Where did the finch go? Insights from radio telemetry of the medium ground finch (Geospiza fortis)

Abstract

Movement patterns and habitat selection of animals have important implications for ecology and evolution. Darwin's finches are a classic model system for ecological and evolutionary studies, yet their spatial ecology remains poorly studied. We tagged and radio-tracked five (three females, two males) medium ground finches (Geospiza fortis) to examine the feasibility of telemetry for understanding their movement and habitat use. Based on 143 locations collected during a 3-week period, we analyzed for the first time home-range size and habitat selection patterns of finches at El Garrapatero, an arid coastal ecosystem on Santa Cruz Island (Galápagos). The average 95% home range and 50% core area for G. fortis in the breeding season was 20.54 ha ± 4.04 ha SE and 4.03 ha ± 1.11 ha SE, respectively. For most of the finches, their home range covered a diverse set of habitats. Three finches positively selected the dry-forest habitat, while the other habitats seemed to be either negatively selected or simply neglected by the finches. In addition, we noted a communal roosting behavior in an area close to the ocean, where the vegetation is greener and denser than the more inland dry-forest vegetation. We show that telemetry on Darwin's finches provides valuable data to understand the movement ecology of the species. Based on our results, we propose a series of questions about the ecology and evolution of Darwin's finches that can be addressed using telemetry.

Keywords: Geospiza fortis; behavior; communal roosting; habitat selection; home range; spatial ecology.

FIGURE 1

Maps of home ranges for radio‐tagged medium ground finches (a–e) at El Garrapatero on Santa Cruz Island, Galápagos. Each point represents the location of a finch

FIGURE 2

Jacobs’ index showing direction of selection for each habitat type for each finch. The gray shading is only for distinguishing the habitat types

FIGURE A1

Attached radio transmitters on Geospiza fortis band number (a) LF1233 and (b) LF1234. Mounting process of the radio transmitter backpack on individual JP4645 (c‐d‐e‐f)

FIGURE A2

Density of sampling effort across time of the day at El Garrapatero. The points at the bottom were y‐jittered to better see the amount of sampling at a particular time of the day. The sunrise (0606 h) and sunset (1814 h) on the first day of March 2019 is shown as the dotted vertical lines

FIGURE A3

Bearing estimation from razimuth package (shown for only 4 points of individual KGSK2033 as an example; Gerber et al., 2018). The points behind the transmitter estimate and posterior mode are the MCMC iterations (50,000)

FIGURE A4

Diagnostic plot for razimuth model outputs for the individual JP4645

FIGURE A5

Differences between habitats at El Garrapatero. The arid natural zone of El Garrapatero (dry‐forest, a‐c‐e) and the beach transitional zone more utilized by tourists (b‐d‐f). The photo (b) was taken at the beach, and (d‐f) were taken at the site referred to as “inland water” which was considered the edge of the manzanillo forest. The same site can be seen at El Garrapatero (a and c), but (a) is in a wetter season in 2019 than (c) in 2018

FIGURE A6

Minimum convex polygon rarefaction curve for each finch (a‐e). We used the function mcp (100%) from the adehabitatHR package to calculate the polygons (Calenge, 2006). All the points are ordered based on their sampling date. The blue line represents a Nonlinear Least Squares calculated with the function nlsLM from the minpack.lm package (Elzhov et al., 2016)

FIGURE A7

Bearing estimation from razimuth package (Gerber et al., 2018) quantifying the error location of the VHF emitters. The points not shown in the legend (the colour scale from yellow to purple) are the MCMC iterations (50,000)

FIGURE A8

Average distances traveled by finches from their nest during the diurnal and nocturnal activities

FIGURE A9

Estimation of nest location with the centroid of all diurnal activity locations of each finch. The yellow triangle is the known location of the nest of a bird. The black dot represents the centroid of all diurnal activity locations and the area around it is a buffer of 36.6 m determined by the upper limit of the 95% confidence interval of the distance between the nest and the centroid of all diurnal activity locations. Note that for female JP4645, the roosting points at the bottom of the map are from 2019‐02‐28 and 2019‐03‐01, before incubation started. From then on, this female remained on the nesting territory at night, as indicated by the three additional roosting points recorded during the incubation period, on 2019‐03‐04, 05, and 13. For the incubating female LF1234, the estimated location of the nest and that of nocturnal roosts are relatively close to each other, suggesting that this female also did not abandon the nesting territory during the night