Skip to main content
. 2021 Feb 9;16(2):e0245695. doi: 10.1371/journal.pone.0245695

Table 3. Chi-square test results by tree species from bat roost site selection study conducted in February-March 2019 and December-January 2019–2020 at Camp Blanding, Clay County, Florida.

Residuals Contributions Use vs. Available
Tree Species Roost Quadrant Roost Quadrant
Loblolly bay Gordonia lasianthus 1.74 -0.89 5.91 1.55 Greater
Sweetbay magnolia Magnolia virginiana -0.22 0.11 0.1 0.03 Less
Red bay Persea borbonia 0.22 -0.11 0.1 0.03 Greater
Slash pine Pinus elliottii 1.61 -0.82 5.04 1.32 Greater
Longleaf pine Pinus palustris 2.75 -1.41 14.81 3.88 Greater
Loblolly pine Pinus taeda 3.31 -1.69 21.36 5.6 Greater
Turkey oak Quercus laevis -1.48 0.76 4.3 1.13 Less
Water oak Quercus nigra -1.65 0.84 5.32 1.39 Less
Other1 -3.38 1.73 22.32 5.85 Less
χ2 51.17
p-value <0.001

1Other = bald cypress (Taxodium distichum), boxelder (Acer negundo), red maple (Acer rubrum), American hornbeam (Carpinus caroliniana), common persimmon (Diospyros virginiana), Carolina holly (Ilex ambigua), sweetgum (Liquidambar styraciflua), tulip tree (Liriodendron tulipifera), rusty staggerbush (Lyonia ferruginea), Southern magnolia (Magnolia grandiflora), black cherry (Prunus serotina), swamp white oak (Quercus bicolor), turkey oak (Quercus cerris), Shumard oak (Quercus shumardii), post oak (Quercus stellata), live oak (Quercus virginiana), winged elm (Ulmus alata), and American elm (Ulmus americana).

Residuals with values greater than two denotes a major influence on the chi-square test statistic. Contributions denote the difference between expected and observed values with larger contributions signifying greater difference.