Abundance variations within feeding guilds reveal ecological mechanisms behind avian species richness pattern along the elevational gradient of Mount Cameroon
Data files
Mar 22, 2023 version files 23.62 KB
-
Mt_Cameroon_bird_feeding_ecology_and_abundance.csv
-
README.md
Abstract
Two distinct diversity patterns are observed along tropical elevations: (a) decreasing number of species towards high elevations and (b) a hump-shaped pattern with the peak at mid-elevations. As diversity is likely supported by ecological capacity of the environment, decomposition of the overall richness into ecological facets and considering number of individuals within them is crucial for the proper understanding of richness patterns. We examined abundances of different avian guilds along the forested part of the elevational gradient on Mt. Cameroon. We (a) compared richness and abundance elevational patterns, (b) assessed the effective contribution of multiple guilds to richness and abundance patterns, and (b) assessed to which extent observed abundances of guilds differ from those expected by chance. We sampled birds in 2011–2015 during the dry season at seven elevations (30 m, 350 m, 650 m, 1100 m, 1500 m, 1850 m, 2200 m a.s.l.). For each assemblage, we estimated proportions of species and individuals that use particular diets, foraging modes, and feeding strata. We found that a rather decreasing pattern of species richness turns into a hump-shaped one if we look at the total abundances, implying different mechanisms behind these patterns. The number of species and individuals thus do not seem to be directly related, contrary to ‘the more-individuals hypothesis’. Abundances of foliage gleaners at mid-elevations, nectarivores at high elevations, and frugivores at low elevations deviated from random expectations. Our results suggest that parts of ecological space are filled partly separately by bird species and individuals along elevation of Mt. Cameroon.
Methods
Bird community sampling
Bird communities were sampled in 2011–2015 during the driest season (November - December) at seven forested elevations (30 m, 350 m, 650 m, 1100 m, 1500 m, 1850 m, 2200 m a.s.l.) using a standardized point count method (Bibby et al., 2000) and, additionally, random walks to capture the occurrence of rare species. Birds were recorded in the morning (6 am – 10 am) at 16 points at each elevation. The sample sites had a radius of 50 m and were located at least 150 m apart to diminish the risk of multiple counting of individuals while remaining logistically tractable for the study. Each point was visited on three different days, the order of points was changed during each visit to avoid biases due to daytime, and birds were recorded in three consecutive 5 min intervals during each visit. The maximum number of individuals of a species recorded at any of these 5 min intervals was recorded as its abundance at the given point, and the sum across all 16 points as its abundance at the given elevation (hereafter “elevational abundance”). We gave 1 estimate to singing individuals (1 pair) and a 0.5 individual estimate (half of the pair) to individuals that were just visually observed and were not singing. This is based on the assumption that singing individuals are usually accompanied by non-singing mates (Moller, 1991; Nowicki & Searcy, 2004). Duetting birds were counted as a pair, so 1 was assigned to a duetting pair. Additional random walks by three observers have been conducted for three days to capture rare species not recorded during regular point counts. The abundance assigned uniformly to these rare species was 0.5 for the given elevation as only vagrant and non-breeding individuals may be present. We are aware that our approach has its limitations, but it is rather conservative and probably underestimates the actual number of individuals in the forest. However, it focuses on reproductively active individuals and since we did it uniformly across the gradient, and do not expect a higher abundance of singing females in a specific elevation, the comparisons between elevations are informative. The taxonomy used in this study followed the IOC World Bird List version 4.2 (Gill & Donsker, 2014).
A distance sampling protocol was used, with birds recorded in five 10 m wide radial distance bands (Buckland et al., 2001). Here, we utilized the Distance v6.2 software (Thomas et al., 2010) to account for differences in species detectability and to estimate the number of unobserved individuals based on fitted detection functions (Buckland et al., 2001). Application of detection adjustments is, however, generally found problematic in tropical environments (Banks‐Leite et al., 2014) and we used the observed abundance estimates instead of the distance sampling-based estimates in further analyses. See Ferenc et al. (2016) and supplementary analyses therein for a detailed description of analyses on distance sampling-based data. Overall, estimates based on distance sampling provided worse results because (i) many rare species could not be included in the analyses, (ii) estimates derived from distance sampling were highly correlated with the observed abundance estimates, (iii) exceptions exist for which abundance estimates based on detection functions are unrealistically high. To standardize for number of individuals sampled we rarefied the observed diversities at particular elevations (Fig. S1)
Food and foraging niche estimates
The data on food and foraging niches were obtained from literature (Brown et al., 1982; Urban et al., 1986, 1997; Fry et al., 1988; Keith et al., 1992; Fry & Keith, 2000, 2004). The following niche dimensions (with categories) were recorded: food type – insectivores (insects and other invertebrates), carnivores (vertebrates), nectarivores (nectar and pollen), frugivores (fruits), and granivores (seeds); and foraging height classes - ground, low stratum, middle stratum, and canopy. Foraging behavior was categorized for insectivores as follows: (i) aerial feeding – catching prey actively in the air, (ii) foliage gleaning – gathering prey actively from leaves, mosses, twigs, and small branches, (iii) bark probing – gathering pray from branches and tree trunks, (iv) flycatching – waiting for prey and catching it in the air, (v), sit-and-wait – waiting for prey and catching it on the surface, (vi) ground search – gathering from the ground. We quantified the relative composition of the diet (% of diet or relative amount of time spent feeding on nectar) and relative use of foraging modes and foraging strata (% of time spent by each foraging mode and in particular strata while foraging).