Birds, Buckthorn and Oaks Final

by Carl Strang

As I mentioned in my previous post on this subject, the songbird migration was nearly done by mid-May. In fact I have only 3 observations to add to the ones from the early part of the season, two in the oaks and one in the buckthorn-dominated woodland. That mildly reinforces my earlier conclusion, that the only reason buckthorn woodlands appeal to birds early in the season in our area is that oaks, which dominate most of our woodlands, leaf out late. In this year when the oaks were in leaf throughout the migration season, songbird migrants nearly abandoned the insect-depauperate buckthorn at Mayslake Forest Preserve and spent their time in the oak woodlands.

South savanna at Mayslake, showing oaks well leafed.

So far all the analysis has been of the total species counts of neotropical migrants that do not nest at Mayslake. However, some of those species are brush specialists that might be expected to prefer the buckthorn over the more open oak woodlands regardless of what is happening in the canopy.

The Tennessee warbler is an example of a canopy species, on the other hand, which will not want to spend a lot of time foraging down in the buckthorn, though they might want to have it handy for resting and as a refuge from predators.

When I look at the two years’ data, sorting out brush from non-brush birds, the most curious observation is that hardly any brush-loving birds stopped by Mayslake this year. Only two birds out of the total sample of 30 were of brush species. Last year, 38 of 166 were brush birds. Looking at 2011, then, both brush and non-brush birds preferred the buckthorn woodland early in the season, though the preference was stronger for the brush species. Late last year, after the oaks were leafing out, non-brush species preferred oaks while the brush birds did not show a strong preference for either woodland type.

All of these data, coming as they do from one site over a short period of time, amount to a pilot study at best. Nevertheless, they support the notion that the apparent preference of migrant songbirds for woodlands dominated by invasive shrubs is an illusion. Birds avoid the buckthorn and prefer restored native woodlands when the latter are in leaf, capable of providing both food and essential shelter.

Birds, Buckthorn and Oaks

by Carl Strang

Now that the early part of the migration season is past, I can do a preliminary test of the Birds-Buckthorn-Oaks hypothesis. The idea is that migrating birds seem to prefer buckthorn infested woodlands early in the migration season because the buckthorn understory provides shelter unavailable in restored, oak-dominated woodlands because usually the oaks have not yet leafed out to provide food and shelter. Data from last year supported this idea, because late-season migrants showed a shift from the buckthorns to the newly leafing oaks. This year a further, better test was made possible by the warm early season stimulating oaks to break buds early, so that they were well in leaf for the early part of the migration season. So, what did the data show? I compiled the numbers of neotropical migrants that don’t nest on Mayslake Forest Preserve that I counted prior to May 19 (when oaks began leafing last year), comparing last year’s counts to this year’s. The results were stark: I counted only 5 of those birds in the buckthorn woodland, compared to 22 in the restored savannas (compared to 80 and 34, respectively, last year. Numbers were low this year, in part because my back strain limited my outings, but also a lot of migrants seemed to be bypassing Mayslake).

Things are moving along. This olive-sided flycatcher, a late season migrant, stopped by Mayslake on May 14, earlier than usual.

Using last year’s proportions to calculate expected values if there had been no difference between years, the values would have been 18.9 in the buckthorn-dominated woodland, 8.1 in the savanna. It should come as no surprise that the resulting chi-squared test statistic showed a highly significant difference (34.07, even with only 1 degree of freedom, is far above the threshold). With the migration progressing so rapidly, I don’t know if there will be enough observations in the late season to consider separately, but if so there will be an additional post. Either way, I am satisfied that the data support my point.

Testing the Birds and Buckthorn Hypothesis

by Carl Strang

A year ago I posted a hypothesis that this most unusual of seasons will allow me to test. To recap: When restored savannas are compared to woodlands with buckthorn and honeysuckle thickets in the understory in early May, when the bird migration is entering its peak, it seems that the birds prefer the invasive-degraded areas to the native savannas. Some birders take this as evidence that restoration is bad for birds. My hypothesis was that this observation is tied to the fact that oaks, the dominant trees in our savannas and woodlands, are among the latest trees to break bud and leaf out. Therefore they are not supporting leaf-eating insects, and also not providing the shelter that the birds need on their daytime migratory stops. As I pointed out a few weeks ago, like our flowering phenology, the oaks broke buds a few weeks early this year.

Remember this photo? I took it in early April, more than a month before oaks typically reach this point.

The stage thus is set. I have the records of where I saw the migrants last year, a late year when the oaks were not leafing out until the second half of May. I remarked in my notes that they were doing so around May 19, so I will take that as my dividing point. I can look at last year’s data, and this year’s, and see if the birds lose their apparent preference for the buckthorn woodlands now that they have leafy oaks as an alternative.

As a starting point I compared the bird counts from April 20 to May 18 last year to those from May 19 to May 31, when migration was essentially done. I considered only species that spend their winters south of the continental U.S. and that do not breed on the Mayslake preserve, to keep things as uniform and unbiased as possible. Even with those restrictions, I had 22 species to work with. For the moment ignoring species by species comparisons, here are the 2011 totals. In buckthorn woodlands before May 19, I made 80 observations of birds in the target group. The corresponding total for restored savanna areas in that same time period was 34 (these counts are not normalized for the relative areas of the two habitats, but the buckthorn area I used for this comparison is smaller, at 5.7 acres, than the savanna at 8.5). So the data support the notion that, at least early in the migration season before the oaks leafed out, the lower quality, buckthorn- and honeysuckle-dominated woodland harbored more birds. What about the latter part of the migration season, after May 18? Things had slowed down at that point, and the migrant species composition changed somewhat, but the totals last year were 24 observations in the buckthorn woodland and 28 in the savannas. Clearly the tide turned after the oaks began to leaf out (for the statistics cognoscenti, the chi-squared contingency table produced a test statistic value of 14.36 at 1 degree of freedom, highly significant). I will report on what happens this year, but if my hypothesis is correct, the oak savanna should prove more attractive to these migrants this year in both parts of the season.

%d bloggers like this: