Endosulfan in freshwater and marine water
Toxicant default guideline values for protecting aquatic ecosystems
Extracted from Section 8.3.7 ‘Detailed descriptions of chemicals’ of the ANZECC & ARMCANZ (2000) guidelines.
The default guideline values (previously known as ‘trigger values’) and associated information in this technical brief should be used in accordance with the detailed guidance provided in the Australian and New Zealand Guidelines for Fresh and Marine Water Quality.
Description of chemical
Most organochlorine pesticides have been phased out of use in recent years, mainly because of their residual properties and potential for bioaccumulation. The guideline trigger values stated are for toxicity only and need to be adjusted for bioaccumulation where appropriate. Where the statistical distribution method was used, figures quoted are the 95% protection levels, usually applicable to slightly to moderately disturbed systems although 99% protection figures are recommended for chemicals that bioaccumulate.
Endosulfan (CAS 115-29-7) is a broad-spectrum insecticide of the cyclodiene group of chlorinated hydrocarbons but it does not share the bioaccumulating properties of other cyclodienes such as dieldrin or chlordane (Maier-Bode 1968). Technical grade endosulfan comprises two stereoisomers, (α) alpha-endosulfan and (β) beta-endosulfan, present in the ratio around 64 to 67% α and 29 to 32% β. Each has slightly different physico-chemical properties, fate and transport in the environment but have similarly high toxicities, as has the major biotransformation product, endosulfan sulfate.
Its IUPAC name is 6,7,8,9,10,10-hexachloro-1,5,5a,6,9,9a-hexahydro-6,9-methano-2,4,3-benzodioxathiepine3-oxide, formula is C9H6Cl6O3S and molecular weight 406.9. The solubility of the α- and β- isomers is 0.32-0.33 mg/L at 22°C and log Kow, 4.74-4.79, at pH5 (Tomlin 1994).
Endosulfan has an important place in the pest management strategy of various crops in Australia, particularly cotton in inland northern NSW and Queensland, where up to 400 tonnes are applied between October and February (Sunderam et al. 1992).
Endosulfan is also used as insecticide and acaricide on a variety of other crops, including vegetables, fruit, nuts, cereal as well as in plant nurseries, on lawn, pasture and fodder, flowers and ornamentals (NRA 1997a). It has over 1300 registered uses in Australia (NRA 1997a). The current analytical practical quantitation limit (PQL) for DDT is 0.05 µg/L (NSW EPA 2000), although specialised methods may detect lower concentrations.
Endosulfan is degraded in soil to the sulfate and to less harmful diols and ethers (Ghadiri et al. 1994). Its DT50 in soil is 30 to 70 days (Tomlin 1994) but is extended to 5 to 8 months if the equally toxic sulfate is included in total endosulfan residues (Tomlin 1994). (Ghadiri et al. 1994) observed that it persists for about 30 weeks in soil and even longer in sediment. The half-life of α-endosulfan in natural river waters is around 2 days while the β-isomer is slightly more persistent, 4 to 7 days (Sunderam 1990). The soluble forms of endosulfan disappear from water bodies in a few weeks (Peterson & Batley 1991).
Endosulfan can persist in fish tissue longer than in water but is readily excreted, and in wild fish in the NSW cotton areas it did not accumulate from one season to the next (Nowak & Julli 1991).
Endosulfan has very high toxicity to fish (Sunderam et al. 1992) and has been suspected of causing fish kills in NSW (Bowmer et al. 1995). It is also highly toxic to some invertebrate species.
Freshwater fish: 42 spp, 96-hour LC50 (48 only available for three species), 0.1 to 63.0 µg/L. Only for six species were figures > 20 µg/L, although outlying figures were also found for Catla catla (371 and 424 µg/L; 48 hours), AQUIRE (1994) and Tilapia mossambicca (140 and 150 µg/L).
There were few chronic fish data for endosulfan. Macek et al. (1976) reported that chronic toxicity to the fathead minnow Pimephales promelas occurred at 0.28 µg/L and the acute-to-chronic ratio was 3. Barry et al. (1995a) reported a 6-day no observed effect concentration (NOEC) figure for Melanotaenia fluviatilis of 6.8 µg/L of mortality of eggs from hatch but hatching NOEC was as high as 224 µg/L.
Freshwater amphibians: 48 to 96-hour LC50 to Rana tigrina of 1.8 to 2.0 µg/L.
Freshwater crustaceans: 48 to 96-hour LC50 values varied with taxa (20 species); most sensitive were copepods (two species; 0.1 to 0.6 µg/L) and ostracods (one species, 0.9 µg/L). Two species of water fleas were very sensitive (0.2 to 0.3 µg/L) although four other similar species had LC50 values between 56 and 720 µg/L. Shrimp and Gammarus were quite sensitive (six species; 2.0 to 7.8 µg/L) followed by crayfish (one species, 24 to 423 µg/L). Crabs were generally least sensitive (three species; 360 to 17,780, above the water solubility) although Paratelphusa jacquemontii was sensitive to 0.16 µg/L.
Sunderam et al. (1994) found a nominal NOEC (14-day, reproduction-impairment) for the Australian cladoceran, Ceriodaphnia dubia, of 10 µg/L and for Moinodaphnia macleayi of 20 µg/L. Patra et al. (1996) reported a measured NOEC figure for C. dubia of 1.0 µg/L at 25°C but this awaits peer review. A 64-day NOEC (mortality) of 2.7 µg/L for D. magna has been reported.
Freshwater insects: 48 to 96-hour LC50, four species, 0.1 to 17.5 µg/L.
Freshwater molluscs: 96-hour LC50 for two species was 6-44 µg/L but for Lymnaea stagnalis the 48-hour LC50 was 7370 µg/L and for Melanopsis dufori, the 96-hour LC50 was 37,330 to 42,760 µg/L. Figures for those species were above the water solubility.
Freshwater algae and ciliates: one algal species, Chlorella vulgaris, 14-day NOEC (growth) of 700 µg/L; one ciliate species, Paramecium aurelia, 5-day NOEC (growth) of 100 µg/L.
Marine fish: 11 species, 48 to 96-hour LC50, 0.1 to 23.3 µg/L. The only figures above 4 µg/L were for Mugil cephalus, and figures around 0.4 µg/L were also reported for this species.
Marine crustaceans: 11 species; 48 to 96-hour LC50 for prawns, shrimp and mysids (seven species) were between 0.03 and 5 µg/L, although Penaeus monodon (additional species) was less sensitive (4.6 to 37.3 µg/L). Crabs were less sensitive (four species, 15 to 178 µg/L).
Marine molluscs: six species, 48 to 96-hour LC50, 2.0 to 65 µg/L.
Marine annelids: two species, 96-hour LC50, 161 to 1135 µg/L.
Marine echinoderm: one species, Strongylocentrotus purpurtus, 5-day LC50, 230 µg/L.
Marine algae: one species of red algae, Champia parvula, 14-day NOEC (reproduction) of 80 µg/L.
1) α-endosulfan – toxicity
Freshwater fish: three species, 96-hour LC50, 0.16 to 1.3 µg/L.
Freshwater crustacean: one species, 48-hour LC50, 249 µg/L.
2) β-endosulfan – toxicity
Freshwater fish: three species, 96-hour LC50, 6.6 to 8.8 µg/L.
Freshwater crustacean: one species, 48-hour LC50, 205 µg/L.
Australian and New Zealand data
A number of studies have been undertaken on endosulfan, associated with its use on cotton in eastern Australia (Chapman et al. 1993). Sunderam et al. (1992) and others have reported 96-h LC50 figures for native fish (measured, where available): silver perch Bidyanus bidyanus (2.3 to 5.7 µg/L); firetail gudgeon Hypseleotris gallii (2.2 µg/L); golden perch Macquaria ambigua (0.3 to 1.4 µg/L); rainbowfish Melanotaenia duboulayi (0.5 to 5.9 µg/L); rainbowfish Melanotaenia fluviatilis (5.7 µg/L); and bony bream Nematolosa erebi (0.2 to 1.3 µg/L). Figures obtained for introduced fish were: Cyprinus carpio (0.1 to 0.6 µg/L); Gambusia holbrooki (2.3 to 3.8 µg/L); Oncorhynchus mykiss (0.7 to 1.6 µg/L); Rasbora sp (0.2 µg/L). The chronic rainbowfish hatching data reported by Barry et al. (1995a) is on the Australian M. fluviatilis.
The following invertebrate data have been reported: shrimp Caridinides sp. (48-hour EC50, 2.0 to 7.8 µg/L); Paratya australiensis (96-hour EC50, 6.1 to 11.7 µg/L); waterfleas Ceriodaphnia dubia (48-hour EC50, 151 to 491 µg/L and 14-day NOEC for reproduction of 10 µg/L); Moinodaphnia macleayi (48-hour EC50, 215 µg/L and 14-day NOEC of 20 µg/L) (Sunderam 1990, Sunderam et al. 1994). More recently, Patra et al. (1996) have reported measured figures for waterfleas and effects of temperature described below (to be peer reviewed). 48-hour EC50 figures for the water boatman Notonecta sp varied from 0.1 to 5.0 µg/L (Sunderam 1990).
Other field data are reported below.
Factors that affect toxicity
pH: There are no data to indicate that the toxicity of endosulfan is affected by pH but its breakdown in water is accelerated in slightly alkaline waters (Eichelberger & Lichtenberg 1971).
Hardness: There are no data to indicate a change in toxicity with hardness.
Salinity: The marine guideline value is lower than the freshwater value but this may be an artefact of the species tested. No other salinity data are available.
Temperature: Endosulfan is commonly used in inland Australia during summer when water temperatures can be high (this also accelerates breakdown). Patra et al. (1995a) found almost a two-fold increase in the 24-hour LC50 to the silver perch, Bidyanus bidyanus, over the range of 15 to 35°C but no change in 96-hour LC50. LT50 values for three species of Australian fish exposed to 1.5 µg/L endosulfan decreased at 35°C.
Critical thermal maximum temperatures for three Australian fish species (Patra et al. 1995b) decreased significantly for fish exposed to sublethal levels (0.3 to 1.0 µg/L) of endosulfan.
Elevated temperatures caused a more significant increase in toxicity of endosulfan to cladocerans (Patra et al. 1996). The acute measured 48-hour EC50 for the Australian Ceriodaphnia dubia decreased from 166 µg/L at 15°C to 2.4 µg/L at 30°C. Reproductive impairment NOEC values decreased from 3.3 µg/L at 20°C to 1.0 µg/L at 25°C and 0.1 µg/L at 30°C.
Suspended matter: Endosulfan is known to adsorb to particulate matter, which is particularly significant in the turbid rivers and billabongs in the cotton growing areas. Most of the studies have focussed on the effects of suspended material on acute toxicity of endosulfan and not on the effects at low concentrations of endosulfan. Sunderam et al. (1992) found that the acute toxicities of endosulfan to native species in turbid water from the Mehi River were not significantly different from those of the animals tested in the filtered Sydney main water.
Leigh et al. (1997) assessed the effects of different concentrations of suspended sediment, modelled on the particulate size characteristics of Namoi River water, on the toxicity of endosulfan to the eastern rainbowfish, Melanotaenia duboulayi, under static conditions over 24 hours. A sediment concentration 1.4 µg/L, maintained in suspension, did not ameliorate the effect of endosulfan at acutely toxic concentrations (10 µg/L). However, higher sediment concentrations up to 52 µg/L, resulting in the precipitation of particles to form both suspended and bottom sediment, achieved a significant concentration dependent reduction in the toxicity of endosulfan. Maximum reduction in toxicity occurred when the aqueous sediment mixture had been pre-treated with endosulfan (10 µg/L) for 8 hours. Fish mortality was reduced by approximately 75%, however, further increases in sediment concentration did not cause any further amelioration in the endosulfan toxicity. It is necessary to undertake further investigations to see if endosulfan at low concentrations (e.g. around 0.01 µg/L) is significantly bound up by suspended sediment at turbidity levels typical of the cotton growing areas.
Environmental effects in the field: Several Australian studies were unable to either detect detrimental effects of endosulfan in the field or to link observed effects with endosulfan. Napier (1992) studied fish and invertebrates in four lagoon ecosystems with differing exposure patterns to endosulfan. High residues in fish were found in some lagoons at the peak of the spraying season, reflecting the exposure history and a fish kill occurred in one when endosulfan concentrations of 0.15 µg/L were measured three days later (Chapman et al. 1993). The overall patterns of aquatic communities varied markedly from one lagoon to another, complicated by water extraction and other factors, and the seasonal changes could not be clearly related to endosulfan. Small native fish were found in one lagoon in which endosulfan was measured up to 0.22 µg/L (Napier 1992).
The NSW Department of Land and Water Conservation have been undertaking chemical and biological monitoring on a broad geographical scale in the north-western rivers of NSW for a number of years. Royal and Brooks (1995) found that, in the 1994–95 irrigation season, there was some evidence of suspended recolonisation rates of one macroinvertebrate type in April at sites within irrigated areas, but could not establish a link with endosulfan or any other pesticide. A more recent survey (Brooks & Cole 1996) confirmed that ephemeropteran populations were adversely affected but was still unable to establish a clear link.
Leonard et al. (1995) found reductions in the populations of five common benthic macroinvertebrate taxa (mayflies and caddisflies) in the Namoi River in 1995–96 and, with the aid of solvent-filled dialysis bags was able to demonstrate a relationship with endosulfan both in solvent and in sediment. They indicated that it was entering the river through surface runoff during storm events. It was not possible to identify the average or peak concentrations in water over the sampling period.
A freshwater high reliability guideline figure of 0.2 µg/L was calculated for endosulfan using the statistical distribution method with 95% protection. The 99% protection level was 0.03 µg/L, and this is recommended as a trigger value for slightly to moderately disturbed ecosystems. Users should note that the 95% level fails to protect some important Australian species from acute toxicity. Endosulfan has been studied extensively in the Namoi and Gwydir rivers in NSW (Schofield 1998). Recent data (Brooks 1998, Chapman 1998, Hyne et al. 1998) indicate that effects on invertebrate populations may occur at concentrations not far above 0.03 µg/L. Hence, application of the 99% protection figure of 0.03 µg/L is strongly recommended and any relaxation from 0.03 µg/L should be assessed carefully.
A marine moderate reliability trigger value of 0.01 µg/L was calculated for endosulfan using the statistical distribution with 99% protection and an acute-to-chronic ratio (ACR) of 7.3. The 99% protection figure is 0.005 µg/L and is recommended for slightly to moderately disturbed systems.
Endosulfan has some potential to bioaccumulate and also for this reason users are advised to apply the 99% protection level if there are no data to adjust for bioaccumulation at the specific site (Section 184.108.40.206 of the ANZECC & ARMCANZ 2000 guidelines).
Guidelines for α- and β-endosulfan
Freshwater low reliability trigger values of 0.0002 µg/L (0.2 ng/L) were derived for alpha (based on the lowest LC50 of 0.16 µg/L) and 0.007 µg/L for beta (based on the lowest LC50 of 6.6 µg/L) using an assessment factor (AF) of 1000. These suggest a difference in toxicity of the two and more work is required. No marine data were available and the freshwater figures could be adopted. These figures should only be used as indicative interim working levels. In general, it is preferable to use the total endosulfan trigger value.
ANZECC & ARMCANZ 2000. Australian and New Zealand Guidelines for Fresh and Marine Water Quality, Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand, Canberra.
AQUIRE (Aquatic Toxicity Information Retrieval Database) 1994. AQUIRE standard operating procedures. USEPA, Washington, DC.
Barry MJ, Logan DC, Ahokas JT & Holdway DA 1995a. Sublethal effects of endosulfan and oxygen concentration on embryos of the Australian crimson-spotted rainbowfish (Melanotaenia fluviatilis). Australasian Journal of Ecotoxicology 1, 71-76.
Bowmer KH, Fairweather PG, Napier GM & Scott AC 1995. Review of data on the biological impact of cotton pesticides. CSIRO Division of Water Resources, Consultancy Report 95/13.
Brooks A & Cole R 1996. Central and north west regions water quality program: Biological monitoring, 1995/96 Report on biological monitoring. Department of Land and Water Conservation, Parramatta.
Brooks A 1998. Biological monitoring: Central and north-west regions water quality program. In Minimising the impact of pesticides on the riverine environment: Key findings from research within the cotton industry. Conference Proceedings, 22–22 July 1998. Land & Water Resources Research & Development Corp (LWRRDC), Cotton Research & Development Corp, Murray-Darling Basin Commission. LWRRDC Occasional Paper 23/98, 73–78.
Chapman J 1998. Laboratory ecotoxicology studies and implications for key pesticides. In Minimising the impact of pesticides on the riverine environment: Key findings from research within the cotton industry. Conference Proceedings, 22–22 July 1998. Land & Water Resources Research & Development Corp (LWRRDC), Cotton Research & Development Corp, Murray-Darling Basin Commission. LWRRDC Occasional Paper 23/98, 62–67.
Chapman JC, Napier GM, Sunderam RIM & Wilson SP 1993. The contribution of ecotoxicological research to environmental research. Australian Biologist 6, 72-81.
Eichelberger JW & Lichtenberg JJ 1971. Persistence of pesticides in river water. Environmental Science and Technology 5, 541-544.
Ghadiri H, Rose CW & Connell D 1994. Characteristics and field-to-stream transport of pesticides and other agricultural chemicals. Final report: Land and Water Resources Research and Development Corporation, R&D Project GRU2, Canberra.
Hyne RV, Lim RP & Leonard AW 1998. Relationship between endosulfan concentrations and maininvertebrate densities in the Namoi River over two cotton growing seasons. In Minimising the impact of pesticides on the riverine environment: Key findings from research within the cotton industry. Conference Proceedings, 22–22 July 1998. Land & Water Resources Research & Development Corp (LWRRDC), Cotton Research & Development Corp, Murray-Darling Basin Commission. LWRRDC Occasional Paper 23/98, 68–72.
Leigh K, Hyne RV & Lim RP 1997. Effect of sediment on the toxicity of endosulfan to eastern rainbowfish (Melanotaenia duboulayi). Abstract 05.4. In Proceedings 4th Annual Conference of the Australasian Society for Ecotoxicology, Brisbane, July 1997.
Leonard AW, Hyne RV, Lim RP & Chapman JC 1995. Relationships between abundance of selected macroinvertebrates and bottom sediment endosulfan levels in the Namoi River (NSW). In 34th Australian Society for Limnology Congress, Program and abstracts, Katoomba, September 1995, p44.
Macek KJ, Lindberg MA, Scott S, Bauston KS & Costa PA 1976. Toxicity of four pesticides to water fleas and fathead minnows: Acute and chronic toxicity of acrolein, heptachlor, endosulfan and trifluralin to the water flea (Daphnia magna) and the fathead minnow (Pimephales promelas). US EPA Duluth, Minnesota, EPA-600/3-76-099.
Maier-Bode H 1968. Properties, effect, residues and analytics of the insecticide endosulfan. Residue Reviews 22, 1-44.
Napier GM 1992. Application of laboratory-derived data to natural aquatic ecosystems. PhD Thesis, Macquarie University and Centre for Ecotoxicology, Sydney.
Nowak B & Julli M 1991. Residues of endosulfan in wild fish from cotton growing areas in NSW, Australia. Toxicological and Environmental Chemistry 33, 151-167.
NRA 1997a. Database extraction of selected pesticides: Registered uses in Australia, National Registration Authority, July 1997, Canberra.
NSW EPA 2000. Analytical Chemistry Section, Table of Trigger Values 20 March 2000, LD33/11, Lidcombe, NSW.
Patra R, Chapman JC, Gehrke PC & Lim RP 1995b. Effects of sub-lethal concentrations of endosulfan on the critical thermal maxima of freshwater fish. Poster abstract PW113. In Proceedings Second SETAC World Congress (16th Annual Meeting) Vancouver BC, 5-9 November 1995.
Patra R, Chapman JC, Lim RP & Gehrke PC 1995a. Effects of temperature on the acute toxicity of endosulfan to silver perch, Bidyanus bidyanus. Poster abstract PW 129. In Proceedings Second SETAC World Congress (16th Annual Meeting), Vancouver BC, 5-9 November 1995, 252.
Patra R, Chapman JC, Lim R & Gehrke P 1996. Effects of temperature on the toxicity of endosulfan, chlorpyrifos and phenol to Australian Ceriodaphnia dubia. Poster abstract 0112. In Proceedings InterSect’ 96, International Symposium on Environmental Chemistry and Toxicology, Sydney, 14-18 July 1996.
Peterson SM & Batley GE 1991. Fate and transport of endosulfan and diuron in aquatic ecosystems. Final Report AWRAC Project 88/20, CSIRO, Centre for Advanced Analytical Chemistry, Menai, NSW.
Royal M & Brooks A 1995. Central and northern west regions water quality program: Biological monitoring. Ecological Services Unit, NSW Department of Land and Water Conservation, Parramatta, TS95.179.
Schofield NJ 1998. Origins and design of the cotton pesticides program. In Minimising the impact of pesticides on the riverine environment: Key findings from research within the cotton industry. Conference Proceedings, 22–22 July 1998. Land & Water Resources Research & Development Corp (LWRRDC), Cotton Research & Development Corp, Murray-Darling Basin Commission. LWRRDC Occasional Paper 23/98, 9–12.
Sunderam RIM 1990. Toxicology of endosulfan in Australian freshwater ecosystems. MSc Thesis, University of Technology Sydney.
Sunderam RIM, Cheng DMH & Thompson GB 1992. Toxicity of endosulfan to native and introduced fish in Australia. Environmental Toxicology and Chemistry 11, 1469–1476.
Sunderam RIM, Thompson GB, Chapman JC & Cheng DMH 1994. Acute and chronic toxicity of endosulfan to two Australian cladocerans and their applicability in deriving water quality criteria. Archives of Environmental Contamination and Toxicology 38, 742–747.
Tomlin C 1994. The pesticide manual: A world compendium. 10th edn, British Crop Protection Council & Royal Society of Chemistry, Bath, UK.