Final Activity Report Summary - PHENO-RANGE-EDGE (Modelling constraints on tree range shifts under climate change: regional and local processes)
The PHENO-RANGE-EDGE project aimed to provide more accurate predictions of the influence of global warming on the distribution and abundance of tree species living in the temperate zone and especially in north America. To do so, both experimental and modelling efforts were conducted throughout this project.
The project was broken down into the following three parts:
1. identification of the factors limiting species range at global scale and implication for diversity. We firstly considered the distribution pattern of all North American tree species, i.e. of 599 species. Using observed range area, we showed evidence for a latitudinal gradient in species range size, with larger ranges in the north. We furthermore showed that tree species richness and mean species range area were negatively covariant across latitudes, suggesting the possibility that the causes of the increase in range area at higher latitudes might in turn also underpin the decreasing latitudinal trend in species diversity. Going deeper in the search of causal factors for this described patterns, we showed that the gradient in range size was consistent with the interspecific variation of thermal tolerance of trees in north America. In other words, this result highlighted the impact of climate on species distribution at continental scale.
2. finding the local ecological factors that were driving species distribution at finer scales. In order to define local ecological factors driving species microhabitat, we used two complementary studies, one involving modelling and the other an experiment. In terms of modelling, we began to quantify the realised niche at the microclimatic scale for the 43 tree species in the reserve using existing geographic information system (GIS) techniques, fine-scale distributional surveys and the network of temperature and moisture recorders in the reserve. During the last months of 2007, we also initiated an out-planting experiment at 15 sites inside the Mont St. Hilaire so as to monitor the phenology of 20 species, 10 native and 10 exotic ones, in 2008. In 2009 we monitored the phenology of nine species, because of the insufficient number of individuals that had germinated in 2008 for the other 11 species. This experiment showed that germination requirements were key factors in limiting the distribution of species and that warming might locally favour the germination of species which were currently more southerly located, and thus probably improve their migration success in the next decades. We also carried out a greenhouse experiment to test the physiological tolerance to soil water stress, according to a drier climate scenario, in four tree species from Mont St. Hilaire. We found clear stress-tolerance differences across species, which could have a large influence on the spatial and temporal distributions of species under a drier climate.
3. forecasting predictions of species distribution changes with migration and colonisation constraints. We compared the prediction of the process-based model Phenofit with the widely-used habitat models for northeast America. A general pattern emerged from our comparisons. Niche-based models tended to predict a stronger level of extinctions and a greater proportion of colonisation than the process-based model. Such a finding highlighted the great relevance of using process-based models to make predictions on the fate of species under climate change in the near future and also called for taking habitat models predictions with more caution than currently done.
The theoretical framework we proposed to provide relevant ways to downscale existing distribution models at the regional level was detailed in a manuscript by the time of the project completion. This project thus led to theoretical advancements, better predictions with modelling approaches and experimental evidence on changes in species distribution in response to climate change.
The project was broken down into the following three parts:
1. identification of the factors limiting species range at global scale and implication for diversity. We firstly considered the distribution pattern of all North American tree species, i.e. of 599 species. Using observed range area, we showed evidence for a latitudinal gradient in species range size, with larger ranges in the north. We furthermore showed that tree species richness and mean species range area were negatively covariant across latitudes, suggesting the possibility that the causes of the increase in range area at higher latitudes might in turn also underpin the decreasing latitudinal trend in species diversity. Going deeper in the search of causal factors for this described patterns, we showed that the gradient in range size was consistent with the interspecific variation of thermal tolerance of trees in north America. In other words, this result highlighted the impact of climate on species distribution at continental scale.
2. finding the local ecological factors that were driving species distribution at finer scales. In order to define local ecological factors driving species microhabitat, we used two complementary studies, one involving modelling and the other an experiment. In terms of modelling, we began to quantify the realised niche at the microclimatic scale for the 43 tree species in the reserve using existing geographic information system (GIS) techniques, fine-scale distributional surveys and the network of temperature and moisture recorders in the reserve. During the last months of 2007, we also initiated an out-planting experiment at 15 sites inside the Mont St. Hilaire so as to monitor the phenology of 20 species, 10 native and 10 exotic ones, in 2008. In 2009 we monitored the phenology of nine species, because of the insufficient number of individuals that had germinated in 2008 for the other 11 species. This experiment showed that germination requirements were key factors in limiting the distribution of species and that warming might locally favour the germination of species which were currently more southerly located, and thus probably improve their migration success in the next decades. We also carried out a greenhouse experiment to test the physiological tolerance to soil water stress, according to a drier climate scenario, in four tree species from Mont St. Hilaire. We found clear stress-tolerance differences across species, which could have a large influence on the spatial and temporal distributions of species under a drier climate.
3. forecasting predictions of species distribution changes with migration and colonisation constraints. We compared the prediction of the process-based model Phenofit with the widely-used habitat models for northeast America. A general pattern emerged from our comparisons. Niche-based models tended to predict a stronger level of extinctions and a greater proportion of colonisation than the process-based model. Such a finding highlighted the great relevance of using process-based models to make predictions on the fate of species under climate change in the near future and also called for taking habitat models predictions with more caution than currently done.
The theoretical framework we proposed to provide relevant ways to downscale existing distribution models at the regional level was detailed in a manuscript by the time of the project completion. This project thus led to theoretical advancements, better predictions with modelling approaches and experimental evidence on changes in species distribution in response to climate change.