Environ Sci Pollut Res DOI 10.1007/s11356-015-4758-2
RESEARCH ARTICLE
Evaluation of the ecotoxicological impact of the organochlorine chlordecone on soil microbial community structure, abundance, and function Chloé Merlin 1 & Marion Devers 1 & Jérémie Béguet 1 & Baptiste Boggio 1 & Nadine Rouard 1 & Fabrice Martin-Laurent 1
Received: 26 February 2015 / Accepted: 18 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract The insecticide chlordecone applied for decades in banana plantations currently contaminates 20,000 ha of arable land in the French West Indies. Although the impact of various pesticides on soil microorganisms has been studied, chlordecone toxicity to the soil microbial community has never been assessed. We investigated in two different soils (sandy loam and silty loam) exposed to different concentrations of CLD (D0, control; D1 and D10, 1 and 10 times the agronomical dose) over different periods of time (3, 7, and 32 days): (i) the fate of chlordecone by measuring 14Cchlordecone mass balance and (ii) the impact of chlordecone on microbial community structure, abundance, and function, using standardized methods (-A-RISA, taxon-specific quantitative PCR (qPCR), and 14C-compounds mineralizing activity). Mineralization of 14C-chlordecone was inferior below 1 % of initial 14C-activity. Less than 2 % of 14C-activity was retrieved from the water-soluble fraction, while most of it remained in the organic-solvent-extractable fraction (75 % of initial 14C-activity). Only 23 % of the remaining 14C-activity was measured in nonextractable fraction. The fate of chlordecone significantly differed between the two soils. The soluble and nonextractable fractions were significantly higher
in sandy loam soil than in silty loam soil. All the measured microbiological parameters allowed discriminating statistically the two soils and showed a variation over time. The genetic structure of the bacterial community remained insensitive to chlordecone exposure in silty loam soil. In response to chlordecone exposure, the abundance of Gram-negative bacterial groups (β-, γ-Proteobacteria, Planctomycetes, and Bacteroidetes) was significantly modified only in sandy loam soil. The mineralization of 14C-sodium acetate and 14C-2,4-D was insensitive to chlordecone exposure in silty loam soil. However, mineralization of 14C-sodium acetate was significantly reduced in soil microcosms of sandy loam soil exposed to chlordecone as compared to the control (D0). These data show that chlordecone exposure induced changes in microbial community taxonomic composition and function in one of the t w o so i l s, su g g e s t i n g m i c r o b i a l t o x i c i t y o f t h i s organochlorine. Keywords Soil . Organochlorine . Chlordecone . Microbial ecotoxicology
Introduction Responsible editor: Philippe Garrigues Electronic supplementary material The online version of this article (doi:10.1007/s11356-015-4758-2) contains supplementary material, which is available to authorized users. * Fabrice Martin-Laurent [email protected] 1
INRA, UMR 1347 Agroécologie, Pôle Ecoldur, 17 rue Sully, BP 86510, 21065 Dijon Cedex, France
The cyclodiene insecticide chlordecone (CLD) (1,1a,3,3a,4,5, 5,5a,5b,6-decachlorooctahydro-1,3,4-metheno-2Hcyclobuta[cd]pentalen-2-one) is one of the persistent organic pollutants (POPs) listed on Annex A of the Stockholm convention. It was classified as a persistent organic pollutant (POP) in 2009, but unfortunately, it had been used for 2 decades to control the development of the banana weevil, Cosmopolites sordidus, in banana plantations of the French West Indies (FWI, Guadeloupe and Martinique). Over that
Environ Sci Pollut Res
period (1972–1978 under the trademark Kepone® and 1981– 1993 under the trademark Curlone®), approximatively 300 t of CLD was applied in banana plantations and contained approximately 20,000 ha representing 25 % of agricultural soils in the FWI (Le Déaut and Procaccia 2009). Soils of banana plantations directly exposed to this insecticide are contaminated with CLD concentrations ranging between 0.1 and 37.4 mg kg−1 (Cabidoche et al. 2009; Martin-Laurent et al. 2014). In addition, although CLD displays high affinity for soil organic matter, its residues leach from agricultural soils to water resources and pollute drinking water in the FWI but also contaminate aquatic biota, including fish and giant shrimp locally called ouassou (Coat et al. 2006, 2011). Moreover, due to its biomagnification potential, CLD can be transferred to crops (tuber vegetables; cucurbits) and accumulate in poultry or cattle fed with CLD-contaminated resources (Faroon et al. 1995; Le Déaut and Procaccia 2009; Robert 2011; Cabidoche and Lesueur-Jannoyer 2012; Joindreville et al. 2013, 2014). Monitoring contamination revealed the presence of CLD not only in soils but also in rivers, springs, drinking water, and food crop products such as tuber vegetables, cucurbits, food fishing products (shrimp, fish), and food breeding products too (ducks, hens). Consequently, contamination of the environment and of the food chain by CLD chronically exposes FWI populations, with consequences on human health such as higher prevalence of prostate cancer in FWI populations exposed to CLD (Multigner et al. 2010), or delays in cognitive, visual, and motor development in young children exposed to CLD (Dallaire et al. 2012). A number of studies about CLD toxicity assessment have been published, notably by US scientists at the request of US Environmental Protection Agency after the Hopewell (Virginia, USA) incident in a CLD-manufacturing plant of Allied Chemical group. Yet, little is known on the mode of action of this insecticide. CLD toxicity is thought to mainly result from its hydrophobic nature (CLD solubility is 1 to 3 mg L−1 at pH 7, and log KOW is between 4.5 and 5), hence high affinity for the lipid bilayer of cell membranes (Faroon et al. 1995; Bonvallot and Dor 2004). Interaction of CLD with the lipid bilayer disorganizes animal cell membranes. This in turn alters transmembrane proteins such as calcium-dependent ATPases and causes the proton gradient to break down (Desaiah 1981). Interaction with cell membranes partly explains the action of CLD, which is known to alter cellular functions of the central nervous system. In addition, low CLD concentrations interact with hormone receptors (estrogenic receptors and other endocrine receptors) and induce a hormone-like effect that explains CLD toxicity on animal reproduction (Fujimori et al. 1982; Squibb and Tilson 1982; Linder et al. 1983; Desaiah 1985; Swartz et al. 1988; Kodavanti et al. 1990, 1993; Vaccari and Saba 1995). These studies were mainly carried out on animals. They reveal that the main mechanism involved in CLD toxicity relies on its
interaction with the lipid bilayer of the cell membrane of any living organism, suggesting that CLD can be toxic to a wide range of organisms. This hypothesis was recently verified on plants: CLD triggered a cascade of mechanisms, notably an increase in cytosolic Ca2+ and caspase-3-like activity, causing cell cycle disruption and apoptosis (Blondel et al. 2014). Although CLD contamination currently affects 20,000 ha of arable soils in the FWI and pesticides in general are known to impact soil microorganisms in various ways, to our best knowledge, only one study has assessed the ecotoxicological impact of CLD on the soil microbial community to date (Mercier et al. 2013). This study showed that three CLDcontaminated soils from FWI (andosol, ferralsol, and nitisol) had different bacterial community structures mainly attributable to soil physicochemical properties but could not be exclusively attributed to the CLD contamination level. However, further studies are still needed to evaluate the ecotoxicological impact of CLD on the soil microbial community and on supported ecosystemic functions. Using a microcosm study, we investigated the impact of CLD on the soil microbial community, which plays a pivotal role in soil ecosystemic services. We chose to work with two soils differing in their physicochemical properties, which had not been exposed to chlordecone and to any pesticide for at least 10 years. Soil microcosms were exposed to different doses of CLD (D0, control; D1, agronomical dose; D10, 10 times the agronomical dose) within the contamination range recorded in the FWI. To define the scenario of exposure of the microbial community, we monitored the fate of chlordecone in these two soils by establishing a mass balance of 14Cchlordecone. To monitor the impact of CLD on soil microbial community genetic structure, abundance, and function (degradation capability), we applied a range of standardized methods with different resolution levels. Overall, this approach provided a more accurate assessment of the ecotoxicity of this insecticide to the soil microbial community.
Materials and methods Chemicals [12C U]-Chlordecone Pestanal® (1,1a,3,3a,4,5,5,5a,5b,6decachloroctahydro-1,3,4-metheno-2Hcyclobuta[cd]pentalen-2-one) was purchased from SigmaAldrich (Schnelldorf, Germany) (chemical purity 99.7 %). [ 14 C U]-Chlordecone was purchased from Moravek Biochemicals (Brea, CA, USA) at a specific activity of 1443 MBq mmol−1 (radiochemical purity 99.9 %). CLD solubility ranged between 1 and 3 mg L−1 at ambient temperature (ATSDR 1995). [14C U]-2,4-D (analytical grade purity >99 %, specific activity 4600 MBq mmol−1) was purchased from Sigma-Aldrich (France). [ 1 4 C U]-sodium acetate
Environ Sci Pollut Res
(radiochemical purity 98.9 %, specific activity 1924 MBq mmol − 1 ) was kindly provided by CEA (Commissariat à l’Energie Atomique). Soil sampling and soil physicochemical properties Samples were collected in April 2014 from a sandy loam soil (Varenne-Saint-Germain, France, 46° 25′ 35″ N, 4° 1′ 35″ E) and a silty loam soil (Corcelles-lès-Citeaux, France, 47° 10′ 29″ N, 5° 4′ 32″ E) cropped with grasslands and unexposed to pesticides. Samples were collected in accordance with ISO 10381-6 recommendations (Soil quality-Sampling-Part 6: Guidance on the collection, handling and storage of soil for the assessment of aerobic microbial processes in the laboratory). Soil samples were sieved to 4 mm and stored at 4 °C until use. Soil physicochemical properties were determined by the Laboratory of Soil Analysis (INRA, Arras, France) following standardized procedures. They are presented in Table 1. The sandy loam soil was rich in coarse sand (84.5 %), whereas the silty loam soil was rich in silt and clay (54.9 and 21.1 %, respectively). The first was richer in organic C (2.87 %) and organic matter (4.96 %) than the later (1.52 and 2.62 %, respectively). Microcosm experiment Each soil was submitted to three different treatments: In D1, chlordecone was prepared in dichloromethane and applied at 3 mg kg−1, a dose corresponding to 3 kg ha−1 yr−1 in the field according to Le Déaut and Procaccia (2009); in D10, chlordecone was prepared the same way but applied at Table 1 Soil physicochemical properties
Sampling site
30 mg kg−1, i.e., ten times the agronomic dose; and D0 was the control, with only dichloromethane added. Each microcosm contained 80 g of soil dry weight equivalent. Under a laboratory hood, chlordecone solution (D1 or D10) or dichloromethane (D0) was added to 8 g of soil. Dichloromethane was evaporated and then thoroughly mixed with the rest of the sample and added with sterilized water to reach 60 % of soil water-holding capacity. All microcosms were incubated in airtight jars at 20 °C in the dark. At each sampling date (3, 7, and 32 days of incubation), soil samples were collected from each microcosm as follows: (i) 1 g of soil was kept at −20 °C for DNA-based analyses (q-PCR and A-RISA); (ii) 20 g of soil was directly used for biochemical analyses (radiorespirometry). For each condition, microcosms were made in triplicate (ntot =18). A second series of microcosms (10 g of soil dry weight equivalent) was prepared as described above except that 14C10-chlordecone was added as a tracer (73.3 Bq g−1 and 126.6 Bq g−1 for the D1 and D10 treatments, respectively), to study the fate of 14C-labelled chlordecone in the two soils.
Fate of chlordecone in soil microcosms We defined the scenario of exposure of soil microorganisms by monitoring the fate of 14C-chlordecone in the two soils: We estimated the distribution of 14C residues in 14CO2 resulting from 14C-chlordecone mineralization in the water-soluble and extractable fractions as well as in the nonextractable 14C residue (NER). The amount of 14CO2 evolved from 14C-chlordecone as a result of degradation by microorganisms was estimated by Varenne-Saint-Germain
Corcelles-lès-Citeaux
Horizon 0-20 cm
Horizon 0-20 cm
Sandy loam soil
Silty loam soil
Clay (
RESEARCH ARTICLE
Evaluation of the ecotoxicological impact of the organochlorine chlordecone on soil microbial community structure, abundance, and function Chloé Merlin 1 & Marion Devers 1 & Jérémie Béguet 1 & Baptiste Boggio 1 & Nadine Rouard 1 & Fabrice Martin-Laurent 1
Received: 26 February 2015 / Accepted: 18 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract The insecticide chlordecone applied for decades in banana plantations currently contaminates 20,000 ha of arable land in the French West Indies. Although the impact of various pesticides on soil microorganisms has been studied, chlordecone toxicity to the soil microbial community has never been assessed. We investigated in two different soils (sandy loam and silty loam) exposed to different concentrations of CLD (D0, control; D1 and D10, 1 and 10 times the agronomical dose) over different periods of time (3, 7, and 32 days): (i) the fate of chlordecone by measuring 14Cchlordecone mass balance and (ii) the impact of chlordecone on microbial community structure, abundance, and function, using standardized methods (-A-RISA, taxon-specific quantitative PCR (qPCR), and 14C-compounds mineralizing activity). Mineralization of 14C-chlordecone was inferior below 1 % of initial 14C-activity. Less than 2 % of 14C-activity was retrieved from the water-soluble fraction, while most of it remained in the organic-solvent-extractable fraction (75 % of initial 14C-activity). Only 23 % of the remaining 14C-activity was measured in nonextractable fraction. The fate of chlordecone significantly differed between the two soils. The soluble and nonextractable fractions were significantly higher
in sandy loam soil than in silty loam soil. All the measured microbiological parameters allowed discriminating statistically the two soils and showed a variation over time. The genetic structure of the bacterial community remained insensitive to chlordecone exposure in silty loam soil. In response to chlordecone exposure, the abundance of Gram-negative bacterial groups (β-, γ-Proteobacteria, Planctomycetes, and Bacteroidetes) was significantly modified only in sandy loam soil. The mineralization of 14C-sodium acetate and 14C-2,4-D was insensitive to chlordecone exposure in silty loam soil. However, mineralization of 14C-sodium acetate was significantly reduced in soil microcosms of sandy loam soil exposed to chlordecone as compared to the control (D0). These data show that chlordecone exposure induced changes in microbial community taxonomic composition and function in one of the t w o so i l s, su g g e s t i n g m i c r o b i a l t o x i c i t y o f t h i s organochlorine. Keywords Soil . Organochlorine . Chlordecone . Microbial ecotoxicology
Introduction Responsible editor: Philippe Garrigues Electronic supplementary material The online version of this article (doi:10.1007/s11356-015-4758-2) contains supplementary material, which is available to authorized users. * Fabrice Martin-Laurent [email protected] 1
INRA, UMR 1347 Agroécologie, Pôle Ecoldur, 17 rue Sully, BP 86510, 21065 Dijon Cedex, France
The cyclodiene insecticide chlordecone (CLD) (1,1a,3,3a,4,5, 5,5a,5b,6-decachlorooctahydro-1,3,4-metheno-2Hcyclobuta[cd]pentalen-2-one) is one of the persistent organic pollutants (POPs) listed on Annex A of the Stockholm convention. It was classified as a persistent organic pollutant (POP) in 2009, but unfortunately, it had been used for 2 decades to control the development of the banana weevil, Cosmopolites sordidus, in banana plantations of the French West Indies (FWI, Guadeloupe and Martinique). Over that
Environ Sci Pollut Res
period (1972–1978 under the trademark Kepone® and 1981– 1993 under the trademark Curlone®), approximatively 300 t of CLD was applied in banana plantations and contained approximately 20,000 ha representing 25 % of agricultural soils in the FWI (Le Déaut and Procaccia 2009). Soils of banana plantations directly exposed to this insecticide are contaminated with CLD concentrations ranging between 0.1 and 37.4 mg kg−1 (Cabidoche et al. 2009; Martin-Laurent et al. 2014). In addition, although CLD displays high affinity for soil organic matter, its residues leach from agricultural soils to water resources and pollute drinking water in the FWI but also contaminate aquatic biota, including fish and giant shrimp locally called ouassou (Coat et al. 2006, 2011). Moreover, due to its biomagnification potential, CLD can be transferred to crops (tuber vegetables; cucurbits) and accumulate in poultry or cattle fed with CLD-contaminated resources (Faroon et al. 1995; Le Déaut and Procaccia 2009; Robert 2011; Cabidoche and Lesueur-Jannoyer 2012; Joindreville et al. 2013, 2014). Monitoring contamination revealed the presence of CLD not only in soils but also in rivers, springs, drinking water, and food crop products such as tuber vegetables, cucurbits, food fishing products (shrimp, fish), and food breeding products too (ducks, hens). Consequently, contamination of the environment and of the food chain by CLD chronically exposes FWI populations, with consequences on human health such as higher prevalence of prostate cancer in FWI populations exposed to CLD (Multigner et al. 2010), or delays in cognitive, visual, and motor development in young children exposed to CLD (Dallaire et al. 2012). A number of studies about CLD toxicity assessment have been published, notably by US scientists at the request of US Environmental Protection Agency after the Hopewell (Virginia, USA) incident in a CLD-manufacturing plant of Allied Chemical group. Yet, little is known on the mode of action of this insecticide. CLD toxicity is thought to mainly result from its hydrophobic nature (CLD solubility is 1 to 3 mg L−1 at pH 7, and log KOW is between 4.5 and 5), hence high affinity for the lipid bilayer of cell membranes (Faroon et al. 1995; Bonvallot and Dor 2004). Interaction of CLD with the lipid bilayer disorganizes animal cell membranes. This in turn alters transmembrane proteins such as calcium-dependent ATPases and causes the proton gradient to break down (Desaiah 1981). Interaction with cell membranes partly explains the action of CLD, which is known to alter cellular functions of the central nervous system. In addition, low CLD concentrations interact with hormone receptors (estrogenic receptors and other endocrine receptors) and induce a hormone-like effect that explains CLD toxicity on animal reproduction (Fujimori et al. 1982; Squibb and Tilson 1982; Linder et al. 1983; Desaiah 1985; Swartz et al. 1988; Kodavanti et al. 1990, 1993; Vaccari and Saba 1995). These studies were mainly carried out on animals. They reveal that the main mechanism involved in CLD toxicity relies on its
interaction with the lipid bilayer of the cell membrane of any living organism, suggesting that CLD can be toxic to a wide range of organisms. This hypothesis was recently verified on plants: CLD triggered a cascade of mechanisms, notably an increase in cytosolic Ca2+ and caspase-3-like activity, causing cell cycle disruption and apoptosis (Blondel et al. 2014). Although CLD contamination currently affects 20,000 ha of arable soils in the FWI and pesticides in general are known to impact soil microorganisms in various ways, to our best knowledge, only one study has assessed the ecotoxicological impact of CLD on the soil microbial community to date (Mercier et al. 2013). This study showed that three CLDcontaminated soils from FWI (andosol, ferralsol, and nitisol) had different bacterial community structures mainly attributable to soil physicochemical properties but could not be exclusively attributed to the CLD contamination level. However, further studies are still needed to evaluate the ecotoxicological impact of CLD on the soil microbial community and on supported ecosystemic functions. Using a microcosm study, we investigated the impact of CLD on the soil microbial community, which plays a pivotal role in soil ecosystemic services. We chose to work with two soils differing in their physicochemical properties, which had not been exposed to chlordecone and to any pesticide for at least 10 years. Soil microcosms were exposed to different doses of CLD (D0, control; D1, agronomical dose; D10, 10 times the agronomical dose) within the contamination range recorded in the FWI. To define the scenario of exposure of the microbial community, we monitored the fate of chlordecone in these two soils by establishing a mass balance of 14Cchlordecone. To monitor the impact of CLD on soil microbial community genetic structure, abundance, and function (degradation capability), we applied a range of standardized methods with different resolution levels. Overall, this approach provided a more accurate assessment of the ecotoxicity of this insecticide to the soil microbial community.
Materials and methods Chemicals [12C U]-Chlordecone Pestanal® (1,1a,3,3a,4,5,5,5a,5b,6decachloroctahydro-1,3,4-metheno-2Hcyclobuta[cd]pentalen-2-one) was purchased from SigmaAldrich (Schnelldorf, Germany) (chemical purity 99.7 %). [ 14 C U]-Chlordecone was purchased from Moravek Biochemicals (Brea, CA, USA) at a specific activity of 1443 MBq mmol−1 (radiochemical purity 99.9 %). CLD solubility ranged between 1 and 3 mg L−1 at ambient temperature (ATSDR 1995). [14C U]-2,4-D (analytical grade purity >99 %, specific activity 4600 MBq mmol−1) was purchased from Sigma-Aldrich (France). [ 1 4 C U]-sodium acetate
Environ Sci Pollut Res
(radiochemical purity 98.9 %, specific activity 1924 MBq mmol − 1 ) was kindly provided by CEA (Commissariat à l’Energie Atomique). Soil sampling and soil physicochemical properties Samples were collected in April 2014 from a sandy loam soil (Varenne-Saint-Germain, France, 46° 25′ 35″ N, 4° 1′ 35″ E) and a silty loam soil (Corcelles-lès-Citeaux, France, 47° 10′ 29″ N, 5° 4′ 32″ E) cropped with grasslands and unexposed to pesticides. Samples were collected in accordance with ISO 10381-6 recommendations (Soil quality-Sampling-Part 6: Guidance on the collection, handling and storage of soil for the assessment of aerobic microbial processes in the laboratory). Soil samples were sieved to 4 mm and stored at 4 °C until use. Soil physicochemical properties were determined by the Laboratory of Soil Analysis (INRA, Arras, France) following standardized procedures. They are presented in Table 1. The sandy loam soil was rich in coarse sand (84.5 %), whereas the silty loam soil was rich in silt and clay (54.9 and 21.1 %, respectively). The first was richer in organic C (2.87 %) and organic matter (4.96 %) than the later (1.52 and 2.62 %, respectively). Microcosm experiment Each soil was submitted to three different treatments: In D1, chlordecone was prepared in dichloromethane and applied at 3 mg kg−1, a dose corresponding to 3 kg ha−1 yr−1 in the field according to Le Déaut and Procaccia (2009); in D10, chlordecone was prepared the same way but applied at Table 1 Soil physicochemical properties
Sampling site
30 mg kg−1, i.e., ten times the agronomic dose; and D0 was the control, with only dichloromethane added. Each microcosm contained 80 g of soil dry weight equivalent. Under a laboratory hood, chlordecone solution (D1 or D10) or dichloromethane (D0) was added to 8 g of soil. Dichloromethane was evaporated and then thoroughly mixed with the rest of the sample and added with sterilized water to reach 60 % of soil water-holding capacity. All microcosms were incubated in airtight jars at 20 °C in the dark. At each sampling date (3, 7, and 32 days of incubation), soil samples were collected from each microcosm as follows: (i) 1 g of soil was kept at −20 °C for DNA-based analyses (q-PCR and A-RISA); (ii) 20 g of soil was directly used for biochemical analyses (radiorespirometry). For each condition, microcosms were made in triplicate (ntot =18). A second series of microcosms (10 g of soil dry weight equivalent) was prepared as described above except that 14C10-chlordecone was added as a tracer (73.3 Bq g−1 and 126.6 Bq g−1 for the D1 and D10 treatments, respectively), to study the fate of 14C-labelled chlordecone in the two soils.
Fate of chlordecone in soil microcosms We defined the scenario of exposure of soil microorganisms by monitoring the fate of 14C-chlordecone in the two soils: We estimated the distribution of 14C residues in 14CO2 resulting from 14C-chlordecone mineralization in the water-soluble and extractable fractions as well as in the nonextractable 14C residue (NER). The amount of 14CO2 evolved from 14C-chlordecone as a result of degradation by microorganisms was estimated by Varenne-Saint-Germain
Corcelles-lès-Citeaux
Horizon 0-20 cm
Horizon 0-20 cm
Sandy loam soil
Silty loam soil
Clay (
