Sunday, February 3, 2013

Big January for American Tree Sparrows?

I cannot remember a year where I have seen so many American Tree Sparrows everywhere I go. Let me start this off by saying that the sentiment within this entry is unscientific in that it is reliant upon subjective views, feelings, and general impressions. I am not alone in this though and many anecdotal remarks concerning either American Tree Sparrow presence or abundance in Connecticut in January 2013 have been made publicly on places like the CT Birds List Serv or privately between birding friends. Personally I think these feelings, especially from experts who have spent decades watching the state's birds, are nearly always on the money. In some cases I had a limited amount of data to easily back this assumption of a big American Tree Sparrow winter up, such as around nine years of eBirding my own sightings, none of which come close to some of the numbers I had last month.

However, this can only go so far, and my personal data should back up what I "think" if I have any sort of memory. The problem is that places like CT Birds are inherently fraught with positive reporting - people post when they see an unusual feeder bird, but they do not post as often to say I do not have this species present this year. Negative reports are far less common. Another problem is that the areas I am surveying or watching, the data I have recorded myself over the years and most of the friends I regularly speak to are all centered in southwest Connecticut.  While it is a small state Connecticut has a very diverse geography and many different habitat and climate zones. For example, birds like White-throated Sparrows and Dark-eyed Juncos are year-round breeding residents in the northern regions while they only spend the winter with us on the coast! Other winter feeder species like the Common Grackle and Red-winged Blackbird are common for me but rare for those far inland friends.

I wanted to take a closer look at population numbers running on the seemingly widespread belief that there are an unusually high number of American Tree Sparrows in parts of Connecticut at least, both in group sizes when the species is spotted and birds in places where they typically do not spend the winter or are much rarer. One of the most helpful aspects of eBird is the ability to explore all of its data free and relatively easily. Even if it does not always come "out" the way we all want it to, you can still see the numbers for every single bird entered in a given period. I did just that for American Tree Sparrows in January 2013, 2012, 2011, and 2010 for the entire state. Bear in mind that January 2013 may not have had all of its checklists submitted yet even though the month has now ended as people may be a bit behind. Nevertheless, since some of what we are dealing with is averages it should work for these simple purposes. People regularly submit data going years back, so all of these numbers may change a little over time. Going back further than I did would let us see more and provide a clearer more scientific snapshot, but the data gets thinner and this is just for fun.

I created five graphs containing lines of each of those four years, all of which contain data points of January 1-5, 6-10, 11-15, 16-20, 21-25, and 26-31. In order they depict:

1) Raw numbers of American Tree Sparrows reported to eBird within each period which is biased to the number of reports entered (more reports would likely mean more birds)

2) The number of checklists entered that contained at least one American Tree Sparrow within each period which would have the same sort of bias as the first graph

3) Species abundance which is the average birds per all checklists entered in the period regardless of whether or not American Tree Sparrow was recorded

4) The percentage of all checklists that reported American Tree Sparrows in each period

5) Group size which is the average number of birds seen only when American Tree Sparrow was reported

Within this limited data set it certainly seems to me that our impressions in southwest Connecticut were both wrong and right in different regards. I do not want to draw a lot of conclusions from the first two graphs. Given that eBird is growing exponentially in time, 2012 sure seemed like a weak year for the species as both 2010 and 2011 trumped it in raw numbers.

Species abundance is where to fun starts. Yes, 2012 was weak once again, and 2013 beats it as well as 2011. However, 2010 is above 2013 in five out of six periods. 2013 and 2010 are tied at three each for periods with the higher frequency, but 2011 is a strong contender here as well. Group size shows us the final piece of the puzzle in that 2011 had fewer birds recorded when the species was found as opposed to the stronger 2010 and 2013. Despite my impressions of this being a banner January, 2010 still takes the crown for the most American Tree Sparrows in Connecticut of the four years. The current year is a solid second, and at some points is in the number one position. Maybe if I had completed this analysis for only Fairfield County the results would be different.

There are so many variables not accounted for in this small sample size examination that it would take hours to list them out and days to properly analyze it. Data collection, observers, their geographical locations and skill, weather in Connecticut and across the region, food supplies, feeder watches vs. other counts in other habitats and more are all part of the noise in the data we would have to consider. I do hope it shows you what eBird is capable of and why it is important to enter all of the data you collect whether it is from your backyard, a sea or hawk watch, a Breeding Bird Survey route - anything and everything. All of it matters and can really help us find the facts among our beliefs.

Scott Kruitbosch
Conservation Technician

No comments:

Post a Comment