The last week leading up to the 16-21 August North American Ornithological Conference (NAOC, the Olympics of bird study conferences, see previous post) consisted of sporadic bursts of communication and new analyses from my family vacation in a cabin on a small lake in Maine. While this location was much better for catching Yellow Perch than it was for preparing for a conference the poster below was produced just in time to printed and shipped from Williamsburg to D.C.
As you can see, there are two graphs here, each depicting the relationship between housing price and one other variable. In the top graph I compared the density* of bird species of high regional conservation value (as determined by the Partners In Flight conservation project) to the sale price of a home. There is a weak, yet positive and significant relationship between the two, which suggests that more expensive houses are, for some reason, better for birds. If we flip things around and look at the relationship between non-native, invasive species (Rock Pigeon, House Sparrow and European Starling) and housing prices we find that the regression line is almost a mirror image. Such a result suggests that humans may be willing to pay more for properties where the avian community is healthier. It is important for me to mention that correlation does not prove causation and that there are many possible reasons for the relationships I’ve just described above. For instance, it is very likely that landscaping has something to do with both a person’s assessment of a home’s value and a bird’s assessment of a property’s suitability for nesting. As I progress in my research throughout the coming semester I plan on trying to tease this tangle apart.
Looking forward to the fall semester I am planning on conducting further analyses on my data as well as mining geospatial habitat data for clues on the influence of landscaping on housing price and developing house pricing models with Professor Rob Hicks of the economics department.
With that I would like to thank Professor Dan Cristol for waking up at the crack of dawn with me to do bird surveys and Professor Matthias Leu for his help with my study design and data analysis. Finally, without the generous financial support of both the Charles Center and the Williamsburg Bird Club for this summer research would not have been possible.
*The reason I am using density instead of simply the number of birds that fall into the high conservation value category is that the current standards for publication of survey dictate that analyses are done in such a fashion. The idea is that when surveys are conducted, some birds that are present go undetected. Each species has its own probability of detection based on its behavioral tendencies. For example, a cryptic Chipping Sparrow is much more difficult to detect than a raucous Carolina Wren so the probability of detecting a Carolina Wren if it is present is higher than detecting a Chipping Sparrow. In order to account for these biases I have used the computer program Distance Sampling, which uses species area detection curves to approximate the density of each species.