Predictive models aren't for causal inference

Amazon forest biogeography predicts resilience and vulnerability to drought

Amazonia contains the most extensive tropical forests on Earth, but Amazon carbon sinks of atmospheric CO2 are declining, as deforestation and climate-change-associated droughts1-4 threaten to push these forests past a tipping point towards collapse5-8. Forests exhibit complex drought responses, indicating both resilience (photosynthetic greening) and vulnerability (browning and tree mortality), that are difficult to explain by climate variation alone9-17. Here we combine remotely sensed photosynthetic indices with ground-measured tree demography to identify mechanisms underlying drought resilience/vulnerability in different intact forest ecotopes18,19 (defined by water-table depth, soil fertility and texture, and vegetation characteristics). In higher-fertility southern Amazonia, drought response was structured by water-table depth, with resilient greening in shallow-water-table forests (where greater water availability heightened response to excess sunlight), contrasting with vulnerability (browning and excess tree mortality) over deeper water tables. Notably, the resilience of shallow-water-table forest weakened as drought lengthened. By contrast, lower-fertility northern Amazonia, with slower-growing but hardier trees (or, alternatively, tall forests, with deep-rooted water access), supported more-drought-resilient forests independent of water-table depth. This functional biogeography of drought response provides a framework for conservation decisions and improved predictions of heterogeneous forest responses to future climate changes, warning that Amazonia's most productive forests are also at greatest risk, and that longer/more frequent droughts are undermining multiple ecohydrological strategies and capacities for Amazon forest resilience.

General reversal of N-decomposition relationship during long-term decomposition in boreal and temperate forests

Decomposition of dead organic matter is fundamental to carbon (C) and nutrient cycling in terrestrial ecosystems, influencing C fluxes from the biosphere to the atmosphere. Theory predicts and evidence strongly supports that the availability of nitrogen (N) limits litter decomposition. Positive relationships between substrate N concentrations and decomposition have been embedded into ecosystem models. This decomposition paradigm, however, relies on data mostly from short-term studies analyzing controls on early-stage decomposition. We present evidence from three independent long-term decomposition investigations demonstrating that the positive N-decomposition relationship is reversed and becomes negative during later stages of decomposition. First, in a 10-y decomposition experiment across 62 woody species in a temperate forest, leaf litter with higher N concentrations exhibited faster initial decomposition rates but ended up a larger recalcitrant fraction decomposing at a near-zero rate. Second, in a 5-y N-enrichment experiment of two tree species, leaves with experimentally enriched N concentrations had faster decomposition initial rates but ultimately accumulated large slowly decomposing fractions. Measures of amino sugars on harvested litter in two experiments indicated that greater accumulation of microbial residues in N-rich substrates likely contributed to larger slowly decomposing fractions. Finally, a database of 437 measurements from 120 species in 45 boreal and temperate forest sites confirmed that higher N concentrations were associated with a larger slowly decomposing fraction. These results challenge the current treatment of interactions between N and decomposition in many ecosystems and Earth system models and suggest that even the best-supported short-term controls of biogeochemical processes might not predict long-term controls.

Legislative incapacity and underreporting of COVID-19 mortality

The prevalent interpretation of COVID-19 mortality underreporting typically focuses on authoritarian regimes' propensity for data manipulation. This study, however, posits that the demand side is integral to enhancing the veracity of COVID-19 mortality figures. Through quantitative analysis, it is demonstrated that legislative oversight of the executive significantly correlates with the divergence between excess mortality and officially reported COVID-19 mortality. Moreover, such oversight is shown to bolster the influence of bureaucratic capacity on the precision of mortality data. Consequently, these findings suggest that the notion of "autocratic advantage" in COVID-19 management is not solely a byproduct of regime-led data falsification but also a reflection of deficiencies in legislative and bureaucratic capacities.

Body size shapes song in honeyeaters

Birdsongs are among the most distinctive animal signals. Their evolution is thought to be shaped simultaneously by habitat structure and by the constraints of morphology. Habitat structure affects song transmission and detectability, thus influencing song (the acoustic adaptation hypothesis), while body size and beak size and shape necessarily constrain song characteristics (the morphological constraint hypothesis). Yet, support for the acoustic adaptation and morphological constraint hypotheses remains equivocal, and their simultaneous examination is infrequent. Using a phenotypically diverse Australasian bird clade, the honeyeaters (Aves: Meliphagidae), we compile a dataset consisting of song, environmental, and morphological variables for 163 species and jointly examine predictions of these two hypotheses. Overall, we find that body size constrains song frequency and pace in honeyeaters. Although habitat type and environmental temperature influence aspects of song, that influence is indirect, likely via effects of environmental variation on body size, with some evidence that elevation constrains the evolution of song peak frequency. Our results demonstrate that morphology has an overwhelming influence on birdsong, in support of the morphological constraint hypothesis, with the environment playing a secondary role generally via body size rather than habitat structure. These results suggest that changing body size (a consequence of both global effects such as climate change and local effects such as habitat transformation) will substantially influence the nature of birdsong.

The potential impact fraction of population weight reduction scenarios on non-communicable diseases in Belgium: application of the g-computation approach

BACKGROUND

Overweight is a major risk factor for non-communicable diseases (NCDs) in Europe, affecting almost 60% of all adults. Tackling obesity is therefore a key long-term health challenge and is vital to reduce premature mortality from NCDs. Methodological challenges remain however, to provide actionable evidence on the potential health benefits of population weight reduction interventions. This study aims to use a g-computation approach to assess the impact of hypothetical weight reduction scenarios on NCDs in Belgium in a multi-exposure context.

METHODS

Belgian health interview survey data (2008/2013/2018, n = 27 536) were linked to environmental data at the residential address. A g-computation approach was used to evaluate the potential impact fraction (PIF) of population weight reduction scenarios on four NCDs: diabetes, hypertension, cardiovascular disease (CVD), and musculoskeletal (MSK) disease. Four scenarios were considered: 1) a distribution shift where, for each individual with overweight, a counterfactual weight was drawn from the distribution of individuals with a "normal" BMI 2) a one-unit reduction of the BMI of individuals with overweight, 3) a modification of the BMI of individuals with overweight based on a weight loss of 10%, 4) a reduction of the waist circumference (WC) to half of the height among all people with a WC:height ratio greater than 0.5. Regression models were adjusted for socio-demographic, lifestyle, and environmental factors.

RESULTS

The first scenario resulted in preventing a proportion of cases ranging from 32.3% for diabetes to 6% for MSK diseases. The second scenario prevented a proportion of cases ranging from 4.5% for diabetes to 0.8% for MSK diseases. The third scenario prevented a proportion of cases, ranging from 13.6% for diabetes to 2.4% for MSK diseases and the fourth scenario prevented a proportion of cases ranging from 36.4% for diabetes to 7.1% for MSK diseases.

CONCLUSION

Implementing weight reduction scenarios among individuals with excess weight could lead to a substantial and statistically significant decrease in the prevalence of diabetes, hypertension, cardiovascular disease (CVD), and musculoskeletal (MSK) diseases in Belgium. The g-computation approach to assess PIF of interventions represents a straightforward approach for drawing causal inferences from observational data while providing useful information for policy makers.

Drivers of arthropod biodiversity in an urban ecosystem

Our world is becoming increasingly urbanized with a growing human population concentrated around cities. The expansion of urban areas has important consequences for biodiversity, yet the abiotic drivers of biodiversity in urban ecosystems have not been well characterized for the most diverse group of animals on the planet, arthropods. Given their great diversity, comparatively small home ranges, and ability to disperse, arthropods make an excellent model for studying which factors can most accurately predict urban biodiversity. We assessed the effects of (i) topography (distance to natural areas and to ocean) (ii) abiotic factors (mean annual temperature and diurnal range), and (iii) anthropogenic drivers (land value and amount of impervious surface) on the occurrence of six arthropod groups represented in Malaise trap collections run by the BioSCAN project across the Greater Los Angeles Area. We found striking heterogeneity in responses to all factors both within and between taxonomic groups. Diurnal temperature range had a consistently negative effect on occupancy but this effect was only significant in Phoridae. Anthropogenic drivers had mixed though mostly insignificant effects, as some groups and species were most diverse in highly urbanized areas, while other groups showed suppressed diversity. Only Phoridae was significantly affected by land value, where most species were more likely to occur in areas with lower land value. Los Angeles can support high regional arthropod diversity, but spatial community composition is highly dependent on the taxonomic group.

Long-term monitoring of cycles in Clethrionomys rutilus in the Yukon boreal forest

Baseline studies of small rodent populations in undisturbed ecosystems are rare. We report here 50 years of monitoring and experimentation in Yukon of a dominant rodent species in the North American boreal forest, the red-backed vole Clethrionomys rutilus. These voles breed in summer, weigh 20-25 g, and reach a maximum density of 20 to 25 per ha. Their populations have shown consistent 3-4-year cycles for the last 50 years with the only change being that peak densities averaged 8/ha until 2000 and 18/ha since that year. During the last 25 years, we have measured food resources, predator numbers, and winter weather, and for 1-year social interactions, to estimate their contribution to changes in the rate of summer increase and the rate of overwinter decline. All these potential limiting factors could contribute to changes in density, and we measured their relative contributions statistically with multiple regressions. The rate of winter decline in density was related to both food supply and winter severity. The rate of summer increase was related to summer berry crops and white spruce cone production. No measure of predator numbers was related to winter or summer changes in vole abundance. There was a large signal of climate change effects in these populations. There is no density dependence in summer population growth and only a weak one in winter population declines. None of our results provide a clear understanding of what generates 3-4-year cycles in these voles, and the major missing piece may be an understanding of social interactions at high density.

A spatially structured mathematical model of the gut microbiome reveals factors that increase community stability

Given the importance of gut microbial communities for human health, we may want to ensure their stability in terms of species composition and function. Here, we built a mathematical model of a simplified gut composed of two connected patches where species and metabolites can flow from an upstream patch, allowing upstream species to affect downstream species' growth. First, we found that communities in our model are more stable if they assemble through species invasion over time compared to combining a set of species from the start. Second, downstream communities are more stable when species invade the downstream patch less frequently than the upstream patch. Finally, upstream species that have positive effects on downstream species can further increase downstream community stability. Despite it being quite abstract, our model may inform future research on designing more stable microbial communities or increasing the stability of existing ones.

Landscape-scale benefits of protected areas for tropical biodiversity

The United Nations recently agreed to major expansions of global protected areas (PAs) to slow biodiversity declines1. However, although reserves often reduce habitat loss, their efficacy at preserving animal diversity and their influence on biodiversity in surrounding unprotected areas remain unclear2-5. Unregulated hunting can empty PAs of large animals6, illegal tree felling can degrade habitat quality7, and parks can simply displace disturbances such as logging and hunting to unprotected areas of the landscape8 (a phenomenon called leakage). Alternatively, well-functioning PAs could enhance animal diversity within reserves as well as in nearby unprotected sites9 (an effect called spillover). Here we test whether PAs across mega-diverse Southeast Asia contribute to vertebrate conservation inside and outside their boundaries. Reserves increased all facets of bird diversity. Large reserves were also associated with substantially enhanced mammal diversity in the adjacent unprotected landscape. Rather than PAs generating leakage that deteriorated ecological conditions elsewhere, our results are consistent with PAs inducing spillover that benefits biodiversity in surrounding areas. These findings support the United Nations goal of achieving 30% PA coverage by 2030 by demonstrating that PAs are associated with higher vertebrate diversity both inside their boundaries and in the broader landscape.

Enhanced dominance of soil moisture stress on vegetation growth in Eurasian drylands

Despite the mounting attention being paid to vegetation growth and their driving forces for water-limited ecosystems, the relative contributions of atmospheric and soil moisture dryness stress on vegetation growth are an ongoing debate. Here we comprehensively compare the impacts of high vapor pressure deficit (VPD) and low soil water content (SWC) on vegetation growth in Eurasian drylands during 1982-2014. The analysis indicates a gradual decoupling between atmospheric dryness and soil dryness over this period, as the former has expanded faster than the latter. Moreover, the VPD-SWC relation and VPD-greenness relation are both non-linear, while the SWC-greenness relation is near-linear. The loosened coupling between VPD and SWC, the non-linear correlations among VPD-SWC-greenness and the expanded area extent in which SWC acts as the dominant stress factor all provide compelling evidence that SWC is a more influential stressor than VPD on vegetation growth in Eurasian drylands. In addition, a set of 11 Earth system models projected a continuously growing constraint of SWC stress on vegetation growth towards 2100. Our results are vital to dryland ecosystems management and drought mitigation in Eurasia.

Biodiversity change under adaptive community dynamics

Compositional change is a ubiquitous response of ecological communities to environmental drivers of global change, but is often regarded as evidence of declining "biotic integrity" relative to historical baselines. Adaptive compositional change, however, is a foundational idea in evolutionary biology, whereby changes in gene frequencies within species boost population-level fitness, allowing populations to persist as the environment changes. Here, we present an analogous idea for ecological communities based on core concepts of fitness and selection. Changes in community composition (i.e., frequencies of genetic differences among species) in response to environmental change should normally increase the average fitnessof community members. We refer to compositional changes that improve the functional match, or "fit," between organisms' traits and their environment as adaptive community dynamics. Environmental change (e.g., land-use change) commonly reduces the fit between antecedent communities and new environments. Subsequent change in community composition in response to environmental changes, however, should normally increase community-level fit, as the success of at least some constituent species increases. We argue that adaptive community dynamics are likely to improve or maintain ecosystem function (e.g., by maintaining productivity). Adaptive community responses may simultaneously produce some changes that are considered societally desirable (e.g., increased carbon storage) and others that are undesirable (e.g., declines of certain species), just as evolutionary responses within species may be deemed desirable (e.g., evolutionary rescue of an endangered species) or undesirable (e.g., enhanced virulence of an agricultural pest). When assessing possible management interventions, it is important to distinguish between drivers of environmental change (e.g., undesired climate warming) and adaptive community responses, which may generate some desirable outcomes. Efforts to facilitate, accept, or resist ecological change require separate consideration of drivers and responses, and may highlight the need to reconsider preferences for historical baseline communities over communities that are better adapted to the new conditions.

Analysis of causal effects of 137Cs deposition on 137Cs concentrations in trees after the Fukushima accident using machine learning

Radioactive contamination of forests by long-lived radionuclides from nuclear accidents such as Chernobyl and Fukushima continues to be studied and quantitatively modeled. Whereas traditional statistical and machine learning (ML) techniques generate predictions by focusing on correlations between variables, quantification of causal effects of radioactivity deposition levels on contamination of plant tissues represents a more fundamental and relevant research goal. Modeling of cause-and-effect relationships is advantageous over standard predictive modeling, particularly by improving the generalizability of results to other situations, where the distributions of variables, including potential confounders, differ from those in the training data. Here we used the state-of-the-art causal forest (CF) algorithm to quantify the causal effect of ¹³⁷Cs land contamination after the Fukushima accident on ¹³⁷Cs activity concentrations in the wood of four common Japanese forest tree species: Hinoki cypress (Chamaecyparis obtusa), konara oak (Quercus serrata), red pine (Pinus densiflora), and Sugi cedar (Cryptomeria japonica). We estimated the average causal effect for the population, quantified how it was influenced by other environmental variables, and produced effect estimates at the individual level. The estimated causal effect was quite robust to various refutation methods, and was negatively influenced by high mean annual precipitation, elevation, and time after the accident. Wood subtype (e.g. sapwood, heartwood) and tree species made smaller contributions to the causal effect. We believe that causal ML techniques have promising potential in radiation ecology and can usefully expand the toolkit of modeling approaches available to researchers in this field.

Reducing bias in experimental ecology through directed acyclic graphs

Ecologists often rely on randomized control trials (RCTs) to quantify causal relationships in nature. Many of our foundational insights of ecological phenomena can be traced back to well‐designed experiments, and RCTs continue to provide valuable insights today. Although RCTs are often regarded as the “gold standard” for causal inference, it is important to recognize that they too rely on a set of causal assumptions that must be justified and met by the researcher to draw valid causal conclusions. We use key ecological examples to show how biases such as confounding, overcontrol, and collider bias can occur in experimental setups. In tandem, we highlight how such biases can be removed through the application of the structural causal model (SCM) framework. The SCM framework visualizes the causal structure of a system or process under study using directed acyclic graphs (DAGs) and subsequently applies a set of graphical rules to remove bias from both observational and experimental data. We show how DAGs can be applied across ecological experimental studies to ensure proper study design and statistical analysis, leading to more accurate causal estimates drawn from experimental data. Although causal conclusions drawn from RCTs are often taken at face value, ecologists are increasingly becoming aware that experimental approaches must be carefully designed and analyzed to avoid potential biases. By applying DAGs as a visual and conceptual tool, experimental ecologists can increasingly meet the causal assumptions required for valid causal inference.

Model selection in occupancy models: Inference versus prediction

Occupancy models are a vital tool for ecologists studying the patterns and drivers of species occurrence, but their use often involves selecting among models with different sets of occupancy and detection covariates. The information-theoretic approach, which employs information criteria such as Akaike's information criterion (AIC) is arguably the most popular approach for model selection in ecology and is often used for selecting occupancy models. However, the information-theoretic approach risks selecting models that produce inaccurate parameter estimates due to a phenomenon called collider bias, a type of confounding that can arise when adding explanatory variables to a model. Using simulations, we investigated the consequences of collider bias (using an illustrative example called M-bias) in the occupancy and detection processes of an occupancy model, and explored the implications for model selection using AIC and a common alternative, the Schwarz criterion (or Bayesian information criterion, BIC). We found that when M-bias was present in the occupancy process, AIC and BIC selected models that inaccurately estimated the effect of the focal occupancy covariate, while simultaneously producing more accurate predictions of the site-level occupancy probability than other models in the candidate set. In contrast, M-bias in the detection process did not impact the focal estimate; all models made accurate inferences, while the site-level predictions of the AIC/BIC-best model were slightly more accurate. Our results show that information criteria can be used to select occupancy covariates if the sole purpose of the model is prediction, but must be treated with more caution if the purpose is to understand how environmental variables affect occupancy. By contrast, detection covariates can usually be selected using information criteria regardless of the model's purpose. These findings illustrate the importance of distinguishing between the tasks of parameter inference and prediction in ecological modeling. Furthermore, our results underline concerns about the use of information criteria to compare different biological hypotheses in observational studies.

Bedrock may dictate the distribution of the fire salamander in the southern border of its global range

Understanding the factors that determine the spatial distribution of species is crucial for conservation planning. In this short communication, we review previous distribution models of the fire salamander (Salamandra infraimmaculata) in northern Israel, produced by the group of the late Prof. Leon Blaustein, while suggesting a biologically-informed reinterpretation of their main predictions. We argue for the prime importance of bedrock, specifically hard limestone, because it is tightly associated with the availability of karstic formations that are key to adult survival throughout the summer. Furthermore, we suggest that the spatial distribution of limestone bedrock also determines large-scale inter-population connectivity, and may explain the observed genetic differentiation among populations, as well as the southernmost limit of the species’ global distribution.