Current Research

How to build resilience to climate change in global terrestrial protected areas

Protected areas are critical for biodiversity conservation, but their static boundaries may make them vulnerable to climate change. This is because biodiversity is often governed by climatic factors; as the climate changes, species may track shifting climates to retain the climatic environment they’re adapted to.

But what happens if that suitable climate that is currently situated in protected areas shifts to areas that are unprotected? Species may need to shift their distributions outside of reserves to track it, and become more vulnerable to other threats from human activities that commonly occur outside of protected areas.

Our new paper in Science Advances looks at this issue in detail for global terrestrial protected areas. We first analyze what climates are present within the boundaries of protected areas, compared to the total available climate. Our first discovery is that there are biases in which parts of ‘climate space’ (the full set of temperature and precipitation combinations, such as warm-wet, warm-dry, cold-wet, cold-dry, and everything in between) protected areas are situated:

Fig. 1
The global outlook for current and future protected climate space.

Globally, protected areas tend to be concentrated in only small parts of this available climate space, particularly in cold-dry and temperate-wet climates. This clumping of protected areas means that as the climate warms, the future climate within protected areas will be vastly different, which suggests that the species found within protected areas may also be different in the future. This could threaten a lot of species that are unable to find suitable climates within protected areas in the future.

So what can be done to increase the degree to which climates that are currently found within protected areas continue to be represented by protected area networks (referred to as ‘protection retention’)?

We were surprised to find that simply adding protection to a PA network didn’t increase protection retention. We think this is because if you add more PAs following past trends, you just end up adding more protected land in the same biased fashion (i.e., into the same portion of climate space). Climate change pushes those suitable climates out of reserves and into unprotected lands, and climates that were formerly not under protection now enter the PA reserve. So you have minimal retention.

What we found that worked well was having a PA network that covered a wide range of climates. Then, as the climate shifts out of one PA, it may shift into another, and thus still be represented.

We compared this adaptation strategy to a mitigation strategy that focuses on reducing climate emissions to slow the pace of climate change. Slowing climate change increases retention a lot, because not as much of the currently protected climates shift outside of the reserve network. But, surprisingly, the adaptation strategy was found to be as effective as, or perhaps even more effective than, the mitigation strategy, in terms of increasing protection retention. Importantly, these two strategies are completely compatible, meaning country governments can employ both simultaneously for maximum conservation benefit.

We know protected areas are important cornerstones for biodiversity conservation, but they need to be made more resilient given the current threats of climate change. This means reimagining how PA networks should be designed. Our results suggest that PA networks that span wide climatic gradients will be more resilient in the future, so it’s essential to take this into consideration in future PA planning.

Related publications:

Elsen, P. R., Monahan, W. B., Dougherty, E. R. & Merenlender, A. M. 2020. Keeping pace with climate change in global terrestrial protected areas. Science Advances 6, eaay0814. Link PDF

See media related to this work:

UC Berkeley College of Natural Resources: Protected areas need across climates to safeguard biodiversity

WCS Newsroom: Ensuring protected areas cover a range of climatic conditions is more effective for safeguarding biodiversity than simply expanding protection


How will topography and human pressure interact to affect biodiversity under climate change?

Our new paper published in Nature Communications seeks to answer this question. The surprising result is that as species shift their ranges under climate change in many mountain ranges around the world, they move out of areas of intense human pressure and into areas with more intact land area, relative to where they started. Sometimes this is helped even more by the fact that species can find more total land area as they shift to higher elevations, given complex topography and a diversity of mountain shapes in the world.

Our results have important conservation implications. First, we often think of mountain ranges as being isolated, remote landscapes and thus devoid of much human pressure. We were shocked to find that nearly 60% of all mountainous area was classified as under intense human pressure: this will undoubtedly put a strain on mountain biodiversity. So implication #1 is that mountains may need additional protection.

If that’s the case, where should we protect in mountains? Another interesting finding from our study was that most human pressure is located at low elevations and at the bases of mountains, in other words, the areas most hospitable and productive for humans. Implication #2 is that there isn’t much intact land area left at low elevations, so protecting what’s left is likely very important to help species persist, especially those that are poor dispersers. Implication #3 is that, given there isn’t much intact land area left at low elevations, we probably need to focus restoration efforts to these elevational zones. This is likely particularly important to enable species to shift their ranges upslope. It doesn’t help that there is more intact land area available at higher elevations if species can’t access them.

Finally, if there is more intact land area available at higher elevations, and this are the areas models predict species are shifting into, implication #4 is that we have a good opportunity to proactively protect intact landscapes that will be increasingly important for species in the coming decades. In the immediate to short-term, protecting these areas is also likely to capture a lot of endemic biodiversity, since endemism tends to increase with elevation for most taxa.

While our results offer a glimmer of hope for montane species, conserving mountain biodiversity under global change will require bold actions from governments and land managers to commit to international policies and achieving conservation targets.

Map of global mountain ranges shows the proportion of the elevational gradient for each range where percentage of change in intact area equals or exceeds the percentage of change in total area following range shifts for a suite of modeled hypothetical montane species. Insets for select mountain ranges show mean and standard error percentage of changes in area across all modeled species in two cases, one where species are allowed to occupy all land area (blue lines ΔAreatotal), and one where species only occupy intact land area (red lines ΔAreaintact).

Related publications:

Elsen, P. R., Monahan, W. B. & Merenlender, A. M. 2020. Topography and human pressure in mountain ranges alter expected species responses to climate change. Nature Communications 11, 1-11. Link PDF

See media related to this work:

Read a press release from the Wildlife Conservation Society.

Spatial patterns of temperature predict continental-scale bird richness patterns during winter

Paper out now in Remote Sensing of Environment is the first to use thermal satellite imagery from Landsat 8 to map spatial patterns of temperature to predict broad-scale species richness patterns. We develop new metrics of relative temperature and thermal heterogeneity–the spatial patterning of temperature–across the conterminous US using the complete collection of thermal imagery from the Landsat 8 Thermal Infrared Sensor.

Our approach captures fine-scale and localized thermal dynamics (see Figure). For example, it captures warm and cool areas of the Great Plains, temperature gradients along the Rocky Mountains and in the Santa Catalina Mountains near Tucson, and individual crop fields with varying thermal environments in the Central Valley. Thermal heterogeneity was also well captured, and was greatest in topographically complex regions and also around city blocks in urban areas and at the edges of fields.

Relative temperature (a) and thermal heterogeneity (b) across the conterminous United States with zoomed insets for the Great Plains in Iowa (c, g); the southern Rocky Mountains (d, h); Tucson, AZ and the surrounding Santa Catalina Mountains (e, i); and fields in the Central Valley near Fresno, CA

We used these data to predict bird species richness patterns during winter, when most species face pronounced thermal stress and may rely on thermal heterogeneity as refugia. Indeed, we found overall bird richness patterns to be positively related to both our thermal predictors–there were more bird species found in areas that were warmer and more thermally heterogeneous. This was true when controlling for confounding factors, such as elevation, terrain ruggedness, and different land cover types. And the patterns were stronger for species that were more sensitive to the thermal environment–small-bodied species (which are poor thermoregulators) and species sensitive to climate change showed particularly positive relationships with our thermal variables.

Our study has important implications for biodiversity conservation. In winter, thermally heterogeneous areas offer important benefits for species that enable more species to occur. This makes prioritizing thermally heterogeneous areas for protection an important part of a conservation strategy. More generally, we were surprised to discover that our study represents the first to use freely available thermal satellite imagery at 30 m resolution for biodiversity studies. These data are globally available, which will enable future studies to investigate similar issues in other areas and for other taxonomic groups.

Related publications:

Elsen, P. R., Farwell, L. S., Pidgeon, A. M., & Radeloff, V. C. 2020. Landsat 8 TIRS-derived relative temperature and thermal heterogeneity predict winter bird species richness patterns across the conterminous United States. Remote Sensing of Environment 236, 111514. Link PDF

Download the data:

Data from our paper are now available to view and download here:

See media related to this work:

See a video produced by NASA’s Goddard Space Flight Center featuring our paper and describing the importance of thermal heterogeneity for bird diversity in the United States:

Why are some species sensitive to forest loss, while others aren’t?

A new paper, published in Ecography, shows that species’ sensitivity to land-use change is influenced by their sensitivity to variation in temperature. Led by Umesh Srinivasan, our study uses observations of birds in forest and agriculture at two ends of the Himalayas spanning a gradient in annual temperature variation: the western Himalayas experience twice the annual temperature variation as the eastern Himalayas. We found that birds that are adapted to greater temperature variation were found to be less reliant on forests, and could tolerate agricultural lands to a much larger degree. By contrast, eastern Himalayan birds, which are adapted to relatively stable climates, were much more dependent on forests, being found in agriculture far less than their western counterparts. This was even true for populations of the same species, with western populations being less sensitive to forest loss than eastern populations.

Forests converted to mixed agriculture in the western Himalayas

When forests are converted to agriculture, the thermal environment changes.  On average, agricultural lands in the tropics can be nearly 8°C warmer than nearby forests. Forests also buffer against temperature fluctuations and extremes. Conversion in the eastern Himalayas would create thermal conditions that birds aren’t generally exposed to, whereas conversion in the western Himalayas would create thermal conditions that are fairly similar to the temperature variation species experience on an annual basis. Using metrics of species thermal sensitivity, we showed that western species’ adaptation to this thermal environment influences their sensitivity to forest conversion.

Our results advance our knowledge of what makes species sensitive to land-use change, and have important implications for conservation: species in tropical regions, adapted to stable climates, are likely more threatened by habitat loss, especially of forests. These regions would benefit most from strict habitat protection. On the other hand, temperate regions or regions experience greater seasonality, will likely contain species more tolerant of habitat loss. These regions might benefit most from a combination of habitat protection, restoration, and managing agricultural lands for biodiversity.

Related publications:

Srinivasan, U., Elsen, P. R. & Wilcove, D. S. 2019. Annual temperature variation influences the vulnerability of montane species to land-use change. Ecography 42, 2084-2094. Link PDF

See media related to this work:

Study helps pinpoint what makes species vulnerable to environmental change – Princeton Environmental Institute

View recent Weston Roundtable Lecture:

“The global outlook for conserving mountain biodiversity under land-use and climate change”

Paper out now in Conservation Biology seeking to address the habitat needs of Himalayan birds throughout their annual cycle. Most Himalayan birds migrate elevationally in spring and fall and are increasingly encountering agricultural landscapes in the process. Our study quantifies the impacts of mixed agriculture and pasture on Himalayan birds during winter and the breeding season to inform comprehensive conservation strategies.

Winter in western Himalayan temperate conifer forests

Surprisingly, we found that the value of agriculture and primary forest changes seasonally: mixed agricultural mosaics are particularly important for bird communities during winter, and primary forests are particularly important during the breeding season. Converting either forest or agriculture to pasture would be detrimental to Himalayan bird communities in any season, but would disproportionately affect breeding birds.

Our results suggest that effective conservation in the Himalayas must do two things: protect intact primary forests for their high value to breeding birds, and protect agriculture from being converted to pasture for their high value to wintering birds. In the existing context of land-sparing and land-sharing, our results suggest that neither approach on its own may be adequate in highly seasonal environments where species have different habitat requirements, diets, and life history strategies seasonally, as is the case with many migratory species. Conservation strategies in such seasonal environments should integrate tenets from both approaches to be successful.

Related publications:

Elsen, P. R., Ramesh, K. & Wilcove, D. S. 2018. Conserving Himalayan birds in highly seasonal forested and agricultural landscapes. Conservation Biology 32, 1313-1324. Link PDF

See media related to this work:

Himalayan birds throng agricultural lands in winter, but many breed in forests – Mongabay India

Cover image of PNAS Vol 115, Issue 23, 2018

New paper featured on the cover of Proceedings of the National Academy of Sciences quantifying the protection of elevational gradients in mountain ranges worldwide. Mountains are biodiversity hotspots, containing exceptionally high numbers of plant and animal species for their extent. Protecting elevational gradients with mountain ranges is essential to capturing this full suite of diversity and also to facilitate range shifts as species track temperatures up and down mountainsides following climate change.

We combined high resolution digital elevation models with maps of protected areas to determine how protection is distributed across elevation for over 1,000 mountain ranges globally. Although mountains are generally considered well-protected, we found that nearly 40% of mountain ranges contained no strictly protected areas, and that roughly half of all mountain ranges fail to meet current conservation targets (17% land protection) at any elevation. We develop optimal models of protection and show that strategic conservation planning is important to reduce the amount of additional land area required under protection to meet these targets for all elevations. Otherwise, we would need to protect roughly half of all mountainous area to achieve this goal.


Map of the proportion of all elevations meeting current conservation targets (17% land protected) for 1,010 mountain ranges worldwide

Related publications:

Elsen, P. R., Monahan, W. B. & Merenlender, A. M. 2018. Global patterns of protection of elevational gradients in mountain ranges. Proceedings of the National Academy of Sciences USA 115, 6004-6009. Link PDF

See media and news related to this work:

Unprotected mountains more vulnerable to climate change – UC Berkeley Dept. of Environmental Science, Policy, and Management

Research highlight in California Agriculture

Response article published in PNAS: Reply to You et al.: The World Database on Protected Areas is an invaluable resource for global conservation assessments and planning.

The spatial and temporal domains of modern ecology, published in Nature Ecology and Evolutionis a new study led by collaborator Lyndon Estes that endeavors to reveal the observational extents of studies across the discipline of ecology, in both space and time. We find that while ecologists have made some advances in studying larger extents and finer resolutions, there are still significant gaps in what we observe and describe. Most observations are at resolutions of less than one square meter and with extents less than 10,000 ha. Ecologists also tend to make repeat observations infrequently and study processes for less than a year. Technologies, such as remote sensing and unmanned aerial vehicles, can pave the way for expanding observational extents and drilling down to finer resolutions, while also potentially doing so at more frequent intervals. Filling in the observation gaps can help ecologists understand more about the processes they aim to study.


The observational domains of modern ecology

Related publications:

Estes, L., Elsen, P. R., Treuer, T., Ahmed, L., Caylor, K., Chang, J., Choi, J. J., & Ellis, E. C. 2018. The spatial and temporal domains of modern ecology. Nature Ecology & Evolution 2, 819-826. Link PDF

See media and news related to this work:

Check out a related News and Views article that further highlights some of the important findings, as well as a blog post written by Lyndon Estes and Tim Treuer, describing the ideas that led to the study, and the process of reviewing and scoring hundreds of academic papers to uncover the scales of ecological observations.

New paper in Proceedings of the Royal Society B on determining how the influence of temperature and competition on bird community structure varies across seasonality and species richness gradients in the Himalayas. We surveyed bird communities during breeding and winter along opposing seasonality and bird species richness gradients in two regions of the Himalayas separated by 1500 km. We deployed temperature loggers to link each bird observation with temperature, and used this to create season-specific thermal niches for 120 bird species. This enabled us to look at the degree to which species track temperatures across seasons (a measure of the abiotic influence) and to look at the degree to which potentially competing species segregate in thermal space (a measure of the biotic influence).

Species in less seasonal environments were more constrained by temperature than species in seasonal environments, suggesting they may be more sensitive to further changes in temperature. Surprisingly, we found that seasonality influenced the strength of competition in both regions, but in context specific ways. Competition was intensified during winter in the more seasonal environment across species; in the less seasonal environment during winter, competition intensified between species with similar ecologies (measured by body size). Our results demonstrate that abiotic and biotic factors interact to structure bird elevational ranges and entire communities, and point to both heightened thermal sensitivity and enhanced competition playing particularly important roles in structuring ecological communities in tropical regions.

Related publications:

Srinivasan, U.*, Elsen, P. R.*, Tingley, M. W., & Wilcove, D. S. 2018. Temperature and competition interact to structure Himalayan bird communities. Proceedings of the Royal Society B, 285, 20172593. Link PDF

See media and news related to this work:

Mongabay India



New paper featured on the cover of Indian Birds – read a new paper in Indian Birds describing the distribution of White-browed Shortwing in the western Himalayas, lead by Gunjan Arora. The White-browed Shortwing has a disjunct range, with very few records west of Nepal. I observed and recorded the song of this species in Sarmoli village near Munsiyari, Uttarakhand in May 2014, one of only a few records in the state.



Research featured on the cover of the February 2017 issue of Ecology

Caption: Mixed coniferous forest at 2900 m in Great Himalayan National Park, India, in December 2012. Stark habitat transitions, such as those between mixed coniferous forest (foreground) and upper temperate forest (seen in background), limit the elevational distributions of many Himalayan birds, but temperature dominates as the primary range-limiting factor. See Elsen et al. in this issue (doi: 10.1002/ecy.1669). Photo credit: Paul R. Elsen

Disentangling the factors limiting species distributions

A fundamental yet unresolved question in ecology is why species occur where they do, and not elsewhere. I explore this question utilizing unique natural climatic and species richness gradients in the Himalayas. Through fieldwork and abundance modeling, my research attempts to disentangle three leading hypotheses – climate, habitat, and

Treeline above mixed Rhododendron and upper temperate forests, Uttarakhand, western Himalayas

competition – thought to limit bird elevational ranges. Results from this work have revealed surprising differences in Himalayan bird communities compared to previous foundational work in the tropics. Whereas the majority of tropical bird species are thought to be primarily limited by competitive interactions, the majority of birds from the temperate Himalayas are primarily limited by temperature. These results suggest that contrasting mechanisms may be limiting species across latitudinal gradients. Furthermore, my findings underscore a significant sensitivity of Himalayan birds to temperature that suggests their distributions may be substantially influenced by ongoing climate change.

Related Publications:

Elsen, P. R., Tingley, M. W., Kalyanaraman, R., Ramesh, K. & Wilcove, D. S. 2017. The role of competition, ecotones, and temperature in the elevational distribution of Himalayan birds. Ecology 98, 337-348. Link PDF

See media and news related to this work:

Princeton University Press Release



Determining the impacts of agriculture and grazing on Himalayan birds

The Himalayas are a global biodiversity hotspot containing nearly 10% of the world’s birds. At the same time, the region has undergone rapid deforestation, largely driven by the expansion and intensification of agriculture and grazing. My research attempts to understand how the conversion of natural forests to agriculture impacts Himalayan bird communities in terms of abundance, species richness, and community composition, as well as understand the ultimate drivers of species loss. My results from fieldwork studying bird communities across a gradient of agricultural intensity suggest that low intensity ag lands (consisting of small-scale subsistence agriculture within a matrix of lightly used forest) as well as medium intensity at lands (consisting of mixed terraced agriculture)

Terraced mixed agriculture in Himachal Pradesh, western Himalayas

harbor significantly greater abundance and bird species richness than high intensity ag lands (heavily grazed pastures), but also greater than primary forest. Surprisingly, this finding was true both during winter, when forest resources are scarce and birds migrate to lower elevations into human-dominated landscapes, as well as during summer, when species are breeding. However, primary forests harbored unique species during both seasons, and had a distinct community composition compared to agricultural lands. Furthermore, intensifying agriculture through heavy grazing significantly reduced both abundance and bird species richness, and led to homogenized communities. My results suggest that low intensity agricultural lands are important for the majority of Himalayan birds throughout their annual cycle. Conservation strategies in the Himalayas must therefore go beyond establishing protected areas and prioritize retention of low intensity agricultural lands, minimizing their intensification, given their relatively high conservation importance for Himalayan birds.

Related Publications:

Elsen, P. R., Kalyanaraman, R., Ramesh, K. & Wilcove, D. S. 2017. The importance of agricultural lands for Himalayan birds in winter. Conservation Biology 31, 416-426. Link PDF

See media and news related to this work:

Blog article summarizing research by Fred Singer

India Water Portal (in Hindi)

CatchNews (in Hindi)


Developing priorities for montane species conservation

Montane species are expected to be particularly threatened by climate change as warming temperatures drive species’ ranges upslope. This is based in part on the assumption that species are squeezed into ever-smaller areas as they ascend, leading to population declines and eventual extinction. I conducted a global analysis of topography to illustrate that a diversity of area-elevation patterns exists within the world’s mountain ranges, which suggests that population responses will be specific to their local,

Complex topography beyond Great Himalayan National Park, Himachal Pradesh, western Himalayas

topographic context. Accounting for underlying topography in conservation planning can therefore target critical “pinch points” where species’ range shifts result in significant area reductions. Protection of these key “topographic corridors” may facilitate species adaptation to climate change. Within this theoretical framework, for my current Smith Fellowship I am developing and implementing a novel approach to conserving montane species under climate change across US mountain ranges, with particular focus on California ranges. Results will identify both priority regions for montane conservation and context-specific recommendations for stakeholders to guide climate-informed, proactive conservation strategies for species adaptation under climate change.

Related Publications:

Elsen, P. R. & Tingley, M. W. 2015. Global mountain topography and the fate of montane species under climate change. Nature Climate Change 5, 772-776. Link PDF

See media and news related to this work:

Washington Post


Carbon Brief

Atlas Obscura