Both real-time and long-term environmental data perform well in predicting shorebird distributions in managed habitat.

Highly mobile species, such as migratory birds, respond to seasonal and interannual variability in resource availability by moving to better habitats. Despite the recognized importance of resource thresholds, species-distribution models typically rely on long-term average habitat conditions, mostly because large-extent, temporally resolved, environmental data are difficult to obtain. Recent advances in remote sensing make it possible to incorporate more frequent measurements of changing landscapes; however, there is often a cost in terms of model building and processing and the added value of such efforts is unknown. Our study tests whether incorporating real-time environmental data increases the predictive ability of distribution models, relative to using long-term average data. We developed and compared distribution models for shorebirds in California's Central Valley based on high temporal resolution (every 16 days), and 17-year long-term average surface water data. Using abundance-weighted boosted regression trees, we modeled monthly shorebird occurrence as a function of surface water availability, crop type, wetland type, road density, temperature, and bird data source. Although modeling with both real-time and long-term average data provided good fit to withheld validation data (the area under the receiver operating characteristic curve, or AUC, averaged between 0.79 and 0.89 for all taxa), there were small differences in model performance. The best models incorporated long-term average conditions and spatial pattern information for real-time flooding (e.g., perimeter-area ratio of real-time water bodies). There was not a substantial difference in the performance of real-time and long-term average data models within time periods when real-time surface water differed substantially from the long-term average (specifically during drought years 2013-2016) and in intermittently flooded months or locations. Spatial predictions resulting from the models differed most in the southern region of the study area where there is lower water availability, fewer birds, and lower sampling density. Prediction uncertainty in the southern region of the study area highlights the need for increased sampling in this area. Because both sets of data performed similarly, the choice of which data to use may depend on the management context. Real-time data may ultimately be best for guiding dynamic, adaptive conservation actions, whereas models based on long-term averages may be more helpful for guiding permanent wetland protection and restoration.

Impact of extreme drought and incentive programs on flooded agriculture and wetlands in California's Central Valley.

BACKGROUND: Between 2013 and 2015, a large part of the western United States, including the Central Valley of California, sustained an extreme drought. The Central Valley is recognized as a region of hemispheric importance for waterbirds, which use flooded agriculture and wetlands as habitat. Thus, the impact of drought on the distribution of surface water needed to be assessed to understand the effects on waterbird habitat availability. METHODS: We used remote sensing data to quantify the impact of the recent extreme drought on the timing and extent of waterbird habitat during the non-breeding season (July-May) by examining open water in agriculture (rice, corn, and other crops) and managed wetlands across the Central Valley. We assessed the influence of habitat incentive programs, particularly The Nature Conservancy's BirdReturns and The Natural Resources Conservation Service's Waterbird Habitat Enhancement Program (WHEP), at offsetting habitat loss related to drought. RESULTS: Overall, we found statistically significant declines in open water in post-harvest agriculture (45-80% declines) and in managed wetlands (39-60% declines) during the 2013-2015 drought compared to non-drought years during the period of 2000-2011. Crops associated with the San Joaquin Basin, specifically corn, as well as wetlands in that part of the Central Valley exhibited larger reductions in open water than rice and wetlands in the Sacramento Valley. Semi-permanent wetlands on protected lands had significantly lower (39-49%) open water in the drought years than those on non-protected lands while seasonal wetlands on protected lands had higher amounts of open water. A large fraction of the daily open water in rice during certain times of the year, particularly in the fall for BirdReturns (61%) and the winter for WHEP (100%), may have been provided through incentive programs which underscores the contribution of these programs. However, further assessment is needed to know how much the incentive programs directly offset the impact of drought in post-harvest rice by influencing water management or simply supplemented funding for activities that might have been done regardless. DISCUSSION: Our landscape analysis documents the significant impacts of the recent extreme drought on freshwater wetland habitats in the Central Valley, the benefits of incentive programs, and the value of using satellite data to track surface water and waterbird habitats. More research is needed to understand subsequent impacts on the freshwater dependent species that rely on these systems and how incentive programs can most strategically support vulnerable species during future extreme drought.

Using ricelands to provide temporary shorebird habitat during migration.

To help mitigate large wetland losses in California, The Nature Conservancy launched a dynamic conservation incentive program to create temporary wetland habitats in harvested and fallow rice fields for shorebirds migrating along the Pacific Flyway. Farmers were invited to participate in a reverse auction bidding process and winning bids were selected based on their cost and potential to provide high quality shorebird habitat. This was done in 2014 and 2015, for separate enrollment periods that overlapped with spring and fall migration, both before and after the traditional post-harvest flooding period. To assess the success of the program, we monitored shorebird use of fields that were enrolled (treatments), and others that were subject to typical rice farm management (controls). To put these observations in context, we used satellites to simultaneously monitor the extent of shallow-water habitat across the ~215,000 ha of ricelands in the area. Results showed that providing habitat during migration, when it is typically unavailable in rice fields, yielded the largest average shorebird densities ever reported for agriculture in the region. Treatment fields had significantly greater shorebird density, richness and diversity than control fields in both spring and fall (especially September-early October, and late March-early April), but in fall the difference was greater. Shorebird responses to habitat provisioning, and regional habitat conditions, were variable from year to year, and highly dynamic within a given season. Overall, shorebirds densities were found to be negatively related to the total amount of flooded habitat in the rice landscape. Factors that affected habitat availability included allocation schedules of water deliveries from reservoirs, and rainfall patterns, both of which were influenced by drought. Collectively, these results suggest that appropriately managed agricultural lands have great potential to provide high value habitat for shorebirds during times of habitat deficit, including migration, and that fall may be a particularly impactful time to create additional habitat. Migratory species face great challenges due to the climate change, conversion of historical stopover sites, and other factors, but dynamic conservation programs offer promise that, at least in certain instances, their needs can still be met.

Dynamic conservation for migratory species.

In an era of unprecedented and rapid global change, dynamic conservation strategies that tailor the delivery of habitat to when and where it is most needed can be critical for the persistence of species, especially those with diverse and dispersed habitat requirements. We demonstrate the effectiveness of such a strategy for migratory waterbirds. We analyzed citizen science and satellite data to develop predictive models of bird populations and the availability of wetlands, which we used to determine temporal and spatial gaps in habitat during a vital stage of the annual migration. We then filled those gaps using a reverse auction marketplace to incent qualifying landowners to create temporary wetlands on their properties. This approach is a cost-effective way of adaptively meeting habitat needs for migratory species, optimizes conservation outcomes relative to investment, and can be applied broadly to other conservation challenges.

Abundance models improve spatial and temporal prioritization of conservation resources.

Conservation prioritization requires knowledge about organism distribution and density. This information is often inferred from models that estimate the probability of species occurrence rather than from models that estimate species abundance, because abundance data are harder to obtain and model. However, occurrence and abundance may not display similar patterns and therefore development of robust, scalable, abundance models is critical to ensuring that scarce conservation resources are applied where they can have the greatest benefits. Motivated by a dynamic land conservation program, we develop and assess a general method for modeling relative abundance using citizen science monitoring data. Weekly estimates of relative abundance and occurrence were compared for prioritizing times and locations of conservation actions for migratory waterbird species in California, USA. We found that abundance estimates consistently provided better rankings of observed counts than occurrence estimates. Additionally, the relationship between abundance and occurrence was nonlinear and varied by species and season. Across species, locations prioritized by occurrence models had only 10-58% overlap with locations prioritized by abundance models, highlighting that occurrence models will not typically identify the locations of highest abundance that are vital for conservation of populations.