GIS Applications for Monitoring Shorebird Populations across California’s Central Valley

An annotated bibliography created for GEOG 560
Fall 2023

by Denise Lopez

Annotated Bibliography

M. Barker, G. S. Maguire, M. A. Weston, D. A. Whisson, Non-breeding habitat selection of a sandy shore obligate shorebird. Estuarine, Coastal and Shelf Science. 271, 107848 (2022).

This study investigated habitat selection of non-breeding hooded plovers in Australia’s sandy beaches. Using known information about the species and remote sensing data, Beachscapes were broken up into four distinct types and using GIS, habitat variables were measured. Once analyzed, results showed sites were driven by ecological factors as well as social factors but were able to be narrowed down by locations. This information is helpful for land managers to be able to prioritize protected areas. This type of study can be performed with mapping shorebird use in the Pacific Flyway as most often it is for non-breeding use. The same template can be used to narrow down which areas need to be protected the most.

E. E. Conlisk et al., Both real‐time and long‐term environmental data perform well in predicting shorebird distributions in managed habitat. Ecological Applications. 32 (2022), doi:10.1002/eap.2510.

In this study, authors outline a challenge in monitoring migratory species and the habitat they reside in based on the changing needs throughout the year. Remote sensing has aided these efforts by collecting real time data, but authors sought to answer if this would improve predictability of distribution models compared to long-term average data. Shorebird distribution was analyzed for California’s Central Valley. Findings suggested that each technique had comparable results, so each had equal merit. Authors recommended long-term average data for permanent wetland restoration plans while real time data techniques could serve useful for adaptive conservation actions. This information would be interesting to see if GIS were applied. If factors measured were shown on a gradient (such as flooding, temperature, etc.) with shorebird distribution mapped on the same map, you can develop relationships that would help form more specific management plans.

J. Daniel, H. Polan, R. C. Rooney, Determinants of wetland- bird community composition in agricultural marshes of the Northern Prairie and parkland region. Wetlands. 41 (2021), doi:10.1007/s13157-021-01409-6.

The losses of wetland habitat in Northern Prairie and Parkland region in Canada is often due to agriculture. Bird use and presence can be linked to surrounding land use. Using cluster analysis, authors identified bird assemblages, used indicator species analysis, and predicted bird assemblage occurrence in wetlands using a classification and regression tree. Results showed the surrounding land use did not show a correlation to bird assemblages. Level of disturbance did not correlate with agricultural land use density. In fact, the most significant predictor was the type of natural habitat such as parkland vs. grassland. Other local factors were also more indicative of assemblage such as pond depth, duration of flooding, etc. One important finding was by combining clustering and indicator species analyses, authors were able to identify distinct avian assemblages and relate these to their functional traits. These techniques can be used to narrow down shorebird habitats in the Central Valley based on traits rather than proximity to agricultural land.

I.A. Dwight et al., Linking nest microhabitat selection to nest survival within declining pheasant populations in the Central Valley of California. Wildlife Research. 47, 391 (2020).

In California’s Central Valley, ring-necked pheasant populations are declining, primarily due to land use changes. Authors sought to measure how microhabitat conditions affect these population rates such as nest site selection related to their survival. 190 female pheasants were collared for radio telemetry or GPS to measure nest site characteristics from 2013-2017. Results showed nesting females preferred sites with herbaceous cover, specifically perennial grass. A link was made between height of said grass and nest survival rates. This type of study can be translated to shorebird use of wetlands based on water depth. Researchers already generally know what depth requirements of wetlands should be, but this can be applied specifically to California’s Central Valley.

C. L. Lauver, W. H. Busby, J. L. Whistler, Testing a GIS model of habitat suitability for a declining grassland bird. Environmental Management. 30, 88–97 (2002).

Due to declining habitat, authors sought to gather information on how to best manage loggerhead shrikes. Using Habitat Suitability Index models (HSI), landscapes were characterized with index values. Once measured, each location’s suitability to be habitat was evaluated resulting in a score from 0 to 1. This model was evaluated on potential habitat for loggerhead shrikes in Fort Riley Military Reservation in Kansas, USA. Authors used a developed model made for loggerhead shrikes as a basis for their model to be used in grassland habitat. A general land cover GIS data set was made to determine potential foraging habitat – more specifically hedgerow and tree density data sets. Results indicated the authors were able to successfully combine the HSI model with the GIS data sets to determine potential suitable habitat within the military reservation. Since the variables could be changed in this GIS model, the same procedure can be performed for non-breeding shorebirds in California’s Central Valley.

E. L. Matchett, J. P. Fleskes, Projected impacts of climate, urbanization, water management, and wetland restoration on waterbird habitat in California’s Central Valley. PLOS ONE. 12 (2017), doi:10.1371/journal.pone.0169780.

Matchett and Fleskes stressed the importance of the Central Valley in California as wintering habitat for waterbirds. As is the case for many other types of habitats, wetlands are quickly disappearing due to factors such as land use changes and climate change. This will continue without intervention thus the authors modeled area of waterbird habitats for 17 climate, urbanization, water supply management, and wetland restoration scenarios for years 2006–2099 using a water resources and scenario modeling framework. Results showed that wetland restoration projects continue to offset climate change effects, however this would be insufficient by the year 2065. The model reflected the greatest reduction in waterbird habitat occurred in scenarios that combined warmer, drier climate and decreased prioritizing of wetlands for water rights. To ensure the conservation of wetland habitat for waterbirds, managers will have to continue to put forth efforts that get ahead of any potential complications coming from future climate change. GIS models that create a visual representation of the projected disappearance of shorebird habitat in the Central Valley are essential to garner support from stakeholders involved.

A. Pfennigwerth, J. Albritton, T. Evans, Using spatial modeling to improve wetland inventories in Great Smoky Mountains National Park. Natural Areas Journal. 39, 482–488 (2019).

In this article, authors addressed the challenge of improving inventory techniques of wetlands found within Great Smoky Mountains National Park. As there are several differences in topography, elevation and distance between these wetlands, surveys could be expansive and time consuming. The current inventory and monitoring had identified 450 wetland sites within the park. However, some areas were more represented in data collected due to ease of data collection compared to others. A spatial modeling technique called Maximum Entropy (Maxent) models were implemented to predict wetland habitat suitability based on previous collected GIS-based data. This model was executed and compared to on the ground surveys. The surveys concluded Maxent models were fairly accurate in identifying wetland habitat within the park. Because of this new modeling technique, 350 new wetland sites were identified. Over three years, this expanded the inventory of wetlands within the park by 150%. Maxent modeling has overall improved monitoring efforts by reducing costs and labor. Most importantly, it allowed biologists to be able to gather information on previously inaccessible wetland sites. This technique can be applied to map wetlands in the Central Valley of California. Mapping wetlands narrows down locations of where shorebirds could or should be.

L.-M. Rebelo, C. M. Finlayson, N. Nagabhatla, Remote Sensing and GIS for Wetland Inventory, mapping and change analysis. Journal of Environmental Management. 90, 2144–2153 (2009).

Rebelo, Finlayson and Nagabhatla discussed the importance of sharing information on a global scale regarding wetland management and their qualities. The Ramsar Convention allowed just that. Information from individual projects and/or sites were compiled using remote sensing and GIS. Wetlands are deteriorating, most often due to land use conversions to agricultural purposes. To demonstrate these changes, a project using remote sensing and GIS data to quantify the status of wetlands on the western shoreline of Sri Lanka and a few other African countries. These projects were able to demonstrate how wetlands are disappearing and what type of land use it has been replaced by. They argued that this information at the global scale can help local managers to mimic such techniques to gain insight on how better manage their own wetlands. As the land in the Central Valley is used for mainly agricultural purposes, this type of GIS analysis can help local managers and shorebird advocates show the pressing need for protecting said habitat.

M. Rönkä, H. Tolvanen, E. Lehikoinen, M. von Numers, M. Rautkari, Breeding habitat preferences of 15 bird species on south-western Finnish Archipelago Coast: Applicability of Digital Spatial Data Archives to Habitat Assessment. Biological Conservation. 141, 402–416 (2008).

Authors of this article sought to assess the applicability of GIS and digital data archives for the analysis of coastal bird habitats by conducting a multivariate analysis on the relationship between physical island characteristics and the breeding site selection of 15 different avian species. Using Geographic Information Systems (GIS), environmental databases could be compiled including bathymetry, shoreline, and elevation data, to list a few elements. By collecting this information and creating an inventory of where the different bird species are found, relationships can be determined between chosen habitat and habitat elements. Information gathered via GIS is comparable to what we already know ecologically about these chosen 15 species. Results concluded GIS was able to corroborate what scientists suspect about each species’ preferred habitats. This would be an excellent application of GIS for shorebirds in the Central Valley of California to help determine where specific types of shorebirds could be found. It can be a very useful tool in aiding habitat restoration of threatened species.

A. Shabani, L. C. McArthur, M. Abdollahian, Comparing different environmental variables in predictive models of bird distribution. Russian Journal of Ecology. 40, 537–542 (2009).

Environmental predictors such as vegetation helps scientists determine the presence of species and their abundance in selected habitats. Other predictors could affect distribution thus authors compared three different analyses forecasting bird distribution and defining habitat suitability: discriminant function analysis, General Linear Models and ANOVA. GIS was utilized to display spatial distribution data in the study sites in Australia. Results indicated that some species have a relationship with vegetation patch aspects, however one species could not be an indicator of a particular aspect of the environment. Authors recommend studying the population change of a species over time to measure the change of environment over time. This can be applied to shorebird distribution mapping in California by using vegetation mapping.

D. A. Shealer, M. J. Alexander, Use of aerial imagery to assess habitat suitability and predict site occupancy for a declining wetland-dependent bird. Wetlands Ecology and Management. 21, 289–296 (2013).

Shealer and Alexander showed in this study how aerial imagery can be used to assess wetland habitat for black terns. Using Google Earth images and the National Wetlands Inventory maps to identify and rank 390 potential wetland sites through Wisconsin, USA. This was based on habitat suitability factors that were pre-determined. Ground surveys of these sites were then conducted afterward. Results were as followed: “Black terns were present at 47 % of the wetlands considered suitable but only 11 % of the sites considered marginal or unsuitable. Of the 42 sites where nesting was confirmed, 79 % were at wetlands classified as suitable; no nesting was recorded in any wetlands deemed unsuitable”. The authors concluded that this method of remote sensing was indeed accurate and could be applied to other species and habitat types.

C. Tattoni et al., Fruit availability for migratory birds: A GIS approach. PeerJ. 7 (2019), doi:10.7717/peerj.6394.

 Migratory patterns have proven difficult to analyze, especially due to the factors that alter such patterns such as food availability. In this study, authors compiled a GIS database containing 52 different wild plants that bear fruit over time in the SE Alps. It is assumed that the highest migrant abundance is correlated with habitats with fruit availability, but this information cannot be concluded at a larger scale. The database they developed using GIS allowed authors to create a raster map of each species of plant. The database structure could be applied to other scenarios such as insect abundance across migratory pathways of shorebirds, specifically the Pacific Flyway that goes over California’s Central Valley.

W. D. Shuford, G. W. Page, J. E. Kjelmyr, Patterns and dynamics of shorebird use of  California’s Central Valley. The Condor. 100, 227–244 (1998).

California’s Central Valley is considered one of the most important stopover sites for migrating North American shorebirds in the western United States. Surveys of shorebirds were conducted year-round at various locations ranging from managed wetlands to evaporation ponds from 1992-1995. Authors found populations of shorebirds averaged above 150,000 monthly based on the presence of suitable habitat in managed wetlands and agricultural fields of harvested rice. Populations varied depending on water availability due to natural or management factors. 33 shorebird species were identified throughout the surveys and their distribution and abundance was directly associated with habitat factors. For example, population numbers were lowest in August when managed wetlands were the least flooded and rice fields were not mature enough. The inverse was also true as populations were highest August-November when managed wetlands are being flooded. Yearly variations were also influenced by the amount of rainfall as the biggest increase in numbers from November to January was in 1994-1995 when precipitation was 67% higher than average. GIS applications could be useful in illustrating these patterns. By mapping distributions associated with average water levels or precipitation, we can see a stronger connection between these factors and shorebird presence

K. Tucker, S. P. Rushton, R. A. Sanderson, E. B. Martin, J. Blaiklock, Modelling bird distributions — a combined GIS and Bayesian rule-based approach. Landscape Ecology. 12, 77–93 (1997).

Authors chose the golden plover, snipe, and coal tit to demonstrate how documented habitat preferences and life-history characteristics compare to breeding distribution models developed using this information with GIS. The study area was in Northern England and authors used a modeling approach based on Bayes Theorem. After performing said study, results varied and even provided poor predictions for the snipe. The potential for these models to be accurate is there, however the models do not consider complex factors such as interspecific competition or predation. Since the publication of this article, distribution models using GIS have improved significantly. However, the concerns authors bring up in their prediction models should be considered when applying GIS to create distribution models of shorebirds in the Central Valley.

T. S. Wilson et al., Climate and land change impacts on future managed wetland habitat: A case study from California’s Central Valley. Landscape Ecology. 37, 861–881 (2022).

Wetlands are one of the most endangered habitat types in the world with only about 10% of historical sites remaining. As they are an extremely biodiverse habitat providing resources for a variety of bird species, their disappearance is detrimental to the organisms that rely on their existence. Authors sought to predict the distribution, abundance and connectivity of surface water and managed wetlands in California’s Central Valley based on expected changes resulting from climate change. Results from their climate-driven hydrologic water use model showed that decreased water availability had a greater impact than land use changes. This will affect waterbird habitat, especially in the month of January when migrating season is at its peak. This information is helpful in aiding land managers prioritize conservation plans on specific areas. This will ultimately aid in protecting shorebird habitats as these populations continue to decline as well.