Search

Identifying species interactions and detecting when ecological communities are structured by them is an important problem in ecology and biogeography. Ecologists have developed specialized statistical hypothesis tests to detect patterns indicative of community‐wide processes in their field data. In this respect, null model approaches have proved particularly popular. The freedom allowed in choosing the null model and statistic to construct a hypothesis test leads to a proliferation of possible hypothesis tests from which ecologists can choose to detect these processes. Here, we point out some serious shortcomings of a popular approach to choosing the best hypothesis for the ecological problem at hand that involves benchmarking different hypothesis tests by assessing their performance on artificially constructed data sets. Terminological errors concerning the use of Type I and Type II errors that underlie these approaches are discussed. We argue that the key benchmarking methods proposed in the literature are not a sound guide for selecting null hypothesis tests, and further, that there is no simple way to benchmark null hypothesis tests. Surprisingly, the basic problems identified here do not appear to have been addressed previously, and these methods are still being used to develop and test new null models and summary statistics, from quantifying community structure (e.g., nestedness and modularity) to analyzing ecological networks. [ABSTRACT FROM AUTHOR]

Nonnative species that escape their native-range parasites may benefit not only from reduced infection pathology, but also from relaxed selection on costly immune defenses, promoting reallocation of resources toward growth or reproduction. However, benefits accruing from a reduction in defense could come at the cost of increased infection susceptibility. We conducted common garden studies of the shore crab Hemigrapsus sanguineus from highly parasitized native (Japan) populations and largely parasite-free invasive (USA) populations to test for differences in susceptibility to infection by native-range rhizocephalan parasites, and to explore differences in host resource allocation. Nonnative individuals showed at least 1.8 times greater susceptibility to infection than their native counterparts, and had reduced standing metabolic rates, suggesting that less of their energy was spent on physiological self-maintenance. Our results support an indirect advantage to parasite escape via the relaxation of costly physiological defenses. However, this advantage comes at the cost of heightened susceptibility, a trade-off of parasite escape that is seldom considered. [ABSTRACT FROM AUTHOR]

Statistical hypothesis testing has been widely criticized by ecologists in recent years. I review some of the more persistent criticisms of P values and argue that most stem from misunderstandings or incorrect interpretations, rather than from intrinsic shortcomings of the P value. I show that P values are intimately linked to confidence intervals and to differences in Akaike's information criterion (△AIC), two metrics that have been advocated as replacements for the P value. The choice of a threshold value of △AIC that breaks ties among competing models is as arbitrary as the choice of the probability of a Type I error in hypothesis testing, and several other criticisms of the P value apply equally to △AIC. Since P values, confidence intervals, and △AIC are based on the same statistical information, all have their places in modern statistical practice. The choice of which to use should be stylistic, dictated by details of the application rather than by dogmatic, a priori considerations. [ABSTRACT FROM AUTHOR]

The linear regression model, with its numerous extensions including multivar-iate ordination, is fundamental to quantitative research in many disciplines. However, spatial or temporal structure in the data may invalidate the regression assumption of independent residuals. Spatial structure at any spatial scale can be modeled flexibly based on a set of uncorrelated component patterns (e.g., Moran's eigenvector maps, MEM) that is derived from the spatial relationships between sampling locations as defined in a spatial weight matrix. Spatial filtering thus addresses spatial autocorrelation in the residuals by adding such component patterns (spatial eigenvectors) as predictors to the regression model. However, space is not an ecologically meaningful predictor, and commonly used tests for selecting significant component patterns do not take into account the specific nature of these variables. This paper proposes "spatial component regression" (SCR) as a new way of integrating the linear regression model with Moran's eigenvector maps. In its unconditioned form, SCR decomposes the relationship between response and predictors by component patterns, whereas conditioned SCR provides an alternative method of spatial filtering, taking into account the statistical properties of component patterns in the design of statistical hypothesis tests. Application to the well-known multivariate mite data set illustrates how SCR may be used to condition for significant residual spatial structure and to identify additional predictors associated with residual spatial structure. Finally, I argue that all variance is spatially structured, hence spatial independence is best characterized by a lack of excess variance at any spatial scale, i.e., spatial white noise. [ABSTRACT FROM AUTHOR]

Maximum entropy (maxent) models assign probabilities to states that (1) agree with measured macroscopic constraints on attributes of the states and (2) are otherwise maximally uninformative and are thus as close as possible to a specified prior distribution. Such models have recently become popular in ecology, but classical inferential statistical tests require assumptions of independence during the allocation of entities to states that are rarely fulfilled in ecology. This paper describes a new permutation test for such maxent models that is appropriate for very general prior distributions and for cases in which many states have zero abundance and that can be used to test for conditional relevance of subsets of constraints. Simulations show that the test gives correct probability estimates under the null hypothesis. Power under the alternative hypothesis depends primarily on the number and strength of the constraints and on the number of states in the model; the number of empty states has only a small effect, on power. The test is illustrated using two empirical data sets to test the community assembly model of B. Shipley, D. Vile, and E. Garnier and the species abundance distribution models of S. Pueyo, F. He, and T. Zillio. [ABSTRACT FROM AUTHOR]

Plant geographers have sought for decades to describe and predict the geographic distribution of vegetation types on the basis of plant function and its relationship with the abiotic environment. Traditionally this has been accomplished using categorical representations such as plant functional types. Increasingly, plant functional ecologists have sought to refine categorical functional types via quantitative functional traits in order to understand the ecological implications of trade-offs in plant form and function. Fewer, works have focused upon testing whether commonly measured functional traits enhance . our understanding of plant biogeography broadly and the geographic distribution of vegetation types in particular. Here we combine a continental-scale forest inventory data set containing 18111 plots with a plant functional trait data set to ask: (1) Is there a strong relationship between the abiotic environment and the distribution of functional trait values in forest inventory plots? And (2) can different Holdridge life zones be distinguished upon the basis of their functional trait distributions? The results show geographic patterns of functional. trait distributions that are often strongly correlated with climate and also show that the Holdridge life zones in the study area can be differentiated using a combination of functional traits. [ABSTRACT FROM AUTHOR]

We propose a framework for hypothesis-testing of stable isotope ratios in ecological studies. Statistical procedures are based on analysis of nested linear models and a residual permutation procedure (RPP) that is employed to evaluate probabilities associated with test statistics. We used simulated examples and a real data set to illustrate the utility and generality of the method. First, we developed a test for differences in centroid location and dispersion of δ13C and δ15N values within and among groups of isotopic data. Second, we evaluated magnitude and direction of change in centroid position (termed "path") of a pair of isotopic samples separated in space/time relative to paths of other paired sample sets. Third, we compared attributes of path trajectories (size, direction, and shape) over sample sets containing more than two samples to provide a quantitative description of how patterns of isotopic ratios change in response to spatial and temporal gradients. Examples are limited to the bivariate case (δ13C-δ15N biplots), but the statistical method can readily be applied to univariate and multivariate cases. [ABSTRACT FROM AUTHOR]

Male Neotropical orchid bees (Euglossini) collect volatile substances (fra- grances) from flowers and other sources (e.g., decaying wood) and store them in specialized hind tibial pockets. The accumulated chemicals are later emitted during courtship display, presumably to lure conspecific females for mating. We analyzed tibial fragrances of males of 15 sympatric Panamanian species in the genus Euglossa to test whether communities of euglossine bees are chemically structured, and to elucidate whether male fragrance signals evolve to convey premating isolation. Our analysis revealed substantial chemical disparity among all lineages. Disparity was mediated by compounds that were exclusive to certain species but also by differences in relative quantity of shared compounds. We mapped tibial fragrance compounds present in each species on a DNA-based phylogeny (reconstructed using partial sequences of COI, EFI-α, ArgK, and Pol-II) and found that most dominant compounds were highly homoplasious. In an analysis of chemical differentiation in relation to phylogenetic divergence through time, disparity was greater than, expected from a null model at any point during evolutionary history, suggesting that diversifying selection has shaped fragrance phenotypes. Notably, chemical disparity was greater within recently diverged lineages than among them, suggesting that chemical preferences in orchid bees evolved rapidly in the early stages of species divergence. We postulate communication interference as the possible mechanism behind the observed fragrance differentiation, which may be the product of reproductive character (fragrance) displacement. Our findings are consistent with the hypothesis that male fragrance signals evolve to convey premating isolation. [ABSTRACT FROM AUTHOR]

Hierarchical statistical models are increasingly being used to describe complex ecological processes. The data cloning (DC) method is a new general technique that uses Markov chain Monte Carlo (MCMC) algorithms to compute maximum likelihood (ML) estimates along with their asymptotic variance estimates for hierarchical models. Despite its generality, the method has two inferential limitations. First, it only provides Wald-type confidence intervals, known to be inaccurate in small samples. Second, it only yields ML parameter estimates, but not the maximized likelihood values used for profile likelihood intervals, likelihood ratio hypothesis tests, and information-theoretic model selection. Here we describe how to overcome these inferential limitations with a computationally efficient method for calculating likelihood ratios via data cloning. The ability to calculate likelihood ratios allows one to dó hypothesis tests, construct accurate confidence intervals and undertake information-based model selection with hierarchical models in a frequentist context. To demonstrate the use of these tools with complex ecological models, we reanalyze part of Gause's classic Paramecium data with state-space population models containing both environmental noise and sampling error. The analysis results include improved confidence intervals for parameters, a hypothesis test of laboratory replication, and a comparison of the Beverton-Holt and the Ricker growth forms based on a model selection index. [ABSTRACT FROM AUTHOR]

The article discusses issues raised by opposing reviews on the research conducted by Bradford A. Hawkins et al. on the metabolic theory of ecology (MTE). The discussion focuses on the structure of theories and the impact of changes in the epistemological framework on the relative strengths of MTE. Moreover, the article discusses the necessary assumptions and data statements in testing the validity of a theory, and this case the metabolic theory of ecology. Furthermore, it illustrates the process of hypothesis testing and model selection using MTE and the original theory presented by Andrew P. Allen in his research in 2002.

I argue that ecological data analyses are often needlessly complicated, and l present two examples of published analyses for which simpler alternatives are available. Unnecessary complexity is often introduced when analysts focus on subunits of the key experimental or observational units in a study, or use a very general framework to present an analysis that is a simple special case. Simpler analyses are easier to explain and understand; they clarify what the key units in a study are; they reduce the chances for computational mistakes; and they are more likely to lead to the same conclusions when applied by different analysts to the same data. [ABSTRACT FROM AUTHOR]

It has been actively discussed recently what statistical methods are appropriate when one is interested in testing hypotheses about the origin of beta diversity, especially whether one should use the raw-data approach (e.g., canonical analysis such as RDA and CCA) or the distance approach (e.g., Mantel test and multiple regression on distance matrices). Most of the confusion seems to stem from uncertainty as to what is the response variable in the different approaches. Here our aim is to clarify this issue. We also show that, although both the raw-data approach and the distance approach can often be used to address the same ecological hypothesis, they target fundamentally different predictions of those hypotheses. As the two approaches shed light on different aspects of the ecological hypotheses, they should be viewed as complementary rather than alternative ways of analyzing data. However, in some cases only one of the approaches may be appropriate. We argue that S. P. Hubbell's neutral theory can only be tested using the distance approach, because its testable predictions are stated in terms of distances, not in terms of raw data. In all cases, the decision on which method is chosen must be based on which addresses the question at hand, it cannot be based on which provides the highest proportion of explained variance in simulation studies. [ABSTRACT FROM AUTHOR]

Species loss can impact ecosystem functioning, but no general framework for analyzing these impacts exists. Here I derive a general partitioning of the effects of species loss on any ecosystem function comprising the summed contributions of individual species (e.g., primary productivity). The approach partitions the difference in ecosystem function between two sites (a ‘pre-loss’ site, and a ‘post-loss’ site comprising a strict subset of the species at the pre-loss site) into additive components attributable to different effects. The approach does not assume a particular experimental design or require monoculture data, making it more general than previous approaches. Using the Price Equation from evolutionary biology, I show that three distinct effects cause, ecosystem function to vary between sites: the ‘species richness effect’ (SRE; random loss of species richness), the ‘species composition effect’ (SCE; nonrandom loss of high- or low-functioning species), and the ‘context dependence effect’ (CDE; post-loss changes in the functioning of the remaining species). The SRE reduces ecosystem function without altering mean function per species. The SCE is analogous to natural selection in evolution. Nonrandom loss of, for example, high-functioning species will reduce mean function per species, and thus total function, just as selection against large individuals in an evolving population reduces mean body size in the next generation. The CDE is analogous to imperfect transmission in evolution. For instance, any factor (e.g., an environmental change) causing offspring to attain smaller body sizes than their parents (imperfect transmission) will reduce the mean body size in the next generation. Analogously, any factor causing the species remaining at the post-loss site to make smaller functional contributions than at the pre-loss site will reduce mean function per species, and thus total function. 1 use published data to illustrate how 'this new partition generalizes previous approaches, facilitates comparative analyses, and generates new empirical insights. In particular, the SCE often is less important than other effects. [ABSTRACT FROM AUTHOR]

Establishing relationships between species distributions and environmental characteristics is a major goal in the search for forces driving species distributions. Canonical ordinations such as redundancy analysis and canonical correspondence analysis are invaluable tools for modeling communities through environmental predictors. They provide the means for conducting direct explanatory analysis in which the association among species can be studied according to theft common and unique relationships with the environmental variables and other sets of predictors of interest, such as spatial variables. Variation partitioning can then be used to test and determine the likelihood of these sets of predictors in explaining patterns in community structure. Although variation partitioning in canonical analysis is routinely used in ecological analysis, no effort has been reported in the literature to consider appropriate estimators so that comparisons between fractions or, eventually, between different canonical models are meaningful. In this paper, we show that variation partitioning as currently applied in canonical analysis is biased. We present appropriate unbiased estimators. In addition, we outline a statistical test to compare fractions in canonical analysis. The question addressed by the test is whether two fractions of variation are significantly different from each other. Such assessment provides an important step toward attaining an understanding of the factors patterning community structure. The test is shown to have correct Type I error rates and good power for both redundancy analysis and canonical correspondence analysis. [ABSTRACT FROM AUTHOR]

A distance-based method is provided for the analysis and modeling of multivariate community data in response to a nonlinear gradient. Any reasonable dissimilarity measure can be used, and the method provides a natural extension from canonical analysis of principal coordinates (CAP) to nonlinear canonical analysis through the use of a link function, much like the extension of linear models to generalized linear models. The form of the nonlinear link function needs to be specified and will depend on the particular ecological system and the nature of the gradient. For example, an exponential decay curve could be used to model community structure after an environmental impact as a nonlinear function of time. This curve is used in our first example, where community structure is modeled as a nonlinear function of habitat size. Our second example uses a logistic curve to model change in community structure through a region of habitat transition from grassland to woodland. Computationally, this methodology uses a standard nonlinear optimization procedure to find the values of the parameters that maximize the correlation of the principal coordinates (obtained from an appropriately chosen distance measure) with the chosen form of nonlinear gradient. A simple randomization procedure is used to test the significance of the fitted nonlinear gradient over and above the fit of the linear gradient, and bootstrap confidence intervals for parameters are readily obtained. Any reasonable form of nonlinear gradient can be used, and it can be modeled as a nonlinear function of multiple environ- mental variables, making this a very flexible and versatile procedure for modeling multi-variate ecological systems. [ABSTRACT FROM AUTHOR]

We simulated the effects of missing information on statistical distributions of animal response that covaried with measured predictors of habitat to evaluate the utility and performance of quantile regression for providing more useful intervals of uncertainty in habitat relationships. These procedures were evaluated for conditions in which heterogeneity and hidden bias were induced by confounding with missing variables associated with other important processes, a problem common in statistical modeling of ecological phenomena. Simulations for a large (N = 10000) finite population representing grid lo- cations on a landscape demonstrated various forms of hidden bias that might occur when the effect of a measured habitat variable on some animal was confounded with the effect of another unmeasured variable. Quantile (0 ≤ τ ≤ 1) regression parameters for linear models that excluded the important, unmeasured variable revealed bias relative to parameters from the generating model. Depending on whether interactions of the measured and unmeasured variables were negative (interference interactions) or positive (facilitation interactions) in simulations without spatial structuring, either upper (τ > 0.5) or lower (τ < 0.5) quantile regression parameters were less biased than mean rate parameters. Heterogeneous, nonlinear response patterns occurred with correlations between the measured and unmeasured variables. When the unmeasured variable was spatially structured, variation in parameters across quantiles associated with heterogeneous effects of the habitat variable was reduced by modeling the spatial trend surface as a cubic polynomial of location co- ordinates, but substantial hidden bias remained. Sampling (n 20-300) simulations demonstrated that regression quantile estimates and confidence intervals constructed by inverting weighted rank score tests provided valid coverage of these parameters. Local forms of quantile weighting were required for obtaining correct Type I error rates and confidence interval coverage. Quantile regression was used to estimate effects of physical habitat resources on a bivalve (Macomona liliana) in the spatially structured landscape on a sandflat in a New Zealand harbor. Confidence intervals around predicted 0.10 and 0.90 quantiles were used to estimate sampling intervals containing 80% of the variation in densities in relation to bed elevation. Spatially structured variation in bivalve counts estimated by a cubic polynomial trend surface remained after accounting for the nonlinear effects of bed elevation, indicating the existence of important spatially structured processes that were not adequately represented by the measured habitat variables. [ABSTRACT FROM AUTHOR]

Under R. H. Whittaker's "multiplicative" perspective of species diversity, regional (γ) diversity is viewed as the product of within-assemblage (α) diversity and among-assemblage (β) diversity. An additive perspective, which views α and β as complements, has recently been endorsed for the study of spatial patterns in species diversity. Here we highlight the relation between the two measures of β diversity, showing the multiplicative beta; as a simple summary statistic of the relation between the additive components. We provide analytical solutions for approximate confidence intervals and hypothesis tests for both measures of beta diversity. We also present results of simulation studies that can help delimit the sample sizes needed for future studies, and which suggest the effect-size likely to be detected with samples of a given size. [ABSTRACT FROM AUTHOR]

The open population Cormack-Jolly-Seber (CJS) capture-mark--recapture model for estimating survival allows for random temporary emigration from the sampling area, but Markovian temporary emigration can bias estimates of survival.. We explore a multistate capture-recapture model that has been proposed for coping with Markovian temporary emigration. We provide a comprehensive assessment of the performance of this model using computer algebra and simulation. We found that most model parameters were identifiable unless survival, emigration, and immigration were all time dependent. Simulation results showed that intrinsically identifiable parameters were estimated without bias and that precision of survival estimates was always high. When temporary emigration was Markovian, precision of estimates of emigration, immigration, and recapture probabilities was acceptable; otherwise it was not. Test component 2.Ct of the goodness-of-fit test for the CJS model had good power to detect Markovian temporary emigration. We conclude that the multistate model works well when temporary emigration is Markovian (i.e., when the CJS model should not be used) and when survival and recapture probabilities are high. [ABSTRACT FROM AUTHOR]

The population dynamics of six beetle pests (Chrysomelidae: Galerucinae) were studied for three seasons in monocultures and polycultures of the corn-bean-squash agroecosystem in Costa Rica. The study was designed to determine if beetle abundance differed among these treatments, and if so, to determine whether the differences were caused primarily by predators and parasites (Enemies Hypothesis) or by movement patterns of adult beetles (Resource Concentration Hypothesis). In polycultures containing at least one nonhost plant, the numbers of beetles per unit host plant (squash or bean) were significantly lower relative to the numbers of beetles on host plants in monocultures. This pattern was generally observed for each of the beetle species studied between 40 and 60 d after planting and continued until the end of the season. When a given beetle species fed on both crop species in a diculture (i.e., the squash-bean diculture), the general polyculture effect was reversed: beetle numbers were significantly greater on one or both of the host plants in the diculture relative to the respective monocultures. These effects were apparently caused by changes in species richness of the plots per se rather than changes in host density or total plant density in the plot. There were no differences in rates of parasitism or predation of beetles between monocultures and polycultures. Neither were there significant differences in beetle emergence between treatments that could account for differences in adult beetle numbers between monocultures and polycultures. Instead, several lines of evidence show that different movement patterns of adult beetles were primarily responsible for the observed pattern of beetle abundance and distribution. Direct measurements of beetle movements in the field showed that beetles tended to emigrate more from polycultures that included a nonhost plant than from host monocultures. Field observations and experiments showed that this was due to several factors: (1) beetles avoided host plants shaded by corn; (2) cornstalks interfered in some way with flight movements of beetles; (3) as beetles moved through polycultures they remained on nonhost plants significantly less time than on host plants and their overall rate of movement in polycultures was higher. These data suggest that the Resource Concentration Hypothesis rather than the Enemies Hypothesis can account for differences in beetle abundance between monocultures and polycultures. The implication of this finding for predicting herbivore abundance in yet untried agreocosystems are discussed. It is suggested that it is not sufficient merely to recognize that herbivores are less common in particular polycultures. Practical issues of agroecosystem design and more theoretical questions of plant-herbivore coevolution depend on knowing exactly what ecological processes create reduced herbivore loads in polycultures. [ABSTRACT FROM AUTHOR]

Describes a statistical test for determining the diversity of a biological community that declines with time or along an environmental or other gradient. Test of the null hypothesis against the ordered alternative hypothesis; Maximum likelihood estimates; Observed value of the likelihood ratio statistic.

An introduction is presented in which the editor discusses various topics within the issue including the importance of P values and hypothesis testing data in statistical ecology and ecological applications and the views of associate professor Paul A. Murtaugh regarding statistical ecology.