In migratory stopover habitats, bird abundance and composition change on a near daily basis. On any given day, the local bird community should reflect local environmental conditions but also the environments that birds encountered previously along their migratory route. For example, during fall migration, the coast of the Yucatán Peninsula in Mexico receives birds that have just crossed the Gulf of Mexico and their abundance and composition may be associated with regional factors such as wind conditions experienced on previous dates but also local factors such as fruit availability. Thus, we used three data sets to quantify the influence of wind and fruit on near daily variation in bird abundance and composition. Our bird data consists of the number of individuals per species captured using mist nets for two coastal national parks in the Yucatán Peninsula during fall migration in 2016 and 2017. The data is provided daily, as are the “net-hours,” i.e., the sum of the hours each net was open summed across nets. Thus, we analyzed bird captures standardized by net-hours. Our fruit data consists of the number of unripe and ripe fruits per species counted on each side of each mist net lane at various points during fall migration. Our wind data consists of the wind costs birds arriving at our sites would have experienced when departing the north coast of the Gulf of Mexico two days prior to their arrival. Wind cost reflects wind speed and direction and it was calculated using the wind.dl_2 function in the rWind package. The wind costs are averaged across the entire US Gulf coastline. We used Moran eigenvector maps to quantify the temporal structure of the bird, wind, and fruit data and we partitioned the variance in the bird data into the components explainable by wind or fruit, the temporal structure of wind or fruit, and temporal structure independent of wind or fruit. After running the analysis, we did not find a strong association between daily changes in bird abundance or community composition with wind conditions and ripe fruit availability. Thus, despite wind and fruit being known to be important to individual birds (influencing stopover duration and departure decisions), their effects might not scale up to be drivers of population and community-level variation.

File descriptions. (Note: further detail on data capture and processing methods can be found in the associated manuscript).

Feldmanetal_birddata.csv - Capture data from running mist nets at two sites Contoy (21.4730° N, 86.7884° W) and Cuyo (21.5300° N, 87.7486° W) in 2016 (Contoy: 17-aug - 14-nov; Cuyo: 4-aug - 14-nov) and 2017 (Contoy: 4-aug - 10-nov; Cuyo: 3-aug - 16-nov). For every capture, the species (scientific name, english name, four letter AOU code), the number of individuals, and the date (including Julian day) are given. We also added information on whether the species is partially frugivorous and whether it is resident or migratory at our sites.

Feldmanetal_fruitdata.csv - Fruit counts from each of the mist net lanes used to capture birds. Counts were made by walking along the 12 m net lane and counting fruit on both sides, observing 2 m into the vegetation. The data indicate the species, the number of unripe and ripe fruit, and the date (including Julian day). Fruit was counted on the following dates:

• Contoy 2016: 18-aug, 25-aug, 1-sep, 8-sep, 16-sep, 22-sep, 30-sep, 3-oct, 6-oct, 13-oct, 20-oct, 27-oct, 10-nov
• Cuyo 2016: 13-aug, 20-aug, 24-aug, 3-sep, 9-sep, 18-sep, 22-sep, 6-oct, 18-oct, 31-oct, 3-nov, 10-nov
• Contoy 2017: 4-aug, 8-aug, 15-aug, 22-aug, 29-aug, 5-sep, 14-sep, 19-sep, 27-sep, 2-oct, 11-oct, 17-oct, 24-oct, 31-oct, 7-nov, 13-nov
• Cuyo 2017: 4-aug, 11-aug, 18-aug, 24-aug, 1-sep, 8-sep, 22-sep, 29-sep, 6-oct, 12-oct, 20-oct, 27-oct, 3-nov, 14-nov 

Feldmanetal_winddata.csv - The wind cost at each site for each day with bird data. Wind cost was calculated using the wind.dl_2 function in the rWind package ver. 1.1.5 (Javier Fernández-López, Klaus Schliep (2019). rWind: Download, edit and include wind data in ecological and evolutionary analysis. Ecography 42: 804–810), which accesses the Global Forecast System of the US National Weather Service (https://www.ncdc.noaa.gov/data-access/model-data/model-datasets/global-forcast-system-gfs). We calculated wind cost for every point on the shoreline of the Gulf of Mexico in the United States, accessing point data for the states of Texas, Louisiana, Mississippi, Alabama, and Florida from the National Assessment of Shoreline Change project (https://coastalmap.marine.usgs.gov/national/shorelc/). We calculated the angle between each point and each of our study sites and calculated the wind cost along that trajectory using the cost.FMGS function in rWind. (The function combines wind angle and wind speed to calculate cost). We calculated the cost at 6 pm, 7 pm, and 8 pm for the date two previous to the date of our bird data. We averaged the cost across the three hours and across all the shoreline points to arrive at a final wind cost, which is given in the "density" column of the file.

Feldmanetal_completedata.csv - The data we used to analyze the temporal pattern in bird abundances, the temporal pattern in fruit counts, the temporal pattern in wind cost, and the relationship between birds, fruit, and wind. The data combines the bird and wind data from the above files. For the fruit data, we had to calculate fruit counts for each mist net date. (Bird data were collected daily but fruit data were collected weekly). We did so by taking the weighted average of fruit counts for the site and year with the weight dependent on how close in time were the dates of fruit capture and bird capture.

Feldmanetal_Rcode.r - Complete R code used in the analysis.

The data and code files were generated on 2021-03-21 by Richard Evan Feldman and corresponds to the following publication:

Feldman RE, Celis Murillo A, Deppe JL, Dorantes Euan A. Local fruit availability and en route wind conditions are poor predictors of bird abundance and composition during fall migration in coastal Yucatán Peninsula. Wilson Journal of Ornithology

Corresponding author information
    
    Dr. Richard Evan Feldman
    Unidad de Recursos Naturales
    Centro de Investigación Científica de Yucatán
    Calle 43 #130 x 32 y 34
    Col. Chuburná de Hidalgo, CP 97205
    Mérida, Yucatán, México
    Tel: +52 999 942 8330 ext. 233
    Email: richard.feldman@cicy.mx

The data in this paper was collected with financial support from the Secretaría de Medio Ambiente y Recursos Naturales of Mexico and the Consejo Nacional de Ciencia y Tecnología of Mexico (Fondo Sectorial de Investigación Ambiental #262986) and permission to band birds was provided by the Secretaria de Medio Ambiente y Recursos Naturales of Mexico under permit SGPA/DGVS/05989/16.