Is there a pattern evident in the successional order of habitat types recolonized by certain diurnal raptor species within the last twenty years along urban/rural gradients in Washington, Oregon, and California? This is the question I hope explore throughout this course. I have yet to finalize which raptor species I will include in this analysis, but I will likely use observations of red-tailed hawks, merlins, American kestrels, and northern harriers. This analysis will be automated with a Python script so the number of species I choose to include is not critical at this stage of the research.
To answer this question, I will use citizen science bird observations and land cover data. The bird observations will come from the eBird database, an online citizen science monitoring and reporting system. For each eBird observation, an observer submits the species, the geographic location, and the date/time. These data exist at a range of spatial scales from several meters to several kilometers. Because I will analyze habitat colonization at the patch scale, I have already written Python code to select observations where lat/long is reported at or above a specified precision. The extent of these data stretches beyond the continental US. Temporally, eBird data are reported with a timestamp specified to the minute, however, my analysis will only require resolution to the month. I will use data from 1995 to 2014. Data exist for observations before 1995, but they are much more sparse.
The land cover data I will use is derived from LandTrendr, a series of algorithms for extracting land cover change information from Landsat time series imagery. For this project, I will use only land cover data from Washington, Oregon, and California as these data already exist. Land cover data for other parts of the country would need to be processed, a time-consuming endeavor that is unnecessary for the purposes of this class. Land cover data for Washington, Oregon, and California exist for all years from 1995 to 2014.
From this research, I expect to see two general patterns: 1) species prevalence for each selected species will increase inside highly developed patches from the beginning of the study period to the end, and 2) a successional order of colonized habitat types will be present for each species. A number of longitudinal studies have documented urban re-colonization by various diurnal raptor species. These re-colonization events may follow successional patterns. This process can be understood in the following way: when an urban area is first developed, existing raptor species likely retreat to remaining high quality habitat. As this high quality habitat reaches carrying capacity, pioneering individuals must find marginal habitat that still suits their ecological needs. This same process of demographic saturation is repeated, likely resulting in a successional pattern, as more marginal habitat is re-colonized.
From my limited knowledge of spatial statistics, it seems that a geographically weighted regression or ordinary least squares would be appropriate to tease out such patterns. eBird data are also notorious for spatial biases where observations are more numerous near easily accessible places (e.g. roads, trails, etc.) Perhaps a spatial autocorrelation analysis may work to assess the accuracy of successional pattern analysis results. To perform these analyses, I am proficient with ArcInfo, ModelBuilder, and Python. However, I don’t have any experience with R.
Within the last several decades, urbanization has rapidly altered available habitat for diurnal raptor species, leading to changes in geographic distribution. Greater understanding of urban raptor habitat selection may lead to more effective conservation. To date, however, data on urbanization patterns and species response are not available at the spatial and temporal scales needed to build this understanding. Answering the above research question will hopefully shed light on the effects of development on raptor distributions and aid city planners and conservationists in building more ecologically habitable cities.
