Jessica Gurevitch

Design and analysis of ecological experiments

Introduction: theories, hypotheses and statistics Exploratory data analysis and graphical display ANOVA: experiments in controlled environments ANOVA and ANCOVA: field competition experiments MANOVA: multiple response variables and multispecies interactions Repeated measures analysis: growth and other time-dependent measures Time-series intervention analysis: unreplicated large-scale experiments Non-linear curve fitting: predation and functional response curves Multiple regression: herbivory Path analysis: pollination Population sampling and bootstrapping in complex designs: demographic analysis Failure time analysis: emergence, flowering, survivorship and other waiting times The bootstrap and the jackknife: describing the precision of ecological indices Spatial statistics: analysis of field experiment Mantel tests: spatial structure in field experiments Model verification: optimal foraging Meta-analysis: combining the results of independent experiments References Author index Subject index

THE META‐ANALYSIS OF RESPONSE RATIOS IN EXPERIMENTAL ECOLOGY

Meta-analysis provides formal statistical techniques for summarizing the results of independent experiments and is increasingly being used in ecology. The response ratio (the ratio of mean outcome in the experimental group to that in the control group) and closely related measures of proportionate change are often used as measures of effect magnitude in ecology. Using these metrics for meta-analysis requires knowledge of their statistical properties, but these have not been previously derived. We give the approximate sampling distribution of the log response ratio, discuss why it is a particularly useful metric for many applications in ecology, and demonstrate how to use it in meta-analysis. The meta-analysis of response-ratio data is illustrated using experimental data on the effects of increased atmospheric CO2 on plant biomass responses.

STATISTICAL ISSUES IN ECOLOGICAL META‐ANALYSES

Meta-analysis is the use of statistical methods to summarize research findings across studies. Special statistical methods are usually needed for meta-analysis, both because effect-size indexes are typically highly heteroscedastic and because it is desirable to be able to distinguish between-study variance from within-study sampling-error variance. We outline a number of considerations related to choosing methods for the meta-analysis of ecological data, including the choice of parametric vs. resampling methods, reasons for conducting weighted analyses where possible, and comparisons fixed vs. mixed models in categorical and regression-type analyses.

A Meta-Analysis of Competition in Field Experiments

A meta-analysis was conducted on field-competition experiments published in six journals over a 10-yr period. We analyzed the effects of competition on the biomass of organisms belonging to 93 species in a wide variety of habitats. Competition had a large effect overall, with a great deal of heterogeneity in that effect among organisms. There were large differences among trophic levels in competitive effects, but the relative magnitude of competition at different trophic levels was contrary to the predictions of ecological theory. Primary producers and carnivores displayed small to medium effects. In these two groups, interspecific effects did not differ from intraspecific effects, nor did effects differ in terrestrial versus aquatic habitats. The effects of competition on herbivores ranged from large effects on anurans and lotic arthropods, to medium effects on marine mollusks and echinoderms, to effects that were not statistically distinguishable from zero for terrestrial arthropods. Interspecific competitive effects among these herbivore groups were generally less than intraspecific effects. Among primary producers, the effects of competition were not different in high- and low-productivity habitats. Across all taxa, large organisms did not experience greater competitive effects than small organisms, and competitive effects did not depend on the size attained in the absence of competitors. The effects of competition were weakly density-dependent in cases in which it was possible to examine the effects of the density of neighbors. Experiments conducted on caged organisms resulted in greater competitive effects than those with free-roaming and unenclosed organisms. Experiments with small sample sizes, short durations, and poor experimental design were more variable than experiments that were larger, longer, and better planned.

Global Warming and Terrestrial Ecosystems: A Conceptual Framework for Analysis

raise global mean temperature over the next century by 1.0–3.5 °C (Houghton et al. 1995, 1996). Ecologists from around the world have begun experiments to investigate the effects of global warming on terrestrial ecosystems, the aspect of global climate change that attracts the most public attention (Woodwell and McKenzie 1995, Walker and Steffen 1999). The effort to understand response to warming builds on a history of investigations of the effects of elevated CO 2 on plants and ecosystems (Koch and Mooney 1996, Schulze et al. 1999). There are important differences, however, between increases in atmospheric CO 2 and temperature change, both in the temporal and spatial patterns of change and in how they affect ecosystems. The scientists involved in temperature change research have had to face new technical and conceptual challenges in designing and interpreting their experiments (Schulze et al. 1999). In this paper we describe these challenges and present a conceptual framework for interpreting experimental results and predicting effects of warming on ecosystems.

Empirical approaches to quantifying interaction intensity: Competition and facilitation along productivity gradients

Quantitative synthesis across studies requires consistent measures of effect size among studies. In community ecology, these measures of effect size will often be some measure of the strength of interactions between taxa. However, indices of interaction strength vary greatly among both theoretical and empirical studies, and the connection between hypotheses about interaction strength and the metrics that are used to test these hypotheses are often not explicit. We describe criteria for choosing appropriate metrics and methods for comparing them among studies at three stages of designing a meta-analysis to test hypotheses about variation in interaction intensity: (1) the choice of response variable; (2) how effect size is calculated using the response in two treatments; and (3) whether there is a consistent quantitative effect across all taxa and systems studied or only qualitatively similar effects within each taxon–system combination. The consequences of different choices at each of these stages are illu...

The Interaction between Competition and Predation: A Meta‐analysis of Field Experiments

Ecologists working with a range of organisms and environments have carried out manipulative field experiments that enable us to ask questions about the interaction between competition and predation (including herbivory) and about the relative strength of competition and predation in the field. Evaluated together, such a collection of studies can offer insight into the importance and function of these factors in nature. Using a new factorial meta‐analysis technique, we combined the results of 20 articles reporting on 39 published field experiments to ask whether the presence of predators affects the intensity of competitive effects and to compare the average effects of competition and predation. Across all studies, the effects of competition in the presence of predators were less than in the absence of predators, and the interaction between competition and predation for most response variables was statistically significant. Removal of competitors had much more positive effects on organisms’ growth and mass than did exclusion of predators. Predator exclusion had much more beneficial effects on organisms’ survival than did competition. The mean effects of competition and predation on density did not differ from one another. The results differed among trophic levels. Further understanding would benefit greatly from more field experiments that manipulate both competition and predation, that focus on a wider range of organisms and environments, that focus on population‐level parameters such as density, and that report results more completely, including data such as sample sizes and variances.

Meta-analysis in ecology

Abstract Meta-analysis is the statistical synthesis of the results of separate studies. It was adapted from other disciplines for use in ecology and evolutionary biology beginning in the early 1990s, and, at the turn of the century, has begun to have a substantial impact on the way data are summarized in these fields. We identify 119 studies concerned with meta-analysis in ecology and evolution, the earliest published in 1991 and the most recent in 2000. We introduce the statistical methods used in modern meta-analysis with references to the well-developed literature in the field. These formal, statistically defensible methods have been established to determine average treatment effects across studies when a common research question is being investigated, to establish confidence limits around the average effect size, and to test for consistency or lack of agreement in effect size as well as explanations for differences in the magnitude of the effect among studies. Problems with popular but statistically flawed methods for the quantitative summary of research results have been pointed out, and their use is diminishing. We discuss a number of challenges and threats to the validity of meta-analysis in ecology and evolution. In particular, we examine how difficulties resulting from missing data, publication bias, data quality and data exclusion, non-independence among observations, and the combination of dissimilar data sets may affect the perceived utility of meta-analysis in these fields and the soundness of conclusions drawn from its application. We highlight particular applications of meta-analysis in ecology and evolution, discuss several controversies surrounding individual meta-analyses, and outline some of the practical issues involved in carrying out a meta-analysis. Finally, we suggest changes that would improve the quality of data synthesis in ecology and evolutionary biology, and predict future directions for this emerging enterprise.

INVASIVE SPECIES ACCELERATE DECOMPOSITION AND LITTER NITROGEN LOSS IN A MIXED DECIDUOUS FOREST

Invasive species can change decomposition rates within an ecosystem by changing the quality of the litter entering a system. It is not known, however, whether or not invasions can also change rates of decomposition irrespective of litter quality. We conducted an experiment to explore the differences in decomposition between leaf litter of native and exotic invasive woody plants and between invaded and uninvaded mesic hard- wood forests on Long Island, New York, USA. We evaluated the mass and nitrogen loss rates from leaf litter of four pairs of native and exotic woody species. Litter from the exotic species decomposed and released nitrogen significantly faster than litter from the native species. The largest differences in decomposition and nitrogen loss occurred between the invaded and uninvaded sites rather than between native and exotic species, with litter of all species types decomposing substantially faster in invaded sites. These results suggest that the invasion of exotic species into hardwood forests alters decomposition and nutrient cycling, irrespective of species-specific litter quality differences between natives and ex- otics.

Competition and the Local Distribution of the Grass Stipa Neomexicana

Within a grassland in southeastern Arizona, Stipa neomexicana occurs only on dry ridge crests with low total grass cover, while total grass cover is greater below the ridge crests in moister, low-lying areas. I hypothesized that Stipa neomexicana was limited to these dry ridges by competitive exclusion. This hypothesis was tested by removal experiments conducted at three positions along the topographic gradient. The responses of Stipa were compared with those ofAristida glauca and other neighboring species. The predictions made under the hypothesis were confirmed: competitors limited seedling establishment, seedling survival, flower production, and the growth of mature plants. The response to the removal of competitors increased from the ridge crest to the downslope positions. Competition depressed estimated finite rates of population increase for Stipa neomexicana, particularly on the lower slope. Thus, increasing competition from neighboring grasses along the topographic gradient was responsible for restricting Stipa neomexicana to the unfavorable ridge crest sites.