Community trait responses to land-use conversion and temperature increase, with new approaches to investigate thermal tolerance and resilience of ectotherms

Abstract

Land-use conversion modifies environmental conditions and threatens local biodiversity. Yet, little is known about the ecological and physiological drivers and mechanisms leading to biodiversity loss. A functional trait-based approach, however, provides a powerful and general framework now widely used to explain the relationship between species community and their environment. In this thesis, I firstly investigated the methodology used in upper thermal limit measurements to accurately estimate this key trait in predicting species responses to temperature changes. I used ants, one of the model organisms in thermal ecology, to review and test the methods used for measuring Critical Thermal Maximum (CTmax). The results show that variation in ramping rate speeds used in CTmax assays generate important methodological biases, and may raise confounding impact ecological interpretations. Therefore, using a combination of dynamic and static thermal assays, with field observations of species maximal temperature activity, I identified a suitable ramping rate. Assays using a fast ramping rate (i.e., 1.0 °C min-1) are recommended, as the retrieved CTmax values should present both biological and ecological relevance. Then, I investigated thermal resilience through the study of post heat-coma recovery, an important trait to estimate species fitness under extreme heat temperatures. This work focused on the relationship between body mass and recovery after heat-shock. The results provided a mechanistic understanding of community weighted mean of body size changes in response to temperature increase following land-use change or global warming. Additionally, field experiments directly linked the role of post heat-coma recovery to fitness benefits in limiting predation risks. In the third chapter, I investigated the impacts of land conversion from forests to rubber plantation on ant assemblage trait composition. Species thermal (CTmax) and morphological traits were used to understand trait responses to habitat change. I compared the relative contributions of thermal and morphological traits in explaining the drivers of land-use change. Here, the model using only thermal traits presented a significant correlation with habitat type, contrary to the morphological traits tested. As a result, thermal traits best characterize ecosystem filtering process than morphological traits between ant assemblages inhabiting forests or rubber plantations. In the fourth chapter, I investigated the impacts of urbanization and its land-use with a trait-based approach at the community-level, while accounting for the scale of effect. Ant communities were sampled along an urban to sub-rural gradient in two subtropical cities. These results show that considering the scale of effect is important in functional trait-based approach, and environmental predictors from different scales may affect the inference of community trait-environment relationship. Overall, my results present new approaches in the study of thermal tolerance and resilience; and support the importance and uniqueness of thermal traits in explaining species community and their responses to land-use changes. The use of different traits and their integration can help understand the relationship between species traits and environmental filters, identifying land-use change drivers. These results should support the development of new ecological applications in the study of individual species and community responses to global changes, ultimately supporting the protection of ecosystems.