ARCI comments on early feeding aggregations of Swallow-tailed Kites

As with other Swallow-tailed Kite breeding-season behaviors, these feeding flocks have always occurred during the same period each year, between fledging (which peaks in early June) and the formation of pre-migration communal roosts (which peak in late July). In addition, ARCI’s monitoring of nests and pre-migration roosts since 1988 has not revealed any changes in the year-to-year timing of fledging and roost development (consistent with predictions of others for long-distance migrants facing climate change), nor have we documented early nesting or unusually high failure rates this year.

One possible explanation for this year’s early feeding aggregations is the pattern of winds during March 2013 when Swallow-tailed Kites (and many other species, obviously) were trying to migrate northward over the Gulf of Mexico. Up to that point, ARCI (with colleague Jennifer Coulson) had been satellite-tracking 13 adult Swallow-tailed Kites tagged in Florida, Georgia, South Carolina, Mississippi, and Louisiana (www.arcinst.org; Research; Tracking studies). As always since this study began in 1996, we were collecting wind data (NOAA) during the kites’ trans-Gulf flights. In March, however, we noted a very unusual three-week series of large high-pressure systems that produced nearly continuous northerly winds from the southeastern U.S. to the northern Yucatan Peninsula, where the kites stage before their challenging over-water flight. A few of the kites made it safely to the U.S. Gulf coast before this began, but most waited for the infrequent spells when the strongest headwinds briefly abated, only to encounter stronger, more persistent northerlies once over the Gulf. Given that we were monitoring 13 birds, there are many stories within this story, but the net effect was that three tagged kites (two from Florida, one from South Carolina) did not survive because their lengthy, convoluted flight paths failed to reach the U.S. coast in time (we have learned that this species’ over-water limit is about four days).  Of the 10 that survived, three have not attempted to nest – a higher proportion than usual. Their late arrival may have resulted in loss of their previous territory or mate, or their preferred mate may not have survived migration and there was not sufficient time to pair with another.  Of the seven that did attempt to nest, at least three started unusually late.

In all, based on our sample of 13 marked birds and the uncommon weather, it seems likely that nesting effort would have been below average this year.  This could help explain the early feeding aggregations of adults. A possible contributing factor is that summer insect abundance might be peaking earlier due to climate change. However, this in itself would not hasten formation of post-breeding feeding aggregations without earlier nesting and/or higher failure rates, neither of which is occurring.  This is only speculation; I do not know if climate-related shifts in regional insect abundance have been documented.

In a related post, Patrick Leary suggested that post-fledging counts of long- versus short-tailed kites (presumably adult versus hatch year) might be useful for monitoring productivity. This method, along with guidance on aging based on plumage (complicated by some molting adults having symmetrically short tails) and a caution about the timing of southbound migration relative to age (adults leave earlier, causing the age ratio to shift continuously over time), was presented in a report to Florida Fish and Wildlife: Meyer, K. D. 1998. Communal roosts of American Swallow-tailed Kite in Florida… (available at ARCI’s website: www.arcinst.org; Publications). While systematic and repeated photographic counts at large pre-migration roosts that control for gradually decreasing ratios of adults to juveniles might support statistically valid estimates of annual productivity, less rigorous efforts likely will not.