Response of pearly-eyed thrashers to two intense hurricanes in the Luquillo mountains, Puerto Rico
Author
Cruz Mendoza, Alberto C.
Advisor
Wunderle Jr., JosephType
ThesisDegree Level
M.S.Date
2022-06-23Metadata
Show full item recordAbstract
Hurricanes may have direct and indirect impacts driving changes in forest structure and productivity, thereby altering local avian species richness, abundance, or density. After hurricanes, birds may shift sites in response to vegetation damage or changes in resource availability. The Pearly-eyed Thrasher (Margarops fuscatus; thrasher hereafter) is recognized as a "supertramp" because of its high dispersal ability, aggressive behavior, and opportunistic diet, which facilitates colonization of vacant niches in disturbed habitats. Consistent with the supertramp concept, are previous studies indicating that the abundance of thrashers increased in storm-damaged sites after hurricanes. Therefore, I expected thrasher site occupancy and abundance to increase in the most damaged sites in the year (2018) after the passage of the 2017 hurricanes Irma and Maria in the Luquillo Experimental Forest in Puerto Rico. Studies here after the hurricanes indicated that the storm damage to the vegetation increased with elevation, and therefore, I predicted that thrasher site occupancy and abundance would increase with elevation. To test this hypothesis, I used historical point count data of thrashers for the pre-hurricane baseline (1998, 2005) for comparison with posthurricane point counts conducted at similar sites along the elevation gradient in 2018. During April-August 2018, teams of two observers sampled 158 georeferenced points, which were visited three times each (April-May; June-July; July-August) to sample thrashers. Thrasher site occupancy was estimated with a single-season model and site abundance was estimated with an N-mixture model. In the first chapter, I used elevation, forest type, a two-category subjective assessment of hurricane damage at each point count site and year as the covariates for the models. In the second chapter, I included the Normalize Difference Vegetation Index (NDVI) derived from satellite imagery as a measure of storm damage at each point count site. In addition to covariates of elevation and forest type gathered at each point count site, I used the Mean Information Gain Index (MIG), Green Index (ExG-ExR), and the Leaf Area Index (LAI) as forest structure covariates.
In the first chapter, analyses of detectability suggest that thrasher breeding was delayed and curtailed in the first breeding season after the hurricanes, as demonstrated previously after other hurricanes. My results indicate a decline in thrasher site occupancy and abundance from 1998 to 2005. However, more recently published findings from 2015 indicate that thrasher site occupancy did not continue to decline after 2005 until after the 2017 hurricanes. Although overall thrasher site occupancy and abundance declined from 2005 values after the hurricanes, thrasher site occupancy and abundance in 2018 remained highest at mid-elevation (400-800 m) sites, consistent with pre-hurricane studies. I found no evidence for a shift in thrasher site occupancy and abundance after the hurricanes.
In the second chapter I found a positive relationship between sierra palm cover with thrasher site occupancy and abundance, which is consistent with the importance of sierra palm fruit in the thrasher diet. Also, I demonstrated that NDVI per site showed a negative linear relationship with elevation corroborating previous findings that vegetation damage was greatest at high elevations. In contrast to my predictions, however, thrashers at high elevations (> 600 m) had a higher probability of occupying sites with minimal or no storm damage (as indicated by NDVI values) than more heavily damaged sites in 2018. These heavily damaged sites may have had lower fruit abundance than less damaged sites as suggested by my field observations. Thrashers may delay colonization of damaged sites at high elevation because of the slow plant regrowth there which retards fruitin abundance declines g. Despite posthurricane in the first year after the hurricanes site occupancy and, the thrasher continue potential threat as a predator and competitor of endangered wildlife at mids to be a elevation forests in the Luquillo Experimental Forest.
In the first chapter, analyses of detectability suggest that thrasher breeding was delayed and curtailed in the first breeding season after the hurricanes, as demonstrated previously after other hurricanes. My results indicate a decline in thrasher site occupancy and abundance from 1998 to 2005. However, more recently published findings from 2015 indicate that thrasher site occupancy did not continue to decline after 2005 until after the 2017 hurricanes. Although overall thrasher site occupancy and abundance declined from 2005 values after the hurricanes, thrasher site occupancy and abundance in 2018 remained highest at mid-elevation (400-800 m) sites, consistent with pre-hurricane studies. I found no evidence for a shift in thrasher site occupancy and abundance after the hurricanes.
In the second chapter I found a positive relationship between sierra palm cover with thrasher site occupancy and abundance, which is consistent with the importance of sierra palm fruit in the thrasher diet. Also, I demonstrated that NDVI per site showed a negative linear relationship with elevation corroborating previous findings that vegetation damage was greatest at high elevations. In contrast to my predictions, however, thrashers at high elevations (> 600 m) had a higher probability of occupying sites with minimal or no storm damage (as indicated by NDVI values) than more heavily damaged sites in 2018. These heavily damaged sites may have had lower fruit abundance than less damaged sites as suggested by my field observations. Thrashers may delay colonization of damaged sites at high elevation because of the slow plant regrowth there which retards fruitin abundance declines g. Despite posthurricane in the first year after the hurricanes site occupancy and, the thrasher continue potential threat as a predator and competitor of endangered wildlife at mids to be a elevation forests in the Luquillo Experimental Forest.