Forest structure affects resource partitioning between pygmy and white-breasted nuthatches

Abstract

Ecologically similar, sympatric species will theoretically drive one or the other extinct unless those species can divide their resources. Resource partitioning is a common way to avoid competitive exclusion and these differences in resource utilization often represent the “ghost of competition past.” I studied how two closely related species, pygmy (Sitta pygmaea) and white-breasted (Sitta carolinensis) nuthatches, divide tree space. Furthermore, I tested whether pygmy and white-breasted nuthatches foraged in different areas on large- and small-diameter trees and whether pygmy nuthatches foraged more frequently on small-diameter trees. Finally, I tested whether each nuthatch species foraged less frequently in forests where most of the small-diameter trees have been removed. I found that in large trees, pygmy nuthatches foraged higher up in the tree and on the branches than white-breasted nuthatches. White-breasted nuthatches foraged low in the tree and on the trunk. In small-diameter trees, pygmy nuthatches foraged mostly on the lower trunk. White-breasted nuthatches foraged significantly less frequently on small-diameter trees than on large-diameter trees. Overall, pygmy nuthatches foraged more on small-diameter trees. Finally, pygmy nuthatches foraged significantly less frequently in areas of the forest where small-diameter trees have been selectively logged. My results concur with the literature on niche partitioning in forest-dwelling birds, where trees are divided spatially among species. Finally, my data may help explain historical niche relations between these species and how changes in forest structure can alter these niche dynamics.

Keywords:

• pygmy nuthatch
• white-breasted nuthatch
• resource partitioning
• niche
• ponderosa pine

Introduction

Competition for limiting resources can cause competitive exclusion (Gause 1932; Hardin 1960). To avoid extinction, competing species must coevolve ways to divide resources (MacArthur 1958; Hutchinson 1959; MacArthur & Levins 1967). Species that recently diverged via allopatry and are experiencing secondary contact exhibit character displacement (Lack 1947; Brown & Wilson 1956; Schluter 1995; Kawano 2002; Pfennig & Pfennig 2012). The observed phenotypic differences between these closely related species can be the result of “the ghost of competition past (Connell 1980).” The primary way in which competing species coexist is through resource partitioning (MacArthur 1958; Hutchinson 1961; Schoener 1974; Pyke 1982; Martin & Martin 2001; Young et al. 2010; Beaulieu & Sockman 2012). As a result, closely related and ecologically similar species that overlap in ranges partition their resources.

Resource partitioning occurs when competing species divide their resources in such a way that one species does not drive the other extinct. Resource partitioning can occur when species forage in different habitats/micro-habitats (MacArthur 1958; Hutchinson 1961; Alatalo & Alatalo 1979; Török 1990; Steen et al. 2007) and at different times (Bosakowski & Smith 1992; Lawler & Morin 1993; Kronfeld-Schor & Dayan 2003; Lara et al. 2009). Furthermore, coexisting species often use different hunting tactics that result in the capture of different prey items (Pulliam 1985; Kent 1986a, 1986b; Kiszka et al. 2011). Even in a seemingly homogenous environment, subtle environmental gradients and habitat heterogeneity often enable species to vary where they obtain their resources, at what time of day or season they obtain them, and offer them an array of resources, which can be accessed in different ways. As such, resource partitioning is an important evolutionary force shaping how competing species obtain their resources without causing competitive exclusion.

I examine how pygmy nuthatches (Sitta pygmaea) and white-breasted nuthatches (Sitta carolinensis) divide their resources. Pygmy and white-breasted nuthatches are closely related species that forage for insects on the bark of trees. Their ranges overlap considerably in the southwestern US. These two species differ in body size, with pygmy nuthatches the smaller of the pair (Kingery & Ghalambor 2001). Previous studies have shown that pygmy nuthatches forage higher up in the tree and use the branches more frequently than the trunk (Stallcup 1968; McEllin 1979). White-breasted nuthatches, on the other hand, forage more frequently lower in trees and more often on the trunks (Stallcup 1968; McEllin 1979). My study attempts to corroborate these findings and expands on them in two ways. First, I test whether pygmy nuthatches forage more frequently in small-diameter trees and whether they switch their primary foraging areas to the lower- and mid-trunk of small-diameter trees. Second, I utilize current forest practices that reduce fuel loads and improve forest health by thinning forests of small-diameter trees with selective logging and prescribed fires. I test whether pygmy nuthatches forage less frequently in parts of the forest that have been selectively logged and burned.

Study site and methods

I carried out this study from February to April 2008 at Schoolhouse Gulch and Groom Creek, 8 and 10 km south, respectively, from Prescott, Yavapai Co., in the central Arizona highlands. The Groom Creek site is untreated forest and consists of dense stands of ponderosa pine (Pinus ponderosa) with an understory of gambel oak (Quercus gambelii), Emory oak (Q. emoryi), and grey oak (Q. grisea). The Schoolhouse Gulch site has been selectively thinned for small-diameter trees and most of the understory trees were removed in 2006 and then small broadcast fires were burned in 2007. The Forest Service’s timber management plan called for no ponderosa pine trees larger than 45.7 cm inches be cut with an average spacing of 10.4 m between trees. The desired forest condition includes mid-aged trees (≈100 years), 36.6 m basal area, and a crown cover of less than 40%. The Schoolhouse Gulch site is more open and there are fewer trees with diameter less than 32 cm diameter at breast height (DBH) than the untreated site, making it less heterogeneous.

Ponderosa pine trees were divided into small and large groups, with small trees smaller or equal to 32 cm DBH and large trees with DBHs larger than 32 cm. I did this by walking transects during January 2007 and measuring the DBH of trees in which nuthatches were found foraging. From this data, I calculated that at 32 cm DBH, the probability that a pygmy nuthatch was foraging on the lower trunk of a larger tree was equal to the probability that a white-breasted nuthatch was foraging on the tree.

I divided ponderosa pine trees into five foraging zones: lower trunk, mid-trunk, lower branches, upper trunk, and upper branches (Figure 1). The lower trunk was defined as the trunk below any branches. The lower branches were defined as all the branches halfway between the first branch and the top of the tree, and the mid-trunk was defined as the trunk associated with this region. The upper branches were defined as all the branches above the lower branches, and the upper trunk as the trunk associated with this region. I measured the frequency of each foraging zone for each species in both the non-thinned and thinned treatments. In both treatments, I measured the frequency of nuthatches along three, 450 m transects that were 200 m apart. I performed observations on each transect once a week and each transect was done a total of nine times. The data for all three transects were compiled into a frequency of nuthatch sightings per hour.

Figure 1. Diagram showing how tree was divided into five foraging locations: lower trunk, mid-trunk, lower branches, upper trunk, and upper branches.

To test for differences in foraging position of the tree between species, I used a general linear mixed effects model where species, foraging location, and species–foraging location interaction were the fixed factors and forest condition was a random effect to control for differences between sites (Crawley 2007). To compare how frequently each species was foraging in thinned and unthinned forests, I used an ANOVA and a Tukey’s honest significant difference to perform post hoc tests. To test for differences in foraging position within each species in small and large trees and between forest conditions, I used a general linear model on each species separately where tree size, forest condition, foraging position, and their interactions were the fixed factors (Crawley 2007). I performed a natural log + 1 transformation of these data (Crawley 2007). I ran an independent samples t-test to compare the DBHs of trees in which pygmy and white-breasted nuthatches were found foraging. All statistical tests were done in R (R Core Team 2012).

Results

White-breasted and pygmy nuthatches differed where they forage in trees and on the size of trees on which they foraged. Compared to pygmy nuthatches, white-breasted nuthatches foraged significantly more on the lower trunk (Figures 2(A) and (B) and 3(A) and (B); t = 6.32, df = 39, 320, p < 0.001), the mid-trunk (t = 4.04, df = 39, 320, p < 0.001), and marginally more frequently on the upper trunk (t = 1.90, df = 39, 320, p = 0.058) than did pygmy nuthatches. Pygmy nuthatches foraged significantly more in the upper branches compared to white-breasted nuthatches (t = –2.59, df = 349, p = 0.01). Pygmy nuthatches (n = 26) more often foraged in small-diameter trees than white-breasted nuthatches (n = 26; t = 10.306, df = 47, p < 0.001). The mean DBH of trees in which pygmy nuthatches were found foraging was 21.3 cm (SD = ± 8.2 cm), while that of white-breasted nuthatches was 48.4 cm (±10.6). Pygmy nuthatches were more frequent than white-breasted nuthatches, both in thinned and unthinned forests (F = 7.684, df = 1,32, p = 0.00921). Tukey’s honest significant difference test revealed that pygmy nuthatches were significantly more frequent than white-breasted nuthatches in thinned forests (p < 0.001) and unthinned forests (p = 0.017826).

Figure 2. (A) Mean frequency per hour in different positions on the tree in large-diameter trees in non-thinned forest. (B) Mean frequency per hour in different positions in small-diameter trees in non-thinned forest.

Figure 3. (A) Mean frequency per hour in different positions on the tree in large-diameter trees in thinned forest. (B) Mean frequency per hour in different positions in small-diameter trees in thinned forest.

When considering white-breasted nuthatches independently, white-breasted nuthatches foraged the most on the lower trunk (Figures 2(A) and (B) and 3(A) and (B); t = 6.97, df = 14, 165, p < 0.001) and very little on the upper trunk (t = –6.87, df = 14, 165, p < 0.001) and upper branches (t = –6.66, df = 14, 165, p < 0.001). White-breasted nuthatches foraged significantly less on the upper branches of small trees than on the upper branches of large trees (t = 5.09, df = 14, 165, p < 0.001). Likewise, white-breasted nuthatches foraged more on the upper trunk of large-diameter trees compared to small-diameter trees (t = 5.45, df = 14, 165, p < 0.001). There was no significant difference in foraging frequencies on the lower trunk (t = –1.55, df = 14, 165, p = 0.123) or mid-trunk (t = 0.58, df = 14, 165, p = 0.552). White-breasted nuthatches foraged less frequently in small-diameter trees (t = –7.201, df = 14, 165, p < 0.001), but not less frequently in unthinned forest than in the thinned forest (t = 1.26, df = 14, 165, p = 0.209). Finally, white-breasted nuthatches foraged less frequently on the lower trunk in unthinned forests than in thinned forests (t = –2.95, df = 14, 165, p = 0.004), but not in the other locations in the tree (mid-trunk: t = 0.48, df = 14, 165, p = 0.632; upper trunk: t = 1.25, df = 14, 165, p = 0.213; branches: t = 0.89, df = 14, 165, p = 0.374).

When considering pygmy nuthatches independently, pygmy nuthatches foraged significantly more often in the branches (Figures 2(A) and (B) and 3(A) and (B); t = 4.71, df = 14, 165, p < 0.001) than on the lower trunk (t = –7.96, df = 14, 165, p < 0.001), mid-trunk (t = –7.19, df = 14, 165, p < 0.001), and upper trunk (t-5.02, df = 14, 165, p < 0.001). Pygmy nuthatches used the lower and mid-trunk sections significantly more often on small-diameter trees than on large-diameter trees (Figures 2(A) and (B) and 3(A) and (B); lower trunk: t = 7.69, df = 14, 165, p < 0.001; mid-trunk: t = 4.81, df = 14, 165, p < 0.001) and foraged significantly less in the branches of small-diameter trees than large-diameter trees (t = –7.37, df = 14, 165, p < 0.001). There was no difference in how frequently pygmy nuthatches foraged in upper trunk between small- and large-diameter trees (t = 0.58, df = 14, 165, p = 0.564). In unthinned forests, pygmy nuthatches were significantly more frequent on small-diameter trees than large-diameter trees (t = –3.24, df = 14, 165, p = 0.001). Pygmy nuthatches were significantly less frequent in thinned areas of the forest (t = –2.006, df = 14, 165, p = 0.046). Pygmy nuthatches showed no significant differences in foraging position between thinned and unthinned forests (lower trunk: t = –0.19, df = 14, 165, p = 0.853; mid-trunk: t = 0.26, df = 14, 165, p = 0.796; upper trunk: t = –0.06, df = 14, 165, p = 0.949: branches: t = –0.26, df = 14, 165, 0.792).

Discussion

My results corroborated those of Stallcup (1968) and McEllin (1979), showing that sympatric pygmy and white-breasted nuthatches partition their resources by foraging in different parts of the tree (Figures 2(A) and (B) and 3(A) and (B)). Overall, pygmy nuthatches foraged higher in the tree and on the branches. White-breasted nuthatches foraged mostly lower in the tree and on the trunk (Figures 2(A) and (B) and 3(A) and (B)). My results additionally show several other aspects of niche partitioning between pygmy and white-breasted nuthatches. First, pygmy and white-breasted nuthatches forage differently on large- and small-diameter trees. Pygmy nuthatches move their foraging to the lower tree and trunk on smaller diameter trees (Figure 3(A) and (B)). White-breasted nuthatches are much less common on smaller diameter trees (Figure 3(A) and (B)). Second, pygmy nuthatches forage more frequently on smaller diameter trees. Finally, when few small-diameter trees are present, as in the selectively logged area, pygmy nuthatches foraged less frequently than in the unlogged forest.

The rugosity of the bark may explain why these two nuthatch species forage in different places in the tree. In general, the bark is less rugged at the top of larger trees and on the smaller, top-most branches. Likewise, the bark on small-diameter trees is less rugged. Pygmy nuthatches have smaller beaks than white-breasted nuthatches (Kingery & Ghalambor 2001). The smaller beaks of pygmy nuthatches may allow them to forage more efficiently on less rugged bark, whereas the larger beaked white-breasted nuthatch may forage more efficiently on more rugged bark. I did not measure bark rugosity and therefore any relationship between it and beak size remains speculative. However, several studies have shown that smaller beaked birds forage more efficiently on smaller seeds than larger beaked birds, and larger beaked birds forage more efficiently on larger seeds (Pulliam 1985; Grant & Grant 2006). The relationship between different beak sizes and foraging positions between pygmy and white-breasted nuthatches may be a similar pattern of resource partitioning as in seed-foraging birds.

There are several other ways in which body size can affect resource partitioning between these two nuthatches. First, white-breasted nuthatches are larger birds and therefore have a higher energy requirement. Insects and arthropods found on the trunk may be larger than insects and arthropods found on the branches. Future work should investigate this. Second, white-breasted nuthatches, as the larger species, may deter pygmy nuthatches from foraging on the trunk of larger trees. Body size has long been suspected as a mechanism by which closely related species can coexist (Hutchinson 1959; Leyequién et al. 2007). However, pygmy nuthatches are considered the competitively superior species in ponderosa pine forests because they outnumber all other species (Kingery & Ghalambor 2001). Finally, white-breasted nuthatches may avoid branches, and especially small twigs, because the branches may not hold their greater weight. This is unlikely because white-breasted nuthatches foraged on the trunks more than on large branches (which were usually the lower branches) and because it would not explain why pygmy nuthatches avoid the lower trunk on large-diameter trees.

Dividing trees spatially is a widespread way in which ecologically similar and closely related forest-dwelling birds partition their resources (MacArthur 1958; Stallcup 1968; Holmes et al. 1979; McEllin 1979; Morrison & With 1987; Török 1990). MacArthur’s (1958) original study on niche partitioning in warblers showed that different species foraged at different heights and at different distances from the trunk. Like the two species studied here, many studies show birds that are closely related and most likely have competed for resources in their evolutionary history (Morrison & With 1987; Török 1990). Foraging in different parts of the tree and any morphological attributes that make foraging in those places more efficient, such as beak size, may represent an ancient character displacement that resulted from “the ghost of competition past (Connell 1980).”

My study also aids in understanding niche relationships in an environment heavily modified by humans. Ponderosa pine forests in Arizona have undergone extensive changes since European descendents settled the area 150 years ago (Savage & Swetnam 1990; Covington & Moore 1994; Swetnam et al. 1999). Ponderosa pine forests were much more open, with few small-diameter trees, a carpet of grasses, and frequent, low-intensity fires (Moore et al. 1999). The current practice of selective logging and prescribed burning attempts to restore forests to these pre-settlement conditions (Moore et al. 1999; Fulé et al. 2005). Historical relations between pygmy and white-breasted nuthatches in Arizona most likely were similar to how they were in the selectively logged forest, where pygmy nuthatches probably foraged high up in the tree and further from the trunk and white-breasted nuthatches foraged lower in the tree and on the trunk. According to my results, humans have inadvertently favored pygmy nuthatches through forest transformations that have increased the proportion of small-diameter trees. Therefore, pygmy nuthatches may have been less frequent in pre-settlement times. It is unlikely that current forest practices will seriously harm pygmy nuthatch populations for a couple of reasons. First, pygmy nuthatches are found foraging more frequently than white-breasted nuthatches, in both thinned and unthinned forests. Second, pygmy nuthatches still have a foraging niche in the thinned forests that is distinct from white-breasted nuthatches, allowing them ecological space to avoid any potential population crashes that could make them threatened or endangered.

Acknowledgements

I would like to thank Tom Fleischner for mentoring this project and Perri Eason and a reviewer for insightful comments on the manuscript.