Mapping fragile areas in ecosystems using only short-term spatial data

Authors: GENIN, ALEXANDRE - University Of Montpellier; LEE, STEVEN - Us Geological Survey (USGS); BERLOW, ERIC - Consultant; Ostoja, Steven; KEFI, SONIA - University Of Montpellier

Submitted to: Biological Conservation
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/15/2020
Publication Date: 1/15/2020
Citation: Genin, A., Lee, S., Berlow, E., Ostoja, S.M., Kefi, S. 2020. Mapping fragile areas in ecosystems using only short-term spatial data. Biological Conservation. 241. https://doi.org/10.1016/j.biocon.2019.108388.
DOI: https://doi.org/10.1016/j.biocon.2019.108388

Interpretive Summary: Global change has presented conditions that are resulting in an increase in the frequency and intensity of extreme conditions, such as sustained and record setting droughts. These extreme events hold the potential to pose real harm to the condition and integrity of ecologically important ecosystems. Effective conservation requires we understand where (in space) on the landscape our most fragile ecosystems are located so that potential management interventions can be deployed that might hold the potential to ameliorate the negative consequences of extreme, system changing events. Using data collected just once we use a simple statistical model to assess how different plant species group, or separate out, according to differences in factors like soil moisture and local disturbance patterns. These factors are suggested to be good proxies for how changes in plant communities are organized are influenced, or re organized due to factors associated with global change.

Technical Abstract: Effective conservation requires prioritizing areas most vulnerable to large, irreversible change, versus ones that are more resilient. Generally, this requires experiments and long-term monitoring to assess how different ecosystem components might differentially may respond to future stressors. Such approaches come at significant costs and for many ecosystems the available data is only short-term and spatial. This paper offers an approach to leverage datasets to anticipate how a given area may respond to future perturbations inform conservation and management decisions. Using a simple model that uses the presence of well-separated discrete states in observations (i.e. multi-modality) as an approach to identify environmental conditions under which perturbations can have an unexpectedly large and/or irreversible effect on a given ecosystem. We then apply these insights to iconic high-elevation protected areas, the subalpine meadows of Sierra-Nevada (California) Wilderness. Changes in modality Our analysis of spatial vegetation data identified which specific areas within wilderness meadows that maybe more fragile than others. In these areas, because of interactions with local water regimes and/or upper soil conditions, the plant communities may shift, potentially durably, between one of several discrete states. These areas can be mapped to identify spatial hotpsots of fragility in protected areas. This work provides a simple workflow that uses commonly available spatial, short-term data, to help prioritize areas of conservation and research based on the potential sensitivity to perturbations. Such an approach could be very valuable to make informed decisions in the current context of fast ongoing global and environmental changes.