Phylogenetic uncertainty and the inference of patterns in community ecology and comparative studies.

Progress in phylogenetic community ecology is often limited by the availability of phylogenetic information and the lack of appropriate methods and solutions to deal with this problem. We estimate the effect of the lack of phylogenetic information on the relations among taxa measured by commonly used phylogenetic metrics in comparative studies and community ecology, namely: Blomberg's K phylogenetic signal, Faith's Phylogenetic Diversity (PD), Mean Phylogenetic Distance (MPD) and Mean Nearest Taxon Distance (MNTD). To overcome this problem, we tested two possible solutions: Polytomic trees and Operational trees. Our results show that the effects on K values strongly depended on the level of phylogenetic signal. In the case of the community metrics, the effects were insensitive to the patterns of species distribution in the communities. Community metrics tended to be overestimated with both Polytomic and Operational trees, but the overestimation was higher with Polytomic trees. PD and MPD metrics were less biased than MNTD metric. We show that the lack of phylogenetic resolution is not necessarily problematic for all analyses and that its effect will depend on the chosen metric and on the solutions used to deal with the problem. Based on our results, we suggest that ecologists should prefer the Operational tree solution to remove polytomies in the phylogenetic tree and take careful consideration while designing experiments, and analyzing and interpreting the results of phylogenetic metrics.

The influence of spatial sampling scales on ant-plant interaction network architecture.

Despite great interest in metrics to quantify the structure of ecological networks, the effects of sampling and scale remain poorly understood. In fact, one of the most challenging issues in ecology is how to define suitable scales (i.e., temporal or spatial) to accurately describe and understand ecological systems. Here, we sampled a series of ant-plant interaction networks in the southern Brazilian Amazon rainforest in order to determine whether the spatial sampling scale, from local to regional, affects our understanding of the structure of these networks. To this end, we recorded ant-plant interactions in adjacent 25 × 30 m subplots (local sampling scale) nested within twelve 250 × 30 m plots (regional sampling scale). Moreover, we combined adjacent or random subplots and plots in order to increase the spatial sampling scales at the local and regional levels. We then calculated commonly used binary and quantitative network-level metrics for both sampling scales (i.e., number of species and interactions, nestedness, specialization and modularity), all of which encompass a wide array of structural patterns in interaction networks. We observed increasing species and interactions across sampling scales, and while most network descriptors remained relatively constant at the local level, there was more variation at the regional scale. Among all metrics, specialization was most constant across different spatial sampling scales. Furthermore, we observed that adjacent assembly did not generate more variation in network descriptor values compared to random assembly. This finding indicates that the spatially aggregated distribution of species/individuals and abiotic conditions does not affect the organization of these interacting assemblages. Our results have a direct impact on our empirical and theoretical understanding of the ecological dynamics of species interactions by demonstrating that small spatial sampling scales should suffice to record some patterns commonly found in ant-plant interaction networks in a highly diverse tropical rainforest.

Analyzing community-weighted trait means across environmental gradients: should phylogeny stay or should it go?

Functional traits mediate ecological responses of organisms to the environment, determining community structure. Community-weighted trait means (CWM) are often used to characterize communities by combining information on species traits and distribution. Relating CWM variation to environmental gradients allows for evaluating species sorting across the metacommunity, either based on correlation tests or ordinary least squares (OLS) models. Yet, it is not clear if phylogenetic signal in both traits and species distribution affect those analyses. On one hand, phylogenetic signal might indicate niche conservatism along clade evolution, reinforcing the environmental signal in trait assembly patterns. On the other hand, it might introduce phylogenetic autocorrelation to mean trait variation among communities. Under this latter scenario, phylogenetic signal might inflate type I error in analysis relating CWM variation to environmental gradients. We explore multiple ways phylogenetic history may influence analysis relating CWM to environmental gradients. We propose the concept of neutral trait diffusion, which predicts that for a functional trait x, CWM variation among local communities does not deviate from the expectation that x evolved according to a neutral evolutionary process. Based on this framework we introduce a graphical tool called neutral trait diffusion representation (NTDR) that allows for the evaluation of whether it is necessary to carry out phylogenetic correction in the trait prior to analyzing the association between CWM and environmental gradients. We illustrate the NTDR approach using simulated traits, phylogenies and metacommunities. We show that even under moderate phylogenetic signal in both the trait used to define CWM and species distribution across communities, OLS models relating CWM variation to environmental gradients lead to inflated type I error when testing the null hypothesis of no association between CWM and environmental gradient. To overcome this issue, we propose a phylogenetic correction for OLS models and evaluate its statistical performance (type I error and power). Phylogeny-corrected OLS models successfully control for type I error in analysis relating CWM variation to environmental gradients but may show decreased power. Combining the exploratory tool of NTDR and phylogenetic correction in traits, when necessary, guarantees more precise inferences about the environmental forces driving trait-mediated species sorting across metacommunities.

Influences of sampling effort on detected patterns and structuring processes of a Neotropical plant-hummingbird network.

Virtually all empirical ecological interaction networks to some extent suffer from undersampling. However, how limitations imposed by sampling incompleteness affect our understanding of ecological networks is still poorly explored, which may hinder further advances in the field. Here, we use a plant-hummingbird network with unprecedented sampling effort (2716 h of focal observations) from the Atlantic Rainforest in Brazil, to investigate how sampling effort affects the description of network structure (i.e. widely used network metrics) and the relative importance of distinct processes (i.e. species abundances vs. traits) in determining the frequency of pairwise interactions. By dividing the network into time slices representing a gradient of sampling effort, we show that quantitative metrics, such as interaction evenness, specialization (H2 '), weighted nestedness (wNODF) and modularity (Q; QuanBiMo algorithm) were less biased by sampling incompleteness than binary metrics. Furthermore, the significance of some network metrics changed along the sampling effort gradient. Nevertheless, the higher importance of traits in structuring the network was apparent even with small sampling effort. Our results (i) warn against using very poorly sampled networks as this may bias our understanding of networks, both their patterns and structuring processes, (ii) encourage the use of quantitative metrics little influenced by sampling when performing spatio-temporal comparisons and (iii) indicate that in networks strongly constrained by species traits, such as plant-hummingbird networks, even small sampling is sufficient to detect their relative importance for the frequencies of interactions. Finally, we argue that similar effects of sampling are expected for other highly specialized subnetworks.

SYNCSA--R tool for analysis of metacommunities based on functional traits and phylogeny of the community components.

SUMMARY: SYNCSA is an R package for the analysis of metacommunities based on functional traits and phylogeny of the community components. It offers tools to calculate several matrix correlations that express trait-convergence assembly patterns, trait-divergence assembly patterns and phylogenetic signal in functional traits at the species pool level and at the metacommunity level. AVAILABILITY AND IMPLEMENTATION: SYNCSA is a package for the R environment, under a GPL-2 open-source license and freely available on CRAN official web server for R (http://cran.r-project.org). CONTACT: vanderleidebastiani@yahoo.com.br.