Species Diversity Hinders Adaptation to Toxicants

Abstract

Environmental toxicants such as pesticides exert strong selection pressure on many species. While the resulting development of pesticide resistance in agricultural pest species is well-known, reports on the extent of adaptation in exposed nontarget species are contradictory. These contradictory reports highlight a continuing challenge in understanding the relevant ecological mechanisms that facilitate or hinder adaptation to toxicants in the field. Here we show that species diversity hinders the adaptation to toxicants. In agricultural streams with low diversity, we observed an up to 8-fold increase in insecticide tolerance in a total of 17 macroinvertebrate species that was not observed in more diverse communities under similar contamination. High species diversity occurred independently from adjacent nonpolluted refuge areas. Therefore, the low level of adaptation in diverse streams cannot be explained by an increased recolonization of sensitive individuals from refuge areas. Instead, high intraspecific competition may facilitate the selection for increased tolerance in low-diverse communities. In diverse communities, by contrast, species interactions may reduce intraspecific competition and, thus, the potential for developing toxicant resistance. We suggest that this mechanism may be the general case in adaptation to environmental stressors. Additionally, we conclude that the current framework for risk assessment of pesticides is not protective against selection for tolerant organisms and the associated risk of genetic erosion.

Pesticide pollution in freshwater paves the way for schistosomiasis transmission

“…Concentrations of the highest exposure (TU max ) ranged from <−5 to −1.21, representing a 10 th to <100,000 th of the acute lethal median concentration for standard test organisms. This range of pesticide exposure is comparable to several previous studies conducted in agricultural streams of Europe and Australia 22,30,35 .…”

“…It cannot be excluded that the overall pesticide toxicity might have been even higher due to additional compounds such as pyrethroid insecticides that require different analytical methods. However, pyrethroids typically occur concurrently with the compounds detected and show a similar range of toxicity 35,37 . In previous studies 22,30,35 pesticides have been sampled during the peak exposure following run-off events after heavy rainfalls.…”

“…However, pyrethroids typically occur concurrently with the compounds detected and show a similar range of toxicity 35,37 . In previous studies 22,30,35 pesticides have been sampled during the peak exposure following run-off events after heavy rainfalls. Event-triggered sampling was not feasible in our study area.…”

“…Toxic units of −1 or −2 represent pesticide concentrations of 1/10 or 1/100 of the LC 50 , respectively. Toxic units of pesticides in agricultural European, Siberian and Australian streams typically reach from ≤−5 to −1 34,35 .…”

“…The most toxic substances found were the organophosphorous and carbamate insecticides bendiocarb (most toxic substance at 17 sites, TU up to −1.66, Table S3 in the Supplementary Material), diazinon (most toxic at 15 sites, TU up to −1.71) and pirimiphos-methyl (most toxic at 5 sites, TU up to −1.21). Due to difficulties in the quantification of very low pesticide concentrations, the minimum TU max was set to −5 for further analyses, a threshold at which typically no ecological effects have been observed in field studies 29,34,35 .…”