Long-term management planning for conservation-reliant migratory songbirds is particularly challenging because habitat quality in different stages and geographic locations of the annual cycle can have direct and carry-over effects that influence the population dynamics. The Neotropical migratory songbird Kirtland's warbler Setophaga kirtlandii (Baird 1852) is listed as endangered under the U.S. Endangered Species Act and Near Threatened under the IUCN Red List. This conservation-reliant species is being considered for U.S. federal delisting because the species has surpassed the designated 1000 breeding pairs recovery threshold since 2001.
To help inform the delisting decision and long-term management efforts, we developed a population simulation model for the Kirtland's warbler that incorporated both breeding and wintering grounds habitat dynamics, and projected population viability based on current environmental conditions and potential future management scenarios. Future management scenarios included the continuation of current management conditions, reduced productivity and carrying capacity due to the changes in habitat suitability from the creation of experimental jack pine Pinus banksiana (Lamb.) plantations, and reduced productivity from alteration of the brown-headed cowbird Molothrus ater (Boddaert 1783) removal programme.
Linking wintering grounds precipitation to productivity improved the accuracy of the model for replicating past observed population dynamics. Our future simulations indicate that the Kirtland's warbler population is stable under two potential future management scenarios: (i) continuation of current management practices and (ii) spatially restricting cowbird removal to the core breeding area, assuming that cowbirds reduce productivity in the remaining patches by ≤41%. The additional future management scenarios we assessed resulted in population declines.
Synthesis and applications. Our study indicates that the Kirtland's warbler population is stable under current management conditions and that the jack pine plantation and cowbird removal programmes continue to be necessary for the long-term persistence of the species. This study represents one of the first attempts to incorporate full annual cycle dynamics into a population viability analysis for a migratory bird, and our results indicate that incorporating wintering grounds dynamics improved the model performance.
Data Dictionary and RAMAS Code
Description of the spatial layers used, and code for parameterizing the Spatial Data subprogram in program RAMAS GIS using the spatial
files provided and equations/values given in the manuscript and Supplementary Information document.
DataDictionaryAndRAMASCode.txt
Historical Projections
HistoricalProjections -> HabitatSuitability
hs1979-hs2013 are spatial layers that distinguish 32 ha cells in Michigan, USA as nonsuitable (0), low suitability (1), moderate suitability (2), and high suitability (3) habitat for Kirtland's warbler annually between 1979 and 2013. The layers can be used as input files for the Spatial Data subprogram in program RAMAS GIS.
HistoricalProjections -> Precipitation
p1979-p2013 are spatial layers that contain total March precipitation values for the Nassau NOAA station (Bahamas) annually between 1979 and 2013 in cells that contain suitable habitat for Kirtland's warbler. The layers can be used as input files for the Spatial Data subprogram in program RAMAS GIS.
HistoricalProjections.zip
FutureProjections
FutureProjections
hr2010 is a spatial layer that distinguishes 32 ha cells in Michigan, USA as nonsuitable (0), low suitability (1), moderate suitability (2), and high suitability (3) habitat for Kirtland's warbler in 2010. The layer was derived from hs2010, where 25% of suitable habitat cells were randomly selected and their habitat suitability values were reduced by 1 suitability class. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
hs2010 is a spatial layer that distinguishes 32 ha cells in Michigan, USA as nonsuitable (0), low suitability (1), moderate suitability (2), and high suitability (3) habitat for Kirtland's warbler in 2010. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
nc2010 is a spatial layer used to assess the impacts of elimination of the cowbird trapping program on Kirtland's warbler population dynamics, assuming a 57% reduction in productivity. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
nc2010_41 is a spatial layer used to assess the impacts of elimination of the cowbird trapping program on Kirtland's warbler population dynamics, assuming a 41% reduction in productivity. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
rc2010 is a spatial layer used to assess the impacts of spatially-restricted cowbird trapping to the core breeding area on Kirtland's warbler population dynamics, assuming a 57% reduction in productivity in satellite breeding patches in the Lower Peninsula of Michigan, USA. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
rc2010_41 is a spatial layer used to assess the impacts of spatially-restricted cowbird trapping to the core breeding area on Kirtland's warbler population dynamics, assuming a 41% reduction in productivity in satellite breeding patches in the Lower Peninsula of Michigan, USA. The layer can be used as an input file for the Spatial Data subprogram in program RAMAS GIS.
FutureProjections -> RandomizedPrecipitation
a1-a50, b1-b50, c1-c50, d1-d50, and e1-e50 are spatial layers that contain randomly generated total March
precipitation values in cells that contain suitable habitat for Kirtland's warbler. The values were generated based on the lower 90th percentile of the historical precipitation distribution recorded at the Nassau NOAA station (Bahamas) between 1994 and 2013. The layers can be used as input files for the
Spatial Data subprogram in program RAMAS GIS.