PhD Project [completed]: Projected effect of global change on species' change in extinction risk
Climate change has become one of the major drivers of biodiversity loss, its effects are not only already evident across all levels of biological organization (from genes to ecosystems) but are projected to increase in the coming decades. The probability of a species or population being negatively impacted by climate change (i.e., risk) is determined by the occurrence of adverse climatic events or trends (i.e., hazard), the occurrence of the species or population in areas that could be impacted (i.e., exposure), and their predisposition to be adversely affected, including their sensitivity or susceptibility and lack of capacity to cope or adapt (i.e., vulnerability). Species or populations can adapt to adverse climatic conditions by shifting their geographical distribution or adapting in situ, generally by changing their phenology, morphology or physiology.
Recent efforts to assess the impacts of climate change have predominantly relied on bioclimatic niche modeling, which predicts species’ or populations’ distributions by linking their geographical range and bioclimatic variables. However, these models assume that all species are affected and will respond to climate change similarly, and do not consider differences in vulnerability and exposure. Trait-based assessments have aimed to address this gap, identifying which traits influence risk, allowing assessing multiple species simultaneously in a simple way and serving as a useful tool for prioritizing conservation actions, especially in the absence of distribution data. However, their applicability can be limited as they are not spatially explicit, the relationship between traits and responses is still uncertain, there are gaps in trait data availability and the approach is generally implemented at the species level, ignoring intraspecific differences in exposure, vulnerability and hazard.
The objective of this thesis was to overcome some of these limitations for birds and terrestrial non-volant mammals. To overcome gaps in mammal trait data availability, I compiled in my first chapter COMBINE: A Coalesced Mammal Database of Intrinsic and Extrinsic traits data for 54 traits for 6,234 mammal species, using data from 14 different data sources. These traits covered aspects such as physiology, reproduction, behavior, longevity, diet, and dispersal. I further filled in gaps in the data through a phylogenetic multiple imputation procedure, providing a complete dataset for 21 traits. All data sources and imputed data were flagged, facilitating identifying the origin of the data. This dataset constitutes a useful tool for large-scale ecological and conservation analyses that use traits, including identifying species at risk from climate change.
In my second analytical chapter, Relative latitude, temperature increase and breadth of climatic niche influence mammal populations’ response to climate change, I identified current terrestrial non-volant mammal responses to climate change and the intrinsic traits and environmental factors influencing risk. To achieve this, I first performed a literature review on responses to climate change and categorized them into changes in (a) distribution and abundance, (b) phenology, and (c) morphology. I also identified the direction of each type of response: expansion or contraction for distribution and abundance, advance or delay for phenology, increase or decrease for body size, and no change if no response was detected. To model the relationship between risk from climate change and intrinsic and environmental factors, I focused exclusively on distribution and abundance responses due to their direct relationship. I then selected and obtained data for a series of intrinsic traits and environmental factors previously associated with climate change risk. To account for intraspecific variability in environmental factors, I identified populations of the species that experience similar climatic conditions. As these populations were distributed across large geographical areas, I grouped the responses by species and country, reducing the number of instances of opposing or mixed responses (i.e., different studies for the same species and country reporting distribution and abundance contractions and expansions or phenological advances and delays) and allowing the inclusion of the location of the response within the population. I obtained 382 responses belonging to 130 species located in 30 countries. Most of these responses were distribution and abundance responses (80.6%) while phenological and morphological changes constituted 4.5% (17 responses) and 10.2% (39 responses) respectively. The remaining 4.7% did not fit into any of these categories. Regarding distribution and abundance responses, there were more than twice as many contractions (46.43%) as expansions (20.78%), while in 32.79% of cases there was no clear response. The results of our model indicated that contractions were more likely at the warm edge of the population, while expansions were more likely at the cold edge. Small litter size, hibernation, high temperature increase, low climate seasonality and low altitudinal breadth were also linked to an elevated risk of experiencing a negative response.
In my third analytical chapter, Local environmental factors influence bird distribution and phenological responses to climate change, I followed the same approach but focused on bird distribution and abundance and spring phenological responses to climate change. I also gathered data for nine intrinsic bird traits that have been previously hypothesized to be relevant in determining responses to climate change. This allowed me to identify which intrinsic traits and environmental factors influence experiencing distribution contractions or expansions and spring phenological advances, delays or no changes. I obtained 3,012 responses for 918 species located in 32 countries, 60% of them were distribution and abundance responses and the remaining 40% were spring phenology responses. I found that environmental factors played an important role in determining both distribution and abundance and phenological responses to climate change. Maximum temperature, restricted climate seasonality, relative latitudinal position, and maximum longevity influenced the probability of experiencing contractions and a subsequent increase in risk. Similarly, maximum temperature, climate seasonality, relative latitudinal position, and temperature increase influenced the probability of experiencing advances in spring phenology.
The results presented in this thesis constitute an advance in current knowledge on the variables influencing responses to climate change locally and serve as a starting point for future research.
Soria, C.D., Pacifici, M., Di Marco, M., Stephen, S.M., Rondinini, C., (2021) COMBINE: a coalesced mammal database of intrinsic and extrinsic traits. Ecology. https://doi.org/10.1002/ecy.3344 → Open Access repository
Lumbierres, M., Dahal, P.R., Soria, C.D., Di Marco, M., Butchart, S.H.M., Donald, P.F., Rondinini, C. (2022) Area of Habitat maps for the world’s terrestrial birds and mammals. Scientific Data 9, 749 (2022). https://doi.org/10.1038/s41597-022-01838-w → Open Access repository
Soria, C. (2022) Where will future Key Biodiversity Areas be identified? Projected effect of global change on species’ change in extinction risk. PhD Thesis. Sapienza Università di Roma, Rome, Italy.→ Open Access Repository
Lumbierres, M., Dahal, P.R., Soria, C.D., Di Marco, M., Butchart, S.H.M., Donald, P.F., Rondinini, C. (2022) Area of Habitat maps for the world’s terrestrial birds and mammals. Dryad, Dataset, https://doi.org/10.5061/dryad.02v6wwq48 → Open Access Repository
Soria, C.D., Pacifici, M., Di Marco, M., Stephen, S.M., Rondinini, C., (2021) Data from COMBINE: a coalesced mammal database of intrinsic and extrinsic traits. (published in Ecology https://esajournals.onlinelibrary.wiley.com/doi/full/10.1002/ecy.3344#support-information-section) → Open Access Repository