Skip to Main content Skip to Navigation
Theses

Modélisation de distribution de sensibilité des espèces pour la protection des communautés

Abstract : Human activities have numerous effects on ecosystems. They can for example result in spillage or leakage of contaminants in ecosystems as during spreading of pesticides in agricultural lands. Environmental risk assessment allows to characterize those effects to maintain them at a level considered acceptable. This assessment is usually based on monospecific bioassays, but concentration-response relationships thus described can then be included in Species Sensitivity Distributions (SSD). Those SSD, widely used in environmental risk assessment, then allows to define a concentration that only have an effect on a small fraction of the species found in an ecosystem (usually 5%, leading to hazardous concentrations for 5% of species). One of the most important problems of this kind of approaches is that it still lacks environmental relevance. Monospecific bioassays indeed providing the information to model SSD are usually realised in non realist conditions (organisms isolated and thus not subject to biotic interactions, experimental conditions far from environmental reality, obtention of a unique sensitivity value for a species whereas this sensibility can be variable from one population to another...). This lack of realism is compensated by the application of an assessment factor which divides the final value by a value that increases as the realised estimation is distant from environmental reality, thus giving lower values that are considered more protectives. The work presented here sought to integrate (i) the effect of intraspecific variability on SSD curves parameters, in particular on hazardous concentrations for 5% of species and (ii) the combined effect of an herbicide and interspecific competition on the vegetal communities that are grass stripes following several modeling approaches for each data processing step. Results show the importance of the integration of all this biological complexity in environmental risk assessment procedures as well as the risk of underestimation of assessment factors usually applied on results from SSD modeling.
Complete list of metadata

https://hal.univ-lorraine.fr/tel-03354004
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Friday, September 24, 2021 - 3:18:41 PM
Last modification on : Tuesday, May 10, 2022 - 3:38:05 PM

File

DDOC_T_2021_0065_BAILLARD.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-03354004, version 1

Citation

Vincent Baillard. Modélisation de distribution de sensibilité des espèces pour la protection des communautés. Sciences de l'environnement. Université de Lorraine, 2021. Français. ⟨NNT : 2021LORR0065⟩. ⟨tel-03354004⟩

Share

Metrics

Record views

83

Files downloads

156