Skip to main content

Characterizing the risk of pesticide use on amphibian and reptile populations based on multiple, ecologically relevant responses

Final Report Summary - HERPESTI (Characterizing the risk of pesticide use on amphibian and reptile populations based on multiple, ecologically relevant responses)

Period: 01/04/2014 – 30/11/2015

Amphibians and reptiles (known in conjunction as ‘herps’) are the two vertebrate taxa with a higher proportion of endangered species (31 and 23% of the evaluated species, respectively[1]). Anthropogenic pollution is recognized as one of the major factors threatening amphibians and reptiles, and consequently, ecotoxicology of these two groups has progressed over the last decade[2-3]. Pesticide pollution is of great concern for amphibians and reptiles because i) pesticides are diffusely released to the environment in great amounts (~250 000 yearly tons of active ingredients in the EU); ii) they are released directly to the habitats inhabited by amphibians and reptiles; iii) they affect large extensions of soils and waters, limiting the ability of amphibians and reptiles to avoid them; and iv) in agricultural areas, natural habitats have undergone a transformation process which means that wildlife populations surviving there are in suboptimal conditions. Pesticide application adds an impact to populations that must deal with a stressful environment.
Although the body of knowledge on pesticides and herps has exploded relative to the early days of ecotoxicology, a disconcerting number of commonly used pesticides has never been studied in regard to amphibians[4] or reptiles[5]. One of the reasons to explain this is that amphibians and reptiles have been the only vertebrates not considered in ecotoxicological risk assessments that manufacturers must present before a pesticide can be approved for use. The former, currently derogated European Directive 91/414 concerning the placing of plant protection products (PPP, the group of chemicals including pesticides) on the market, established that for vertebrate ecotoxicology only fish (for the aquatic environment), bird and mammal (for the terrestrial environment) toxicity data had to be included in the risk assessment. The new Regulation (EC) 1107/2009 introduced some changes in this context. Published information regarding toxicity of an active ingredient on amphibians or reptiles will have to be included and considered in risk assessment; however, this novelty has the obvious concern that nobody but the manufacturer gets access to a new active substance before it is placed on the market, and therefore no published information on its toxicity is available before product registration. Neglecting amphibians and reptiles in pesticide risk assessment is worrisome because both groups exhibit particular mechanisms that make them especially vulnerable to the impact of pesticides. Dermal assimilation of pesticides in the terrestrial environment is ignored by tests with birds and mammals, but the permeable nature of amphibian skin or the strong links of reptiles with soils (especially during egg laying) where pesticides are applied make this route a very significant way of exposure of these animals.
This project was designed to test the degree of protection that European legislation on approval and use of pesticides confers to amphibians and reptiles. With this purpose, a risk characterization was developed in four working objectives, which design and results are explained below.

Objective 1. Analysis of the risk of exposure of amphibians and reptiles to pesticides on the basis of spatial and behavioural studies

We addressed this objective through two experiments and a field study. In one of the experiments, we tested the oviposition site selection by female palmate newts (Lissotriton helveticus) as a function of the presence in the water of predicted environmental concentrations of three pesticides. When offered the possibility of selecting either pesticide-treated or clean waters, the percentage of eggs laid on each side was never significantly different from 50%, which shows a random selection of the site by females.
The field study was conducted in a population of wall lizards (Podarcis muralis) inhabiting a vineyard area in the Südliche Weinstrasse (Rhineland-Palatinate, Germany). Animals were individually identified using Passive Integrated Transponders (PIT) to study their movements and habitat use. We tested the hypothesis that lizards would prefer staying in walls located at the field edges (off-field) rather going into the fields (in-field), and that their eventual occurrence in-field would be conditioned by the pesticide application. Our results showed that lizards did not necessarily prefer to be present off-field, occurring also frequently inside the fields. Actually, phenological observations point to the use of in-field areas for egg laying, which would be explained by the facility of manipulation (and nest building) of cultivated soils. We contrasted the presence of lizards in the field with information on pesticide application provided by land owners and observed that application did not seem either to influence the presence of lizards in potentially exposed areas, as no correlation was found between the frequency of occurrence and the time elapsed since last application.
The absence of patterns of avoidance found in the objective 1 suggests a high potential for exposure, both by direct overspray or by dermal and oral uptake after pesticide application, of aquatic and terrestrial phases of amphibians and reptiles. This would reflect a significant uncertainty with respect to bird and mammal-based risk assessment. In contrast with surrogate taxa, herps would stay exposed during prolonged periods of time because of their low chances of moving away from treated areas.

Objective 2. Comparison of the relative importance of different ways of exposure to pesticides for amphibians and reptiles in aquatic and terrestrial environments

This objective was addressed through four different experiments, two in the aquatic phases of common frogs (Rana temporaria) and palmate newts, and two more with terrestrial stages of common frogs and wall lizards, the later including also an in ovo exposure. In the aquatic experiments, some differences in toxicity depending on the exposure route to certain pesticides was observed; for instance, the herbicide glufosinate ammonium was more toxic to larval newts when spiked to sediments in comparison to spiked water. The simultaneous application of the acaricide fenpyroximate to different matrices (i.e. water, food and sediments), which tended to reflect a more realistic scenario, resulted in stronger effects on larval frog growth than when only one matrix was treated with the pesticide, which is the normal approach in toxicity testing. In the terrestrial part, no significant differences in toxicity derived from the exposure to pesticides through contaminated food and soil was found, although the incubation of wall lizard eggs in contaminated soils resulted in delayed growth rates manifested later during the juvenile stage.
From a risk assessment perspective, it is importance to notice the impact of the different exposure matrices, as normally only one is considered (dermal for the aquatic risk assessment and oral for the terrestrial one). Our results show that for aquatic amphibians, the ingestion of contaminated food and/or sediments may be a relevant way of pesticide uptake. Most interestingly, the manifestation of effects of a lizard embryonic exposure later in the post-hatching stage constitutes a new challenge for risk assessment purposes, as current protocols do not consider this type of lagged effects.

Objective 3) Assessment of effectiveness of toxicity data based on fish, birds or mammals to protect amphibians and reptiles from the impact of pesticides.

We compiled scientific literature to compare sensitivity data among taxa. After reviewing 1274 we found that genotoxicity was clearly related to amphibian sensitivity to pesticides. Other responses showed some partial correlations; immunotoxic effects were explained by exposure time and oxidative stress biomarkers responded to the exposure concentration Genotoxicity could be a useful indicator of amphibian sensitivity to pesticides but, as it is usually tested only in substances suspected to exert such effect, it requires confirmation in a wider range of pesticides. Functional responses like immunotoxocity or oxidative stress appear as suitable indicators of chronic and acute toxicity, respectively. Further steps shall compare this information based on uncommon endpoints with that from surrogate species in order to check the appropriateness of surrogates for covering pesticide risks on herps.
Objective 4) Implementation of ecologically relevant responses specific to amphibians and reptiles as endpoints for the study of the effects of pesticides.

We have focused the study ecologically relevant endpoints in the behavioural responses after or during pesticide exposure. Behaviour is a suitable indicator of neurotoxicity, but at the same time reflects a responses which alteration might have direct effects on both the individual and, depending on the behaviour, the population. We tested prey capture behaviour by juvenile common frogs after exposure to three different pesticides, and also whether the contamination of water with the same three pesticides influenced how female newt laid their eggs. With regards to the latter experiment, although we had hypothesized, based on previous information[6] that the percentage of eggs properly wrapped would decrease in contaminated waters, no differences in wrapping success were found between any of the pesticides and the controls. In order to consider behaviour as a relevant endpoint, it would be necessary to consider the mode of action of each substance, and to estimate whether neurotoxic effects further reflected in behaviour modification may exist.

Overall summary

Taken together, the results obtained from the pool of tasks of the HerPesti project shows a clear necessity of reconsidering the current EU scheme of pesticide risk assessment in vertebrates. We provide some evidence showing that exposure scenarios currently used for fish, birds and mammals are not appropriate to cover the risks to amphibians and reptiles in a significant way.


[1] IUCN (2015) The IUCN Red List of Threatened Species 2015.3.
[2] Ortiz-Santaliestra ME (2009) pp 239-267 In Santos EB: Ecotoxicology Research Developments. Nova Science
[3] Sparling DW et al. (2010) pp 1-11 In Sparling DW et al: Ecotoxicology of Amphibians and Reptiles. CRC Press
[4] Hayes TB (2006) Environ Health Persp 114 (S1), 40-50
[5] Pauli BD et al. (2010) pp 203-224 In Sparling DW et al: Ecotoxicology of Amphibians and Reptiles. CRC Press
[6] Ortiz-Santaliestra et al. (2007) Aquat Toxicol 85, 251-257