Explore the vast range of titles available.

Recognition of the scale dependence of ecological processes helps explain the distribution and abundance of organisms. In 1687, Newton reported that the same laws could describe Galileo's data on balls rolling down ramps and Brahe's data on planets moving around the Sun ( 1 ). This observation implied that a finite list of principles could explain our infinite universe. And it inspired a leap across scales: The rules at human scales are not unique. Newton's laws of motion are still the dominant explanatory tool across scales ranging from a few atoms to solar systems. However, over the past 25 years, ecologists have come to realize that, unlike physics, ecology is scale-dependent ( 2 – 4 ). In a recent paper, Gotelli, Graves, and Rahbek ( 5 ) highlight the importance of this scale dependence: They show that a process that occurs at small spatial scales, namely competition between individuals, plays an important role even at the large scale of an entire country.

Macroecology is a growing and important subdiscipline of ecology, but it is becoming increasingly diffuse, without an organizing principle that is widely agreed upon. I highlight two main current views of macroecology: as the study of large-scale systems and as the study of emergent systems. I trace the history of both these views through the writings of the founders of macroecology. I also highlight the transmutation principle that identifies serious limitations to the study of large-scale systems with reductionist approaches. And I suggest that much of the underlying goal of macroecology is the pursuit of general principles and the escape from contingency. I highlight that there are many intertwined aspects of macroecology, with a number of resulting implications. I propose that returning to a focus on studying assemblages of a large number of particles is a helpful view. I propose defining macroecology as “the study at the aggregate level of aggregate ecological entities made up of large numbers of particles for the purposes of pursuing generality”.

Distribution models are used to predict the likelihood of occurrence or abundance of a species at locations where census data are not available. An integral part of modelling is the testing of model performance. We compared different schemes and measures for testing model performance using 79 species from the North American Breeding Bird Survey. The four testing schemes we compared featured increasing independence between test and training data: resubstitution, random data hold-out and two spatially segregated data hold-out designs. The different testing measures also addressed different levels of information content in the dependent variable: regression R² for absolute abundance, squared correlation coefficient r² for relative abundance and AUC/Somer s D for presence/absence. We found that higher levels of independence between test and training data lead to lower assessments of prediction accuracy. Even for data collected independently, spatial autocorrelation leads to dependence between random hold-out test data and training data, and thus to inflated measures of model performance. While there is a general awareness of the importance of autocorrelation to model building and hypothesis testing, its consequences via violation of independence between training and testing data have not been addressed systematically and comprehensively before. Furthermore, increasing information content (from correctly classifying presence/absence, to predicting relative abundance, to predicting absolute abundance) leads to decreasing predictive performance. The current tests for presence/absence distribution models are typically overly optimistic because a) the test and training data are not independent and b) the correct classification of presence/absence has a relatively low information content and thus capability to address ecological and conservation questions compared to a prediction of abundance. Meaningful evaluation of model performance requires testing on spatially independent data, if the intended application of the model is to predict into new geographic or climatic space, which arguably is the case for most applications of distribution models.

1. Plant phenotypic diversity is shaped by the interplay of trade-offs and constraints in evolution. Closely integrated groups of traits (i.e. trait dimensions) are used to classify plant phenotypic diversity into plant strategies, but we do not know the degree of interdependence among trait dimensions. To assess how selection has shaped the phenotypic space, we examine whether trait dimensions are independent. 2. We gathered data on saplings of 24 locally coexisting tree species in a temperate forest, and examined the correlation structure of 20 leaf, branch, stem and root traits. These traits fall into three well-established trait dimensions (the leaf economic spectrum, the wood spectrum and Corner's Rules) that characterize vital plant functions: resource acquisition, sap transport, mechanical support and canopy architecture. Using ordinations, network analyses and Mantel tests, we tested whether the sapling phenotype of these tree species is organized along independent trait dimensions. 3. Across species, the sapling phenotype is not structured into clear trait dimensions. The trait relationships defining trait dimensions are either weak or absent and do not dominate the correlation structure of the sapling phenotype as a whole. Instead traits from the three commonly recognized trait dimensions are organized into an integrated trait network. The effect of phylogeny on trait correlations is minimal. 4. Our results indicate that trait dimensions apparent in broad-based interspecific surveys do not hold up among locally coexisting species. Furthermore, architectural traits appear central to the phenotypic network, suggesting a pivotal role for branching architecture in linking resource acquisition, mechanical support and hydraulic functions. 5. Synthesis. Our study indicates that local and global patterns of phenotypic integration differ and calls into question the use of trait dimensions at local scales. We propose that a network approach to assessing plant function more effectively reflects the multiple trade-offs and constraints shaping the phenotype in locally co-occurring species.

Trait-based approaches have taken an increasingly dominant role in community ecology. Although trait-based strategy dimensions such as the leaf economic spectrum (LES) have been identified primarily at global-scales, trait variation at the community scale is often interpreted in this context. Here we argue from several lines of evidence that a research priority should be to determine whether global-scale trait relationships hold at more local scales. We review recent literature assessing trait variation at smaller scales, and then present a case study exploring the relationship between the correlation strength of leaf traits and their similarity in variation structure across ecological scales. We find that the correlation strength between pairs of leaf traits does not predict whether the traits respond similarly to different drivers of variation. Instead, correlation strength only sets an upper bound to the dissimilarity in trait variation structure. With moderate correlation strengths, LES traits largely retain the ability to respond independently to different drivers of phenotypic variation at different scales. Recent literature and our results suggest that LES relationships may not hold at local scales. Clarifying under what conditions and at which scales the LES is consistently expressed is necessary for us to make the most of the emerging trait toolbox.

While human activities are known to elicit rapid turnover in species composition through time, the properties of the species that increase or decrease their spatial occupancy underlying this turnover are less clear. Here, we used an extensive dataset of 238 metacommunity time series of multiple taxa spread across the globe to evaluate whether species that are more widespread (large-ranged species) differed in how they changed their site occupancy over the 10–90 years the metacommunities were monitored relative to species that are more narrowly distributed (small-ranged species). We found that on average, large-ranged species tended to increase in occupancy through time, whereas small-ranged species tended to decrease. These relationships were stronger in marine than in terrestrial and freshwater realms. However, in terrestrial regions, the directional changes in occupancy were less extreme in protected areas. Our findings provide evidence for systematic decreases in occupancy of small-ranged species, and that habitat protection could mitigate these losses in the face of environmental change. Whether a species declines under the current biodiversity crisis could partly depend on its range size. Here, the authors use replicated metacommunity data to identify global patterns in the relationship between species’ range size and changes in occupancy through time.

The future of temperate forests in the face of global change and anthropogenic stressors remains uncertain. The regeneration stage, which is a critical bottleneck for many organisms, is a key indicator of forest health, future canopy composition and forest adaptive capacity. In trees, seemingly healthy forests can be at long‐term risk due to insufficient juveniles to replace them (regeneration failure), or compositional differences between juveniles and adults (regeneration mismatch). We propose ‘regeneration debt’ to collectively describe regeneration failure and mismatch in analogy to extinction debt. To demonstrate this concept, we conducted a macroecological analysis of regeneration debt and anthropogenic stressors in eastern US forests. Using U.S. Forest Service‐Forest Inventory and Analysis data, we quantified regeneration debt in 18 states from Maine to South Carolina, and evaluated the influence of site, anthropogenic stressors and climate drivers in the most affected regions. We identified three distinct regions, with little debt in the north, moderate debt in the south and severe regeneration debt in the central, mid‐Atlantic region. In this region, multiple anthropogenic stressors (invasive plants, deer overabundance and land use) were associated with both low‐regeneration abundance and the prevalence of disease‐prone and/or suboptimal species. Synthesis and applications. Without management intervention, the severe regeneration debt in the mid‐Atlantic region will likely lead to long‐term declines in forest cover, with cascading negative effects on forest‐dependent taxa and ecosystem services. Moreover, the location of the regeneration debt, which is at the northern edge of and involves many of the tree species that are predicted to gain suitable habitat in the Northeastern US, has consequences that extend far beyond its current geographic extent. In fact, this regeneration debt may already be functioning as a barrier to poleward tree migration. Our results demonstrate the value of regeneration debt as an indicator of ecosystem health and forest adaptive capacity. Without management intervention, the severe regeneration debt in the mid‐Atlantic region will likely lead to long‐term declines in forest cover, with cascading negative effects on forest‐dependent taxa and ecosystem services. Moreover, the location of the regeneration debt, which is at the northern edge of and involves many of the tree species that are predicted to gain suitable habitat in the Northeastern US, has consequences that extend far beyond its current geographic extent. In fact, this regeneration debt may already be functioning as a barrier to poleward tree migration. Our results demonstrate the value of regeneration debt as an indicator of ecosystem health and forest adaptive capacity.

The objective of science is to understand the natural world; we argue that prediction is the only way to demonstrate scientific understanding, implying that prediction should be a fundamental aspect of all scientific disciplines. Reproducibility is an essential requirement of good science and arises from the ability to develop models that make accurate predictions on new data. Ecology, however, with a few exceptions, has abandoned prediction as a central focus and faces its own crisis of reproducibility. Models are where ecological understanding is stored and they are the source of all predictions – no prediction is possible without a model of the world. Models can be improved in three ways: model variables, functional relationships among dependent and independent variables, and in parameter estimates. Ecologists rarely test to assess whether new models have made advances by identifying new and important variables, elucidating functional relationships, or improving parameter estimates. Without these tests it is difficult to know if we understand more today than we did yesterday. A new commitment to prediction in ecology would lead to, among other things, more mature (i.e. quantitative) hypotheses, prioritization of modeling techniques that are more appropriate for prediction (e.g. using continuous independent variables rather than categorical) and, ultimately, advancement towards a more general understanding of the natural world. Synthesis Ecology, with a few exceptions, has abandoned prediction and therefore the ability to demonstrate understanding. Here we address how this has inhibited progress in ecology and explore how a renewed focus on prediction would benefit ecologists. The lack of emphasis on prediction has resulted in a discipline that tests qualitative, imprecise hypotheses with little concern for whether the results are generalizable beyond where and when the data were collected. A renewed commitment to prediction would allow ecologists to address critical questions about the generalizability of our results and the progress we are making towards understanding the natural world.

One of the fundamental questions of ecology is what controls biodiversity. Recent theory suggests that biodiversity is controlled predominantly by neutral drift of species abundances. This theory has generated considerable controversy, because it claims that many mechanisms that have long been studied by ecologists (such as niches) have little involvement in structuring communities. The theory predicts that the species abundance distribution within a community should follow a zero-sum multinomial distribution (ZSM), but this has not, so far, been rigorously tested. Specifically, it remains to be shown that the ZSM fits the data significantly better than reasonable null models. Here I test whether the ZSM fits several empirical data sets better than the lognormal distribution. It does not. Not only does the ZSM fail to fit empirical data better than the lognormal distribution 95% of the time, it also fails to fit empirical data better even a majority of the time. This means that there is no evidence that the ZSM predicts abundances better than the much more parsimonious null hypothesis.