Final Report Summary - ECOCHANGE (Creating conditions for persistence of biodiversity in the face of climate change)
The Marie Curie Project 254206, entitled Creating conditions for persistence of biodiversity in the face of climate change - ECOCHANGE, had two main research aims, namely:
1) to investigate the potential of using the palaeo-record in order to improve our understanding of the impact of climate change on biodiversity, and
2) to study the effect of the physical characteristics of the landscape (geology, geomorphology) on the presence of vegetation at the local scale with the aim of better understanding the effect of climate change on fragmented populations that in many occasions constitute valuable refugia for many species in the face of adverse climatic conditions.
As explained in the project outline, these questions stemmed from the facts that a) in the last decade and half models have been developed predicting climate change-induced habitat losses and extinction of many plants and animals, but that b) these models are difficult to validate, and c) their coarse scale may have overestimated extinction threats/rates of change due to their inability to capture small-scale habitat heterogeneity.
Answering b) and c) would thus allow the scientific community to better assess the uncertainty of model predictions in a).
The models employed in ECOCHANGE are collectively known as Species Distribution Models (SDMs) and consist of a wide range of algorithms which seek to statistically link species occurrence data with environmental data (climatic data generally). These models have an enormous flexibility to link species and environmental data, and are widely used in the study of climate impacts on species distributions and extinction risks. Whereas they provide a very powerful tool, their 'blindness' to biological processes (they are after all statistical models) means that they might be sensitive to i) the type of algorithm used and ii) predicting into climate and environmental scenarios not accounted for in the training of the model. Finally, scale issues are known to affect SDMs' predictions (e.g. coarseness of a model might not capture small-scale processes important for the final response of a given species to a change in climate).
Answering b) - model validation
Proposed validations of SDMs use current data as calibration and past data as verification. The Quaternary fossil record is thus essential to understand the mechanisms that have allowed past species persistence, and to improve our forecasting ability regarding their future performance. Including species occurrence data in periods other than present allows researchers to expand their understanding of the responses that species have had in periods of past climatic changes.
In this line, ECOCHANGE started with a study of the applicability of SDMs to fossil presence data (i.e. pollen and macrofossil). This involved considering the temporal and spatial characteristics of fossil data and the selection of the optimal resolution at which modelling work was possible.
This was explored and successfully applied to a set of woody taxa at a European scale, resulting in the publication of a paper in Global Ecology and Biogeography with M. Macias-Fauria as a first author (Impact Factor 2012: 7.223). This study demonstrated that, when carefully selected, palaeo-ecological data can be used in conjunction with SDMs in order to track and better understand species responses to changes in climate even in relatively short time periods (centuries).
Answering c) - model sensitivity to scale
In order to study the effects of coarseness in SDMs, we conducted a high-resolution study case (10 m resolution) that incorporated not only smoothly-varying climatic parameters but also geomorphic and geologic characteristics of the landscape. We sought to find out what was the effect of local variables in the predicted upslope treeline advance (i.e. forest advancing to higher elevation in mountain systems) under a warmer climate scenario. Factors other than climate might strongly affect presence of vegetation on the landscape, especially at local scales, which are relevant for the presence of small populations that might be crucial in the survival of fragmented populations, potentially modifying SDMs' predictions.
This phase of the project aimed thus at defining the physical characteristics (climatic and non-climatic) that determine the survival of vegetation on cold environments, and to integrate them in species-envelope models in order to improve our understanding of the role of such variables to understand biotic response to past, current, and future climate changes.
A paper was published in Proceedings of the National Academy of Sciences of the U.S.A. out of this study, with M. Macias-Fauria as a first author (Impact Factor 2012: 9.737) which showed that geomorphic processes are at least as important as climate alone in determining the survival of plants in the sub-alpine forest of the Canadian Rockies, and challenged predictions of widespread treeline advance at a landscape scale. This paper reported that only 5% of the current treeline position in the study area is limited by low temperatures, thus breaking ground in this field, and urging to reconsider the current understanding of a rapid response of treeline position to rising temperatures.
The current dynamics of woody plants in the tundra of Northern Eurasia were also analysed and a study of their relationships with climate, sea ice, and landscape physical properties published in the journal Nature Climate Change with M. Macias-Fauria as a first author (Impact Factor 2012: 14.472). This paper showed a strong effect of summer temperature on the current trends of tundra shrub dynamics at a regional scale, and discussed the effect of other variables in these trends, concluding that sea ice will not affect much tundra dynamics in the near future do to its reduced influence on local tundra climate when far from the coast during the growing season (as it happens now with reduced summer sea ice extent). It also identified geomorphic dynamics related to permafrost as key in the expansion of tall shrubs into areas currently occupied by dwarf shrubs.
Further to the published record, a number of collaborations have arisen during the project with researchers over a range of countries (U.K. Finland, Sweden, Canada, Switzerland, Russia) with new questions that deepen in the topic under research and promise high-impact science in the coming years. In particular,
1) a study is nearly finished which seeks to assess the validity of a dynamic vegetation model for Northern Eurasia using pollen data as an independent validation set. This study is being performed in collaboration with the University of Southampton (U.K.) and the Swiss Federal Institute of Technology, Lausanne (Switzerland) and will be finished in the next few weeks.
2) a study just started that will seek to partition the influence of grazing (by reindeer) vs. climate (warmer temperatures) in northern Eurasian tundra vegetation dynamics. This project is currently being conducted in collaboration with the University of Umeå (Sweden) and the University of Lapland (Finland).
3) a five-year research project funded by the Academy of Finland (Resilience in Social-Ecological Systems of Northwest Eurasia - RISES) in which M. Macias-Fauria participates, which seeks to deepen in the understanding of tundra vegetation response to climate change, and on the use of palaeoecological data to expand this knowledge.
Apart from the interest raised by research conducted during the duration of ECOCHANGE, reported in the next section, results stemming from this project had several impacts on the scientific community and society as a whole, namely:
1. The palaeo-ecological record is an emerging and highly insightful resource to better inform conservation and land-management policies, and can give strong evidence for or against the convenience of concrete conservation of mitigation strategies.
2. Local factors such as geomorphic processes can completely override climate-induced species responses and thus need to be taken into account when planning for the future in (at least) cold ecosystems.
3. An expanded E.U.-based (but with members from non-E.U. states, such as Canada and Russia) network of research activity has been consolidated and promises to give more high-impact science in the coming years.
4. Ideas stemming from the project have resulted in the supervision and completion of two M.Sc. and the start of a Ph.D. project within the University of Oxford, U.K.
5. Likewise, ideas stemming from the project have resulted in M. Macias-Fauria giving a course at a M.Sc. level at the University of Oxford for two consecutive years (the third year being also confirmed) on SDMs which has been highly appreciated by students.
1) to investigate the potential of using the palaeo-record in order to improve our understanding of the impact of climate change on biodiversity, and
2) to study the effect of the physical characteristics of the landscape (geology, geomorphology) on the presence of vegetation at the local scale with the aim of better understanding the effect of climate change on fragmented populations that in many occasions constitute valuable refugia for many species in the face of adverse climatic conditions.
As explained in the project outline, these questions stemmed from the facts that a) in the last decade and half models have been developed predicting climate change-induced habitat losses and extinction of many plants and animals, but that b) these models are difficult to validate, and c) their coarse scale may have overestimated extinction threats/rates of change due to their inability to capture small-scale habitat heterogeneity.
Answering b) and c) would thus allow the scientific community to better assess the uncertainty of model predictions in a).
The models employed in ECOCHANGE are collectively known as Species Distribution Models (SDMs) and consist of a wide range of algorithms which seek to statistically link species occurrence data with environmental data (climatic data generally). These models have an enormous flexibility to link species and environmental data, and are widely used in the study of climate impacts on species distributions and extinction risks. Whereas they provide a very powerful tool, their 'blindness' to biological processes (they are after all statistical models) means that they might be sensitive to i) the type of algorithm used and ii) predicting into climate and environmental scenarios not accounted for in the training of the model. Finally, scale issues are known to affect SDMs' predictions (e.g. coarseness of a model might not capture small-scale processes important for the final response of a given species to a change in climate).
Answering b) - model validation
Proposed validations of SDMs use current data as calibration and past data as verification. The Quaternary fossil record is thus essential to understand the mechanisms that have allowed past species persistence, and to improve our forecasting ability regarding their future performance. Including species occurrence data in periods other than present allows researchers to expand their understanding of the responses that species have had in periods of past climatic changes.
In this line, ECOCHANGE started with a study of the applicability of SDMs to fossil presence data (i.e. pollen and macrofossil). This involved considering the temporal and spatial characteristics of fossil data and the selection of the optimal resolution at which modelling work was possible.
This was explored and successfully applied to a set of woody taxa at a European scale, resulting in the publication of a paper in Global Ecology and Biogeography with M. Macias-Fauria as a first author (Impact Factor 2012: 7.223). This study demonstrated that, when carefully selected, palaeo-ecological data can be used in conjunction with SDMs in order to track and better understand species responses to changes in climate even in relatively short time periods (centuries).
Answering c) - model sensitivity to scale
In order to study the effects of coarseness in SDMs, we conducted a high-resolution study case (10 m resolution) that incorporated not only smoothly-varying climatic parameters but also geomorphic and geologic characteristics of the landscape. We sought to find out what was the effect of local variables in the predicted upslope treeline advance (i.e. forest advancing to higher elevation in mountain systems) under a warmer climate scenario. Factors other than climate might strongly affect presence of vegetation on the landscape, especially at local scales, which are relevant for the presence of small populations that might be crucial in the survival of fragmented populations, potentially modifying SDMs' predictions.
This phase of the project aimed thus at defining the physical characteristics (climatic and non-climatic) that determine the survival of vegetation on cold environments, and to integrate them in species-envelope models in order to improve our understanding of the role of such variables to understand biotic response to past, current, and future climate changes.
A paper was published in Proceedings of the National Academy of Sciences of the U.S.A. out of this study, with M. Macias-Fauria as a first author (Impact Factor 2012: 9.737) which showed that geomorphic processes are at least as important as climate alone in determining the survival of plants in the sub-alpine forest of the Canadian Rockies, and challenged predictions of widespread treeline advance at a landscape scale. This paper reported that only 5% of the current treeline position in the study area is limited by low temperatures, thus breaking ground in this field, and urging to reconsider the current understanding of a rapid response of treeline position to rising temperatures.
The current dynamics of woody plants in the tundra of Northern Eurasia were also analysed and a study of their relationships with climate, sea ice, and landscape physical properties published in the journal Nature Climate Change with M. Macias-Fauria as a first author (Impact Factor 2012: 14.472). This paper showed a strong effect of summer temperature on the current trends of tundra shrub dynamics at a regional scale, and discussed the effect of other variables in these trends, concluding that sea ice will not affect much tundra dynamics in the near future do to its reduced influence on local tundra climate when far from the coast during the growing season (as it happens now with reduced summer sea ice extent). It also identified geomorphic dynamics related to permafrost as key in the expansion of tall shrubs into areas currently occupied by dwarf shrubs.
Further to the published record, a number of collaborations have arisen during the project with researchers over a range of countries (U.K. Finland, Sweden, Canada, Switzerland, Russia) with new questions that deepen in the topic under research and promise high-impact science in the coming years. In particular,
1) a study is nearly finished which seeks to assess the validity of a dynamic vegetation model for Northern Eurasia using pollen data as an independent validation set. This study is being performed in collaboration with the University of Southampton (U.K.) and the Swiss Federal Institute of Technology, Lausanne (Switzerland) and will be finished in the next few weeks.
2) a study just started that will seek to partition the influence of grazing (by reindeer) vs. climate (warmer temperatures) in northern Eurasian tundra vegetation dynamics. This project is currently being conducted in collaboration with the University of Umeå (Sweden) and the University of Lapland (Finland).
3) a five-year research project funded by the Academy of Finland (Resilience in Social-Ecological Systems of Northwest Eurasia - RISES) in which M. Macias-Fauria participates, which seeks to deepen in the understanding of tundra vegetation response to climate change, and on the use of palaeoecological data to expand this knowledge.
Apart from the interest raised by research conducted during the duration of ECOCHANGE, reported in the next section, results stemming from this project had several impacts on the scientific community and society as a whole, namely:
1. The palaeo-ecological record is an emerging and highly insightful resource to better inform conservation and land-management policies, and can give strong evidence for or against the convenience of concrete conservation of mitigation strategies.
2. Local factors such as geomorphic processes can completely override climate-induced species responses and thus need to be taken into account when planning for the future in (at least) cold ecosystems.
3. An expanded E.U.-based (but with members from non-E.U. states, such as Canada and Russia) network of research activity has been consolidated and promises to give more high-impact science in the coming years.
4. Ideas stemming from the project have resulted in the supervision and completion of two M.Sc. and the start of a Ph.D. project within the University of Oxford, U.K.
5. Likewise, ideas stemming from the project have resulted in M. Macias-Fauria giving a course at a M.Sc. level at the University of Oxford for two consecutive years (the third year being also confirmed) on SDMs which has been highly appreciated by students.