 26 C. The higher activity of earthworms at higher temperatures may trigger a higher uptake of pesticides causing larger effects on survival under representative tropical conditions. However, sub-lethal endpoints, such as growth and reproduction, did not show clear differences between the tested temperatures and varied with different soil types indicating that the nature of the pesticide and soil type were more important than temperature. Effects of carbendazim on survival, growth and reproduction of Eisenia andrei were lower at 26 C than at 20 C suggesting temperature may also influence the stability of the pesticide and its degradation kinetics, resulting in a lower toxicity at higher temperatures. In conclusion, Chapter 2 suggests that pesticide toxicity may differ with temperature and soil type; therefore temperate data should only be used with caution in tropical risk assessment. Tropical risk assessment currently uses data generated with the temperate compost worms Eisena fetida and Eisena andrei. These species are less ecologically relevant as they are rarely found in natural soils. One good alternative is to use indigenous species specific to tropical regions. 107

Therefore the toxicity of chlorpyrifos, carbofuran and mancozeb to the tropical earthworm species Perionyx excavatus was investigated in Chapter 3. In addition effects of both pure compounds and formulated products were compared to estimate the possible added toxicity of formulated products that is often neglected in toxicity studies. Carbofuran was most toxic to Perionyx excavatus under tropical conditions followed by chlorpyrifos and mancozeb. In case of survival, comparison with toxicity data generated for chlorpyrifos and carbofuran in Chapter 2 revealed that Perionyx excavatus was more sensitive than the standard species Eisenia andrei. However, for sub-lethal effects differences in sensitivity were relatively small. Generally, the toxicity of formulated products was higher than that of the pure compounds. Nevertheless, results suggest that the toxicity of formulated products may be masked by interactions with the soil resulting in more or less similar toxicities of formulations and the pure compounds. In addition, Chapter 3 shows that Perionyx excavatus may be used as a good alternative test species under tropical conditions in future studies, applying the available standard test guidelines for Eisenia sp. Lack of cost-effective tests with short duration often hampers the risk assessment of pesticides under tropical conditions. Recently the earthworm avoidance test has been considered a good alternative in toxicity studies. In Chapter 4, earthworm avoidance tests were performed with the standard temperate test species Eisenia andrei and the tropical earthworm species Perionyx excavatus, using a in two-chamber test system with artificial soil and a selected natural soil at a representative tropical temperature (26 C). The EC 50 values for the effects on avoidance behaviour for both chlorpyrifos and carbofuran indicated that Eisenia andrei was a factor of 2-3 more sensitive than Perionyx excavatus under tropical conditions. An important finding was the attraction of Perionyx excavatus at lower concentrations of both pesticides tested and in both soils; hence it was suggested to use avoidance tests with local species with extra caution as it might lead to wrong estimations of the effects. Comparison with literature data showed that endpoints generated through earthworm avoidance tests are generally less sensitive than reproduction but more sensitive than survival and therefore could be used in the initial risk assessment to facilitate pesticide regulation in tropical regions. The standard test guidelines for earthworm toxicity testing often use the OECD artificial soil, which is composed of sand, kaolin clay and sphagnum peat. Development of a tropical artificial soil is timely needed due to increasing costs, non-availability and environment concerns associated with sphagnum peat. The suitability of locally available, environmentally friendly substrates such as coco peat (composted and non-composted), saw dust and paddy husk was investigated in Chapter 5. Modified artificial soils were prepared by substituting sphagnum peat in the standard artificial soil. Survival, growth and reproduction of Eisenia andrei in the modified 108

Summary artificial soils were tested in the first phase with comparison to standard artificial soil. The results indicated the suitability of modified artificial soils prepared with composted coco peat and paddy husk. During the second phase toxicity of chlorpyrifos, carbofuran and carbendazim for Eisenia andrei was determined to validate the two selected substrates. For all three pesticides LC 50 values found in standard artificial soil and soils modified with composted coco peat and paddy husk did not differ significantly. EC 50 values for the effects on reproduction were similar in standard artificial soil and soil modified with composted coco peat but lower in soil modified with paddy husk. These results suggest that composted coco peat might be a good alternative for sphagnum peat and can be used in tropical ecotoxicology. Standardization of this new substrate however, warrants further studies. Effects of pesticides on structural properties (species diversity, abundance, biomass) and functions (organic matter breakdown, nutrient cycling) of soil ecosystems have rarely been studied in the field under tropical conditions. And predicting field effects based on data from laboratory studies is a challenge. Hence effects of chlorpyrifos on the diversity, abundance and biomass of earthworms and termites in relation to organic matter break down were studied in tropical soil (Chapter 6). To link laboratory results and realistic field conditions, doses relevant to normal agricultural practice, representing the laboratory EC 50 for earthworm reproduction and double the EC 50 were applied. Diversity, abundance and biomass of earthworms and termites were observed prior to application and 1, 2, 3, 6 and 12 months after pesticide application. In addition, using litter bags, organic matter breakdown was investigated until complete decomposition in the control plots was achieved. Chlorpyrifos caused adverse effects on decomposition even 2-3 months after application. Biomass and abundance of earthworms and termites were also reduced and the decrease could be linked with decreased litter decomposition. Nevertheless effects on termites and earthworms were diminished after 6 months and populations were completely recovered after 12 months. The faster litter decomposition rates under tropical conditions showed the importance of modifying the current standard guidelines on effects of pesticides in field conditions for use under tropical conditions. In conclusion, effects of pesticides on earthworms under tropical conditions may depend on temperature, soil type, type of pesticide, toxicological endpoints measured and their interactions. Future investigations should be done with more pesticides using different endpoints and different soil organisms to increase the general knowledge of pesticide effects under these conditions. Degradation and fate of pesticides under tropical conditions should be linked with observed effects to enable a more realistic risk assessment 109

110

Samenvatting De toename van de agrarische productie in de tropen is in grote mate afhankelijk van het gebruik van bestrijdingsmiddelen. Bestrijdingsmiddelen zijn over het algemeen ook giftig voor niet-doelwit organismen en kunnen daardoor een nadelig effect hebben op het functioneren van de bodem. Om te komen tot een duurzame tropische landbouw is daarom informatie nodig over de effecten van bestrijdingsmiddelen op nuttige organismen, zoals regenwormen, die een belangrijke rol spelen in het bodemecosysteem. Risico s van bestrijdingsmiddelen in de tropen worden echter nog zelden onderzocht. En indien een risicobeoordeling in de tropen wordt uitgevoerd, dan gaat deze vaak uit van gegevens die zijn verzameld onder gematigde omstandigheden. Het is onzeker of een dergelijke benadering ook toepasbaar is voor de tropen gelet op de grote verschillen in klimaatomstandigheden. Het is daarom van groot belang kennis te verzamelen over de giftigheid van bestrijdingsmiddelen en de daaruit voortvloeiende mogelijke ecologische effecten onder tropische omstandigheden. Het doel van het onderzoek dat in dit proefschrift wordt beschreven was het bepalen van de giftigheid van enkele geselecteerde bestrijdingsmiddelen voor regenwormen onder tropische omstandigheden. De effecten werden onderzocht door middel van gestandaardiseerde en aangepaste experimenten in het laboratorium en een veldstudie. Daarnaast was het onderzoek gericht op de ontwikkeling van nieuwe methoden voor ecotoxicologisch onderzoek in de tropen. In Hoofdstuk 2 werd de hypothese onderzocht dat temperatuur en bodemtype de kritische factoren zijn die de giftigheid van een bestrijdingsmiddel bepalen. Het effect van drie veel gebruikte bestrijdingsmiddelen, chloorpyrifos, carbofuran and carbendazim, op de overleving, groei en reproductie van de standaard testsoort Eisenia andrei werd onderzocht in verschillende gronden bij temperaturen die representatief zijn voor gematigde en tropische omstandigheden. Voor overleving was de giftigheid van chloorpyrifos and carbofuran in een standaard kunstgrond hoger bij 26 C dan bij 20 C. In natuurlijke gronden was carbofuran echter giftiger bij 20 C, terwijl chloorpyrifos weer het meest giftig was bij 26 C. De hogere activiteit van regenwormen bij een hogere temperatuur zou kunnen leiden tot een verhoogde opname van bestrijdingsmiddelen en daarmee de hogere gevoeligheid bij tropische temperaturen kunnen verklaren. Subletale effecten op groei en reproductie vertoonden echter geen duidelijk verschil tussen de geteste temperaturen maar verschilden tussen de geteste gronden. Dit suggereert dat het type bestrijdingsmiddel en het bodemtype belangrijker zijn voor de gevoeligheid van subletale parameters dan de temperatuur. Carbendazim was minder giftig voor zowel overleving als groei en reproductie bij 26 C dan bij 20 C. 111

Dit suggereert dat de temperatuur invloed heeft op de stabiliteit van dit middel, met een lagere toxiciteit als gevolg van een snellere afbraak bij een hogere temperatuur. De resultaten verkregen in Hoofdstuk 2 laten zien dat de giftigheid van bestrijdingsmiddelen voor regenwormen kan verschillen afhankelijk van temperatuur en bodemtype. Gegevens verkregen onder gematigde omstandigheden zijn daarom niet direct toepasbaar voor de risicobeoordeling in de tropen. De risicobeoordeling van stoffen in de tropen gaat vaak uit van gegevens verkregen voor de compost- of mestwormen Eisena fetida en Eisena andrei. Deze soorten, die vooral in gematigde zones voorkomen, worden niet of nauwelijks gevonden in natuurlijke gronden en zijn daarom minder ecologisch relevant. Een goed alternatief zou een soort kunnen zijn die wel van nature voorkomt in tropische bodems. Om deze reden is in Hoofdstuk 3 de giftigheid van chloorpyrifos, carbofuran and mancozeb bepaald voor de tropische regenworm Perionyx excavatus. Bestrijdingsmiddelen worden op de markt gebracht in de vorm van geformuleerde producten; de giftigheid van deze formuleringen krijgt echter nog weinig aandacht. In Hoofdstuk 3 is ook de giftigheid van deze drie zuivere actieve stoffen vergeleken met die van geformuleerde producten. Carbofuran was het meest giftig voor Perionyx excavatus onder tropische omstandigheden, gevolgd door chlorpyrifos en mancozeb. Wanneer effecten op de overleving worden vergeleken met de toxiciteitsgegevens die in Hoofdstuk 2 zijn beschreven, blijkt Perionyx excavatus gevoeliger te zijn voor chloorpyrifos en carbofuran dan de standaard testsoort Eisenia andrei. Het verschil in gevoeligheid is echter klein wanneer gekeken wordt naar subletale effecten op groei en reproductie. De formuleringen waren over het algemeen iets giftiger dan de zuivere stoffen, maar het verschil was niet groot. Dit suggereert dat de giftigheid van de formuleringen gemaskeerd kan worden door interacties met de bodem. Hoofdstuk 3 laat ook zien dat Perionyx excavatus een goed alternatief is voor toxiciteitstoetsen met regenwormen in de tropen, en dat daarbij de beschikbare gestandaardiseerde toetsrichtlijnen voor Eisenia sp. een goed uitgangspunt vormen. De risicobeoordeling van bestrijdingsmiddelen in de tropen wordt belemmerd door een gebrek aan snelle en goedkope testmethoden. Recent is voor regenwormen een toets op ontwijkingsgedrag ontwikkeld, die een alternatief zou kunnen zijn voor de langer durende standaard toxiciteitstoetsen. In Hoofdstuk 4 is het ontwijkingsgedrag van regenwormen onderzocht met de standaardsoort Eisenia andrei en de tropische soort Perionyx excavatus. Hiervoor is een tweekamer systeem gebruikt en de testen werden uitgevoerd met de standaard kunstgrond en een natuurlijke grond bij een representatieve tropische temperatuur van 26 C. De EC 50 waarden voor het effect van chloorpyrifos en carbofuran op het ontwijkingsgedrag lieten zien dat Eisenia andrei een factor 2-3 gevoeliger was dan Perionyx excavatus. Opmerkelijk was dat Perionyx excavatus 112

Samenvatting aangetrokken leek te worden door lage concentraties van beide bestrijdingsmiddelen in beide gronden. Dit laat zien dat de resultaten van ontwijkingstoetsen met locale regenwormsoorten tot verkeerde conclusies over de risico s van een bestrijdingsmiddel kunnen leiden. Vergelijking met gegevens uit de literatuur liet zien dat het ontwijkingsgedrag van regenwormen over het algemeen minder gevoelig is dan reproductie maar meer gevoelig dan overleving. De toets op ontwijkingsgedrag zou daarom een eerste stap kunnen zijn in de risicobeoordeling van bestrijdingsmiddelen in de tropen. In toxiciteitstoetsen met regenwormen volgens internationaal gestandaardiseerde richtlijnen wordt gebruik gemaakt van een kunstgrond, die wordt samengesteld uit zand, kaolien (pottenbakkersklei) en veenmos (Sphagnum) turf. Toenemende kosten, beperkte beschikbaarheid en milieuoverwegingen ten aanzien van het gebruik van veenmosturf vragen om de ontwikkeling van een kunstgrond die toepasbaar is in de tropen. Daarom is in Hoofdstuk 5 de geschiktheid van enkele lokaal beschikbare en milieuvriendelijke alternatieven onderzocht. Het betrof hier gecomposteerde of niet-gecomposteerde kokosturf, zaagsel en het kaf van rijst (paddy husk). Kunstgronden werden bereid met deze materialen als vervanging van veenmosturf. Als een eerste stap werden overleving, groei en reproductie van Eisenia andrei bepaald in de aangepaste gronden en in de standaard kunstgrond. De gronden bereid met gecomposteerde kokosturf en paddy husk bleken het meest geschikt te zijn. Deze gronden werden vervolgens gebruikt om de giftigheid van chloorpyrifos, carbofuran en carbendazim voor Eisenia andrei te bepalen. LC 50 waarden voor het effect van deze drie bestrijdingsmiddelen op de overleving van de regenwormen verschilden niet voor de aangepaste gronden en de standaard kunstgrond. EC 50 waarden voor het effect op de reproductie waren vrijwel gelijk in de standaard kunstgrond en de grond bereid met gecomposteerde kokosturf, maar lager in de grond met paddy husk. Dit suggereert dat kokosturf het meest geschikt is als alternatief voor veenmosturf in tropische ecotoxicologische toetsen. De mogelijkheden tot standaardisatie van dit materiaal moeten echter nog verder onderzocht worden. De effecten van bestrijdingsmiddelen op de structuur (soortdiversiteit, aantallen, biomassa) en het functioneren (strooiselafbraak, de omzetting van nutriënten) van bodemecosystemen zijn nog nauwelijks onderzocht in de tropen. Het voorspellen van effecten van stoffen op deze parameters op basis van laboratoriumgegevens is een van de belangrijkste uitdagingen in de ecotoxicologie. Om die reden is in Hoofdstuk 6 het effect van chloorpyrifos bepaald op de diversiteit, de aantallen en de biomassa van regenwormen en termieten en de afbraak van organisch materiaal in de bodem onder tropische omstandigheden. Om de resultaten te kunnen vergelijken met die van laboratoriumtesten werden doseringen gekozen die representatief zijn voor de praktijktoepassing van chloorpyrifos en 113

overeenkomend met de EC 50 voor effecten op de reproductie van regenwormen en twee maal de EC 50. De diversiteit, aantallen en biomassa van regenwormen en termieten werden bepaald voor en 1, 2, 3, 6 en 12 maanden na de toepassing van chloorpyrifos. Daarnaast werden strooiselzakjes (litter bags) ingegraven, waarin de afbraak van organisch materiaal werd gevolgd tot het volledig was afgebroken in de onbehandelde controleveldjes. Chloorpyrifos had een nadelig effect op de strooiselafbraak tot 2-3 maanden na toepassing. De biomassa en de aantallen regenwormen en termieten waren ook afgenomen, wat de verminderde strooiselafbraak kan verklaren. Zes maanden na toepassing waren de effecten op regenwormen en termieten al sterk verminderd, de populaties waren volledig hersteld na 1 jaar. De snelle strooiselafbraak die in dit experiment werd waargenomen laat zien dat de huidige richtlijn voor het uitvoeren van litterbag studies naar het effect van bestrijdingsmiddelen aangepast moet worden bij toepassing onder tropische omstandigheden. Concluderend kan gesteld worden dat het effect van bestrijdingsmiddelen op regenwormen onder tropische omstandigheden kan afhangen van temperatuur, bodemtype, type bestrijdingsmiddel, toxicologisch eindpunt en de interactie tussen deze factoren. Nader onderzoek met meer verschillende bestrijdingsmiddelen, focus op verschillende eindpunten en met meerdere soorten testorganismen is nodig om de kennis van bestrijdingsmiddelen onder tropische omstandigheden te vergroten. Naast inzicht in effecten moet ook kennis over de afbraak en het gedrag van bestrijdingsmiddelen onder tropische omstandigheden worden meegenomen om te komen tot een meer realistische risicobeoordeling. 114

Acknowledgements Finally, time has come to write the acknowledgement which indeed gives me mixed emotions. Certainly, I am very delighted that I am reaching a significant milestone in my career. On the other hand, it also reminds me that the stay in the Department of Animal Ecology will end soon. It has been a long journey and a big challenge for me, especially with the nature of this PhD project, working two years in Amsterdam, the Netherlands and two years in Matara, Sri Lanka. This is actually a product that comes in expenses of great help and support given by various people during the last four years. First of all, Kees, I would like to express my deepest thanks for all your guidance, support, patience and encouragement throughout my work. It was only some years back when I started contacting you via email and you were willing to accept me as your student to work in ecotoxiclogy, a field which I like very much. From theory to practical work, this field was completely new to me and it was the greatest challenge that I have ever had. I would like to emphasize that it was your support that helped me to fulfill the target. I was able to finish this study within the time frame we planned and most importantly with good results. The most interesting thing, which you may not realize, is that you have converted me from a fisheries guy to an earthworm guy. Sometimes I felt that I disturbed you too much on different things; nevertheless it was due to my will to do things, which I was not sure about in a correct way. That gave me more success than failures. I hope you understand. Nico, I am grateful to you, for the opportunity and hospitality to work in the Department of Animal Ecology at VU. I must say that I was very happy to stay in such a professional and warm-hearted environment. Thanks for reading all my chapters and thesis on various occasions. Your suggestions and comments greatly helped me to improve my work. I am also indebted to Prof. Pathiratne, initially accepting my request to work as a co-promoter in this project. Your guidance throughout these years helped me in numerous ways. It is pity that I could not do some planned work with you, but I believe we can do more work in the near future and I am looking forward to seeing you in Amsterdam soon. Many thanks to Jörg, Paul, Els, Joost and Herman for accepting to be in my reading committee and for sending your valuable suggestions, comments in a short period even though many of you were enjoying your summer holidays. I really appreciate your critical thoughts on the chapters and they were really helpful in the final context. I hope that I will see you all on my promotion day here in Amsterdam. My gratitude also goes to Prof. (Mrs) N. J. De. S. Amarasinghe at the Department of Zoology, University of Ruhuna, Matara, Sri Lanka for always being supporting and encouraging me in my career. You were my 115

first supervisor at the Department of Zoology, University of Ruhuna exactly 10 years ago, since then you were behind me in all my success. I know that you would be very happy to see this important occasion and you would wish me all the best from your heart. I will definitely miss you here in Amsterdam. I also thank Dr. (Mrs) M. G. V. Wickramasinghe (Head/Zoology) and Dr (Mrs) H. Wegiriya, former head, Department of Zoology for supporting me during my field work in Sri Lanka. I wish to thank Dr Guruge by helping me on numerous occasions during my time in Amsterdam and also to all my teachers at Ruhuna for their encouragement through out the years. It is my pleasure to thank Dr. L. A. Samayawardhena for your guidance especially in the beginning of my career at Ruhuna. It was your enthusiasm towards ecotoxicology inspired me to select this field of research and I learnt a lot from you, when we were working together in Sri Lanka. Very special thanks to Dr J. M. Julka for creating my interest in taxonomy of earthworms and your hospitality at Solan, India, which I liked quite a lot. I really enjoyed my time there and I am looking forward to working with you in the coming years (we have to find more new species...). Hector. without you.. I do not know how I could finish all this work. I thought that maintaining earthworm cultures at the Department of Zoology, Ruhuna would really be a difficult task. But you made life much easier for me by taking very good care of the cultures and doing all the hard work. It also reminds me all the fun we had while working, which made us feel the work is not that hard. Thank you very much! I also appreciate the help given by Priyantha, Rayston, Rohana, Anil, Sanjeewa, Thilaka, Sujeewa and Dharmasiri on several occasions especially during the field work at Walasgala. I could have needed more people to prepare the litter bags but you all had tried your best to help me to finish the work as planned. I was fortunate to have so many good friends at the Department of Animal Ecology although the time I spent here was relatively short. I would like to thank staff members Jacintha, Dick, Gerard, Joris and Matty for their help and enthusiasm during my stay at VU. Thanks for my roommates, Frans and Rully (in 2005) and Tjalf and Coby (in 2009) for integrating me into Dutch environment. I am also indebted to my friends Mieke, Miriam, Maria, Daniel, Maartje and all the others at the Department of Animal Ecology, for their help and also for the nice time working together. Thanks to Eric for the field visits during my first year and I can still well remember how I learnt to dig worms from you. In 2005, it was all about new things, new labs and new techniques..i was a bit nervous but Rudo you helped me a lot in my experiments and 116

Acknowledgements I really appreciate it. Also, many thanks to Désirée, for helping me in all official matters without any hesitation. My deepest gratitude goes to Amma and Thatta for caring and guiding me throughout these years. Whatever I do, wherever I go, I know I will be your beloved son forever. I am so sad that you are not here to see one of your wishes, which both of you were dreaming for years, comes true. I will miss you very much here.also malli and nangi I will miss you too. Finally, but definitely not at last, Huong I can not imagine that how could I finish this without your support and commitment. I know you were very happy to see my success, you are and you will be. In fact, it is all your sacrifices mostly helped me to finish this work and I love you so much.. My lovely little son, Thisara, of course I can not forget to write about you. Although you never allowed me to study or even for rest at home, it was you who inspired me with your cute smile and sweet laughter. You really gave me energy to work. I love you with all my heart. On a long journey of human life, faith is the best of companions; it is the best refreshment on the journey; and it is the greatest property. Lord Buddha Thank you all for your faith that kept me working! 117