Estimated Extinction Risks of Species by 2100 due to Climate Change

We are currently facing high extinction risks in the near future, and it is becoming essential to understand how and why. First, however, we have to understand the amount of extinction we might be facing. Maclean and Wilson (2011) make these first steps by taking 318 of the top scientific papers on biodiversity loss and calibrating extinction risks, using the International Union for Conservation of Nature’s criteria. Their findings suggest declines of up to 14% of species by 2100. Their results also uncover a bias in research; the lack of data from tropical regions of the world, and non-coral marine ecosystems are contributing to incorrect estimates of extinction risks. There is also a gap in our understanding of the effects of climate change on invertebrates, particularly insects. Maclean and Wilson provide a start for quantitatively understanding the risks we are facing in biodiversity loss, and make clear the need for more work to be done. –Mathew Harreld
Maclean, I. and Wilson R. 2011. Recent ecological responses to climate change support predictions of high extinction risk. PNAS Early Edition: 1–6.

          The recent studies on the effects of climate change on biodiversity appear to be pointing toward great losses of species, some going as far as predicting the next great mass extinction. Many studies released are citing the losses of a particular species that are predicted to occur 2050 and 2100. Other papers are focused on the observable changes in species’ environments and their decline rates. But and how accurate are they? This is the question Maclean and Wilson set out to answer.
          The authors took the top ten scientific journals dealing with effects of climate change, choosing 130 papers that covered observed and ecological responses of species to climate change and 188 papers that covered predicted ones. The papers thus contained data on extinction risks, population changes, and geographic range changes of 305 different species. MacLean and Wilson then used the International Union for Conservation of Nature (IUCN) Red List Criteria to determine estimates for extinction risks for each species. Using the IUCN method gives universal and comparable data, rather than the individual studies’ own derived answers. The authors took into account a variety of possible biases from each individual study, and then compared the results by averaging them across all studies. To remove the bias of closely related species appearing to act similarly because of genetics, the authors created a phylogenetic tree and averaged out the branches and tips to remove any possible discrepancy. The model also took into account spatial patterns in extinctions to avoid a bias in the studies done in regions where species are more at risk of climate change. And lastly, the authors broke up the studies by taxonomic groups (plant, invertebrate, and vertebrate), as well as major ecoregions (i.e. polar, temperate, tropical, marine).
          The model utilized research based both on predicted mortality and observed mortality; for studies in which the mortality was based on predictions, 7.69% of species would be threatened by 2100, and for mortalities based on observations, 37.1% would be threatened by 2100. Species more than likely to go extinct by 2100 are 1.9% based on predictions and 12.0% based on observations. The large range between the predicted and observed extinctions is explained by the where the study was done and on what species were observed. A majority of the observational papers were done on land, focusing on rainfall and temperature changes, whereas the prediction papers were done on sea ice changes and ocean circulation changes which produced much higher extinction risks, but there were many fewer of these papers. The authors suggest more studies done on oceanic circulation and acidity changes would result in much better understanding of possible extinction risks. The results also suggest that there seems to be a slight exaggeration on observational data because of a focus on threatened species. The adjustment for spatial patterns in extinctions increased the predicted risk to 10.3%, compared to a 13.9% in observed after adjustments to phylogenetic independence, suggesting that current models are not fully taking into account regions under higher levels of threat to climate change.
          The results for the ecoregions suggest that species at higher latitudes are most threatened, as are marine species. The authors also suggest that in the marine environment there is a bias to study to corals, which might be greatly increasing the marine extinction risk. The overall threat of biodiversity changes due to climate change would be better understood by more research in the tropics, as this is where most species exist, and where they seem least threatened by climate change.
          Though these results shed much light into the risk biodiversity faces today, it is still important to remember that they are simply estimates that need to be worked on further. First, the IUCN model for categorizing extinction risk is not perfect, as it is extremely difficult to determine the thresholds between criteria, especially when trying to apply it to all species. Also, there was a low risk calculated for invertebrates, but this is most likely due to a lack of knowledge. Most of the data recorded for invertebrate extinction risk came from Lepidoptera, and little research has been done on other insects and the effect climate change might have on them. The results of this paper provide more evidence that anthropologic climate change is one of the major factors leading to extinctions, furthering our need in understanding our impact on our planet. Maclean and Wilson clearly quantify the risk of biodiversity loss, while also highlighting areas of further research. In order to have a better understanding of the loss of biodiversity we might face there must be more studies done on less at-risk species, marine species, subtropical and tropical areas, and invertebrates.
Redesigning and Improving Climate Change Models to Better Show the Impact on Global Environments and Species
          Our understanding of how climate change might affect our planet, the ecosystems on it, and the species within them is dependent on computerized climate ecological models. The accuracy of these models is constantly being improved, and here McMahon et al. (2011) suggest further fixes that could be made to our current modeling programs. The models are dependent on the data gathered in the field, and therefore the authors suggest a unified method of data gathering, as well as a single database to access all the currently gathered data. From there tweaks, adjustments, and wholesale changes need to be made to various models. Some models will benefit from the greater level of accessible and new field data, and others need to be redesigned. Often a model needs to be made larger and more specific, allowing scientists to more accurately take into account all of the potential shifts and changes of climate, the environment, and of species and their interactions. McMahon et al. highlight the five largest gaps in our current models, suggesting how to fix them and how it will benefit us now and in the future. –Mathew Harreld
McMahon, S., Harrison, S., Armbruster, S.W., Bartlein, P., Beale, C., Edwards, M., Kattge, J., Midgley G., Morin, X., Prentice I.C., 2011. Improving assessment and modelling of climate change impacts on global terrestrial biodiversity. Trends in Ecology and Evolution 26, 249–259.
          Species and ecosystems response to climate change is key to our understanding to our environments, and to formulating the best conservation efforts for those environments. Discovering and understanding these responses can only be accomplished through modeling. Modeling allows for simulations of species and ecosystem responses to various changes of climate, but the models are far from perfect. As the authors point out, however, models are constantly being updated and reinvented to improve results. Accurate models would allow a greater understanding of earth systems, and how climate change might affect them. McMahon et al. discuss five gaps and ten potential fixes to modeling, as well as creating a internet-based connection between all scientists, so models can be shared and updated seamlessly.
          The first gap identified by McMahon et al. is the necessity to improve global biodiversity monitoring. The two goals of biodiversity monitoring, as specified by the authors, are to create a baseline of data for natural species levels and estimate the rate of change in biodiversity, and to determine the causes behind the changes. The authors then suggest preexisting biodiversity monitoring sites in Europe and North America, specifically The Smithsonian Institution Global Earth Observatory, as models for future monitoring sites. There is a need for more monitoring sites, but more importantly there is a need for these sites to work together to target potentially at risk areas, and then share these data across the globe. This requires a standardization of procedures to create a simplified system of data sharing amongst researches..
          The second gap highlighted is the ability to quantify how sensitive a species is to climate change. Climate can greatly affect species in many ways; climate change could result in extinctions, migrations, range contractions or expansions, or various other effects. To understand and predict these responses to climate change, models must be used. Traditionally scientists have relied on climate envelope models (CEMs) to calculate these changes, but McMahon et al. believe a new approach may be required. The limitation of CEMs is that they treat species as non-adaptive to the changes of other species, as well as disregarding temporal climate variability and CO2 fertilization. Therefore McMahon et al. suggest using process-based models to derive these predictions in the future. The main limitation of these models is that they require more data about individual species, and more data on how they interact with other species in their ecosystem.
          How full communities of species and how biodiversity as a whole will be affected by climate change is a weak spot in the modeling as well. Mahon et al. suggest that in order to model how climate change will affect species-species relationships, modelers must look at the palaeoecological evidence. A new model that combines life-history changes over time, using the palaeoecological evidence, and how they are affected by different climates may allow for a better understanding to how species relationships will change under a changing environment.
          The forth gap highlighted by the authors is how to model the influence of genetic variability and adaptive tendencies of species in climate change. Sexually reproducing species have a great variety of genetic information, and this allows for a great variation of possible responses to environmental changes. The ability to adapt to a rapid climate shift is dependent on the plasticity of species, which is related to the range of phenotypes that a species can generate from one genotype, and the greater the plasticity the greater the species’ ability to adapt to climate changes. This generic variation could have a huge affect on ecosystems and which species survive climate changes, yet this information is rarely used in the building of models. Mahon et al. suggest using already-established data, along with calculated data and laboratory tests of species under specific climate stresses, to develop a unified information base that can be applied in future models. The authors note that there is some use of mathematical integration for population genetics and phylogenetics to determine the short-term and long-term responses of species and populations to climate change.
          The final gap stressed by the authors is the necessity to improve the way global models define groupings of plant species. Models currently generate vegetation shifts by using plant functional types, which is a grouping of similar plant species that will most likely adapt in a similar manner. However, using plant functional types comes at the cost of generalized, and sometimes useless data. The authors suggests designing plant functional types around species that are known to be key in ecological function or are extremely responsive to certain climate changes. There is a need for more robust groupings to establish more concrete results, however there also needs to be a way to successfully quantify these data. In the past there was an issue with data availability, but now the global trait databases (GLOPNET) allow access to increasing amounts of data. Further studies will have to be done on plant traits and how they relate to key environmental factors, as well as expanding the current databases of plant traits. McMahon et al. suggest the development of a new rule set for plant functional types that requires the trade-offs of investing in certain key traits that may play important roles in climate change.
          McMahon et al. believe that in order to make further progress in the field of climate change modeling there needs to be a greater unification of scientists, globally. They suggest the development of an internet based “repository,” allowing instant access to new data on biodiversity from monitoring sites around the globe. Such an access point should go a long way toward advancing the accuracy of current and future models. McMahon et al. believe that with greater connectivity between the scientific community, and the interlacing of already established models, there can be a great increase in the quality of climate change modeling.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s