Comparing traditional and Bayesian approaches to ecological meta‐analysis

How seasonality influences the thermal biology of lizards with different thermoregulatory strategies: a meta-analysis

Ectotherms that maintain thermal balance in the face of varying climates should be able to colonise a wide range of habitats. In lizards, thermoregulation usually appears as a variety of behaviours that buffer external influences over physiology. Basking species rely on solar radiation to raise body temperatures and usually show high thermoregulatory precision. By contrast, species that do not bask are often constrained by climatic conditions in their habitats, thus having lower thermoregulatory precision. While much focus has been given to the effects of mean habitat temperatures, relatively less is known about how seasonality affects the thermal biology of lizards on a macroecological scale. Considering the current climate crisis, assessing how lizards cope with temporal variations in environmental temperature is essential to understand better how these organisms will fare under climate change. Activity body temperatures (Tb ) represent the internal temperature of an animal measured in nature during its active period (i.e. realised thermal niche), and preferred body temperatures (Tpref ) are those selected by an animal in a laboratory thermal gradient that lacks thermoregulatory costs (i.e. fundamental thermal niche). Both traits form the bulk of thermal ecology research and are often studied in the context of seasonality. In this study, we used a meta-analysis to test how environmental temperature seasonality influences the seasonal variation in the Tb and Tpref of lizards that differ in thermoregulatory strategy (basking versus non-basking). Based on 333 effect sizes from 137 species, we found that Tb varied over a greater magnitude than Tpref across seasons. Variations in Tb were not influenced by environmental temperature seasonality; however, body size and thermoregulatory strategy mediated Tb responses. Specifically, larger species were subjected to greater seasonal variations in Tb , and basking species endured greater seasonal variations in Tb compared to non-basking species. On the other hand, the seasonal variation in Tpref increased with environmental temperature seasonality regardless of body size. Thermoregulatory strategy also influenced Tpref , suggesting that behaviour has an important role in mediating Tpref responses to seasonal variations in the thermal landscape. After controlling for phylogenetic effects, we showed that Tb and Tpref varied significantly across lizard families. Taken together, our results support the notion that the relationship between thermal biology responses and climatic parameters can be taxon and trait dependent. Our results also showcase the importance of considering ecological and behavioural aspects in macroecological studies. We further highlight current systematic, geographical, and knowledge gaps in thermal ecology research. Our work should benefit those who aim to understand more fully how seasonality shapes thermal biology in lizards, ultimately contributing to the goal of elucidating the evolution of temperature-sensitive traits in ectotherms.

'Small Data' for big insights in ecology

Big Data science has significantly furthered our understanding of complex systems by harnessing large volumes of data, generated at high velocity and in great variety. However, there is a risk that Big Data collection is prioritised to the detriment of 'Small Data' (data with few observations). This poses a particular risk to ecology where Small Data abounds. Machine learning experts are increasingly looking to Small Data to drive the next generation of innovation, leading to development in methods for Small Data such as transfer learning, knowledge graphs, and synthetic data. Meanwhile, meta-analysis and causal reasoning approaches are evolving to provide new insights from Small Data. These advances should add value to high-quality Small Data catalysing future insights for ecology.

Statistical inference for seed mortality and germination with seed bank experiments

Plant population ecologists regularly study soil seed banks with seed bag burial and seed addition experiments. These experiments contribute crucial data to demographic models, but we lack standard methods to analyze them. Here, we propose statistical models to estimate seed mortality and germination with observations from these experiments. We develop these models following the principles of event history analysis, and analyze their identifiability and statistical properties by algebraic methods and simulation. We demonstrate that seed bag burial, but not seed addition experiments, can be used to make inferences about age-dependent mortality and germination. When mortality and germination do not change with seed age, both experiments produce unbiased estimates but seed bag burial experiments are more precise. However, seed mortality and germination estimates may be inaccurate when the statistical model that is fit makes incorrect assumptions about the age dependence of mortality and germination. The statistical models and simulations that we present here can be adopted and modified by plant population ecologists to strengthen inferences about seed mortality and germination in the soil seed bank.

Phylogeny explains capture mortality of sharks and rays in pelagic longline fisheries: a global meta-analytic synthesis

Apex and mesopredators such as elasmobranchs are important for maintaining ocean health and are the focus of conservation efforts to mitigate exposure to fishing and other anthropogenic hazards. Quantifying fishing mortality components such as at-vessel mortality (AVM) is necessary for effective bycatch management. We assembled a database for 61 elasmobranch species and conducted a global meta-synthesis to estimate pelagic longline AVM rates. Evolutionary history was a significant predictor of AVM, accounting for up to 13% of variance in Bayesian phylogenetic meta-regression models for Lamniformes and Carcharhiniformes clades. Phylogenetically related species may have a high degree of shared traits that explain AVM. Model-estimated posterior mean AVM rates ranged from 5% (95% HDI 0.1%-16%) for pelagic stingrays and 76% (95% HDI 49%-90%) for salmon sharks. Measures that reduce catch, and hence AVM levels, such as input controls, bycatch quotas and gear technology to increase selectivity are appropriate for species with higher AVM rates. In addition to reducing catchability, handling-and-release practices and interventions such as retention bans in shark sanctuaries and bans on shark finning and trade hold promise for species with lower AVM rates. Robust, and where applicable, phylogenetically-adjusted elasmobranch AVM rates are essential for evidence-informed bycatch policy.

Seed ecology of European mesic meadows

BACKGROUND AND AIMS

European mesic meadows are semi-natural open habitats of high biodiversity and an essential part of European landscapes. These species-rich communities can be a source of seed mixes for ecological restoration, urban greening and rewilding. However, limited knowledge of species germination traits is a bottleneck to the development of a competitive native seed industry. Here, we synthesize the seed ecology of mesic meadows.

METHODS

We combined our own experimental data with data obtained from databases to create a combined dataset containing 2005 germination records of 90 plant species from 31 European countries. We performed a Bayesian meta-analysis of this dataset to test the seed germination response to environmental cues including scarification, stratification, temperature, alternating temperature and light. We also used multivariate ordination to check the relationship between seed traits (germination and morphology) and species ecological preferences, and to compare the seed ecology of mesic meadows with that of other herbaceous plant communities from the same geographic area.

KEY RESULTS

The seed ecology of mesic meadows is characterized by (1) high seed germinability when compared with other herbaceous plant communities; (2) low correspondence between seed traits and species ecological preferences; and (3) a deep phylogenetic separation between the two major families, Poaceae and Fabaceae. Poaceae produce many light seeds that respond to gap-detecting germination cues (alternating temperatures and light); Fabaceae produce fewer heavy seeds, which need scarification to break their physical dormancy.

CONCLUSIONS

High germinability of meadow seeds will reduce their capacity to form persistent seed banks, resulting in dispersal limitations to passive regeneration. For centuries, human activities have shaped the regeneration of meadows, leading to a loss of seed dormancy and decoupling seeds from seasonal cycles, as has been found in many domesticated species. The same anthropic processes that have shaped semi-natural mesic meadows have left them dependent on continued human intervention for their regeneration, highlighting the importance of active restoration via seed supply.

A hierarchical, multivariate meta-analysis approach to synthesising global change experiments

Meta-analyses enable synthesis of results from globally distributed experiments to draw general conclusions about the impacts of global change factors on ecosystem function. Traditional meta-analyses, however, are challenged by the complexity and diversity of experimental results. We illustrate how several key issues can be addressed by a multivariate, hierarchical Bayesian meta-analysis (MHBM) approach applied to information extracted from published studies. We applied an MHBM to log-response ratios for aboveground biomass (AB, n = 300), belowground biomass (BB, n = 205) and soil CO₂ exchange (SCE, n = 544), representing 100 studies. The MHBM accounted for study duration, climate effects and covariation among the AB, BB and SCE responses to elevated CO₂ (eCO₂ ) and/or warming. The MHBM revealed significant among-study covariation in the AB and BB responses to experimental treatments. The MHBM imputed missing duration (4.2%) and climate (6%) data, and revealed that climate context governs how eCO₂ and warming impact ecosystem function. Predictions identified biomes that may be particularly sensitive to eCO₂ or warming, but that are under-represented in global change experiments. The MHBM approach offers a flexible and powerful tool for synthesising disparate experimental results reported across multiple studies, sites and response variables.

An assessment of statistical methods for non-independent data in ecological meta-analyses: Comment

Recently, Song et al. (2020) conducted a simulation study using different methods to deal with non-independence resulting from effect sizes originating from the same paper - a common occurrence in ecological meta-analyses. The main methods that were of interest in their simulations were: 1) a standard random-effects model used in combination with a weighted average effect size for each paper (i.e., a two-step method), 2) a standard random-effects model after randomly choosing one effect size per paper, 3) a multilevel (hierarchical) meta-analysis model, modelling paper identity as a random factor, and 4) a meta-analysis making use of a robust variance estimation method. Based on their simulation results, they recommend that meta-analysts should either use the two-step method, which involves taking a weighted paper mean followed by analysis with a random-effects model, or the robust variance estimation method.

Bayesian Methods for Meta-Analyses of Binary Outcomes: Implementations, Examples, and Impact of Priors

Bayesian methods are an important set of tools for performing meta-analyses. They avoid some potentially unrealistic assumptions that are required by conventional frequentist methods. More importantly, meta-analysts can incorporate prior information from many sources, including experts’ opinions and prior meta-analyses. Nevertheless, Bayesian methods are used less frequently than conventional frequentist methods, primarily because of the need for nontrivial statistical coding, while frequentist approaches can be implemented via many user-friendly software packages. This article aims at providing a practical review of implementations for Bayesian meta-analyses with various prior distributions. We present Bayesian methods for meta-analyses with the focus on odds ratio for binary outcomes. We summarize various commonly used prior distribution choices for the between-studies heterogeneity variance, a critical parameter in meta-analyses. They include the inverse-gamma, uniform, and half-normal distributions, as well as evidence-based informative log-normal priors. Five real-world examples are presented to illustrate their performance. We provide all of the statistical code for future use by practitioners. Under certain circumstances, Bayesian methods can produce markedly different results from those by frequentist methods, including a change in decision on statistical significance. When data information is limited, the choice of priors may have a large impact on meta-analytic results, in which case sensitivity analyses are recommended. Moreover, the algorithm for implementing Bayesian analyses may not converge for extremely sparse data; caution is needed in interpreting respective results. As such, convergence should be routinely examined. When select statistical assumptions that are made by conventional frequentist methods are violated, Bayesian methods provide a reliable alternative to perform a meta-analysis.