Tennessee BBS: It’s the Habitat

As alluded to in the last post, the population trends of birds indicated by numbers from the Breeding Bird Survey in Tennessee appear to be related in some degree to the habitat requirements of each species. I looked at this in more detail by categorizing each species according to very general, landscape-level habitat preferences. The categories I chose were:

Forests -- relatively continuous and closed canopy forests. Occurs throughough the state, but most extensive in the eastern mountains and hill 'n' holler portions of the Highland Rim.

Edge/Mosaic -- the mosaic of woodlots, smaller farm and pasture areas, edges, roadsides, successional clearcuts, lower-density houses, and other mixed habitats typical of rural and less dense suburban areas that archetypally characterize much of Tennessee. Widespread statewide, and especially typical of the Highland Rim, hillier parts of the Coastal Plain and Central Basin, and the less rugged areas of eastern Tennessee.

Open country -- large-scale acriculture, with most land area in croplands and pasture. Characteristic of much of the Coastal Plain and Mississippi Alluvial Valley, the flatter inner Central Basin, and the Barrens areas of the Highland Rim. Small areas of this sort of landscape occur in most regions of the state wherever flatter ground occurs.

Residential/suburban/urban -- higher population density areas with lawns, “landscaping,” parks, commercial areas, etc. Includes centers of small towns as well as suburbs of large metro areas.

For each species, I checked off one or more of these categories based on general pereferences, not strict obligate requrements. For instance, Broad-winged Hawks are scored as “forest” and Horned Larks as “open,” but I have seen and heard both species from the exact same spots on our “edge/mosaic” farm in a Highland Rim hollow.

I also added a fifth category, both because it is of much interest in the conservation community and because a quick inspection seemed to confirm its significance. I checked off species that are neotropical migrants -- those migrants for which the majority of the population winters south of the U.S.

And as a final detail, when I calculated averages and other statistics on the percent changes, I used a logarithmic transformation in the form of X = ln(N_f/N_i) where N_i is the average count in the initial 10 years (1966-1975) and N_f is the average count in the final 10 years (2001-2010). This straightens out and balances the skewed nature of these ratios and gives them a more approximately normal (statistically speaking) distribution. Using the log metric, if one species has dropped to half while another has doubled, they will average out to zero. Using the straight percentages, they would average out to +25%. To give the final numbers, I converted these log transfomed averages back to regular percentages.


Averages by habitat type

When treated as described above, the following average and median changes come out for birds of each habitat preference:

Habitat	No. Species	Median	Mean
Forest	36	+74%	+101%
Edge/Mosaic	59	+35%,	+24%
Open	27	+15%	+32%
Residential	20	+67%	+71%

Though statistical tests do not make these trends significant, they confirm the impression that forest species are doing better than average. This gives a strong suggestion that one of the main driving factors for the rise in numbers of most Tennessee birds has been an improvement in the quantity and quality of forest habitat across the state.


The Neotropical Effect

When looked at in isolation, the difference between the neotropical migrants (NMs) and the non-neotropical migrants (non-NMs) was not significant. The 48 NM species showed an average increase of 21% (median 25%); the 53 non-NM species averaged 70% (median 75%). This appears to be a substantial difference, but because of the very large variability between species this difference is not outside the 95% confidence zone. Remember that individual species range from a drop of 99.8% to a rise of 69400% (excluding the three species that were not recorded in the 1966-1975 period and hence have an infinite percentage increase).

It is more revealing to look at the interaction between the neotropical effect and the general habitat preferences. Here are the median percent changes for the species in each category, with the number of species in parentheses:

Habitat	NM	non-NM
Forest	+20% (20)	+143% (16)
Edge/Mosaic	+32% (29)	+55% (30)
Open	-5% (7)	+43% (20)
Residential	+231% (4)	+69% (16)

In the first three habitats the neotropical migrants all show lower median changes; this reverses in the final category but there are very few neotropical migrant species that prefer residential habitats.

Looking at the average (rather than median) change, in most cases the pattern is similar with larger increases for the non-NM species in all cases:

Habitat	NM	non-NM
Forest	+43%	+208%
Edge/Mosaic	+13%	+35%
Open	-13%	+48%
Residential	+40%	+79%

In the case of the averages, statistics can be applied; the one pattern that is (highly) significant is the very large increase among the non-neotropical migrant forest species in comparison to all other categories. In the cases of the other three habitats, the smaller differences between the NM and non-NM trends are not quite statistically significant. It is interesting, though, that they all fall in the same direction.

Overall, the predominant pattern is large increases (an average of three-fold) in non-neotropical migrant forest birds, and a smaller general increase in other species (including the neotropical migrants). Of course a large number of these species do not show simple monotonic trends up or down over the entire 45 year history of the BBS; this is the topic for the next post.

Tennessee BBS index:
Introduction
1: Waterfowl to Herons
2: Vultures to Doves
3: Cuckoos to Woodpeckers
4: Flycatchers to Corvids
5: Larks to Wrens
6: Gnatcatchers to Waxwings
7: Wood Warblers
8: Towhees to Buntings
9: Icterids to House Sparrow
Ups and Downs
Habitats
Next: Inflection Points
Summary