Luke A. Yates, Zach Aandahl, Shane A. Richards, Barry W. Brook
{"title":"Cross validation for model selection: A review with examples from ecology","authors":"Luke A. Yates, Zach Aandahl, Shane A. Richards, Barry W. Brook","doi":"10.1002/ecm.1557","DOIUrl":"10.1002/ecm.1557","url":null,"abstract":"<p>Specifying, assessing, and selecting among candidate statistical models is fundamental to ecological research. Commonly used approaches to model selection are based on predictive scores and include information criteria such as Akaike's information criterion, and cross validation. Based on data splitting, cross validation is particularly versatile because it can be used even when it is not possible to derive a likelihood (e.g., many forms of machine learning) or count parameters precisely (e.g., mixed-effects models). However, much of the literature on cross validation is technical and spread across statistical journals, making it difficult for ecological analysts to assess and choose among the wide range of options. Here we provide a comprehensive, accessible review that explains important—but often overlooked—technical aspects of cross validation for model selection, such as: bias correction, estimation uncertainty, choice of scores, and selection rules to mitigate overfitting. We synthesize the relevant statistical advances to make recommendations for the choice of cross-validation technique and we present two ecological case studies to illustrate their application. In most instances, we recommend using exact or approximate leave-one-out cross validation to minimize bias, or otherwise <i>k</i>-fold with bias correction if <i>k</i> < 10. To mitigate overfitting when using cross validation, we recommend calibrated selection via our recently introduced modified one-standard-error rule. We advocate for the use of predictive scores in model selection across a range of typical modeling goals, such as exploration, hypothesis testing, and prediction, provided that models are specified in accordance with the stated goal. We also emphasize, as others have done, that inference on parameter estimates is biased if preceded by model selection and instead requires a carefully specified single model or further technical adjustments.</p>","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":"93 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ecm.1557","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47711101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jeffrey A. Harvey, Kévin Tougeron, Rieta Gols, Robin Heinen, Mariana Abarca, Paul K. Abram, Yves Basset, Matty Berg, Carol Boggs, Jacques Brodeur, Pedro Cardoso, Jetske G. de Boer, Geert R. De Snoo, Charl Deacon, Jane E. Dell, Nicolas Desneux, Michael E. Dillon, Grant A. Duffy, Lee A. Dyer, Jacintha Ellers, Anahí Espíndola, James Fordyce, Matthew L. Forister, Caroline Fukushima, Matthew J. G. Gage, Carlos García-Robledo, Claire Gely, Mauro Gobbi, Caspar Hallmann, Thierry Hance, John Harte, Axel Hochkirch, Christian Hof, Ary A. Hoffmann, Joel G. Kingsolver, Greg P. A. Lamarre, William F. Laurance, Blas Lavandero, Simon R. Leather, Philipp Lehmann, Cécile Le Lann, Margarita M. López-Uribe, Chun-Sen Ma, Gang Ma, Joffrey Moiroux, Lucie Monticelli, Chris Nice, Paul J. Ode, Sylvain Pincebourde, William J. Ripple, Melissah Rowe, Michael J. Samways, Arnaud Sentis, Alisha A. Shah, Nigel Stork, John S. Terblanche, Madhav P. Thakur, Matthew B. Thomas, Jason M. Tylianakis, Joan Van Baaren, Martijn Van de Pol, Wim H. Van der Putten, Hans Van Dyck, Wilco C. E. P. Verberk, David L. Wagner, Wolfgang W. Weisser, William C. Wetzel, H. Arthur Woods, Kris A. G. Wyckhuys, Steven L. Chown
{"title":"Scientists' warning on climate change and insects","authors":"Jeffrey A. Harvey, Kévin Tougeron, Rieta Gols, Robin Heinen, Mariana Abarca, Paul K. Abram, Yves Basset, Matty Berg, Carol Boggs, Jacques Brodeur, Pedro Cardoso, Jetske G. de Boer, Geert R. De Snoo, Charl Deacon, Jane E. Dell, Nicolas Desneux, Michael E. Dillon, Grant A. Duffy, Lee A. Dyer, Jacintha Ellers, Anahí Espíndola, James Fordyce, Matthew L. Forister, Caroline Fukushima, Matthew J. G. Gage, Carlos García-Robledo, Claire Gely, Mauro Gobbi, Caspar Hallmann, Thierry Hance, John Harte, Axel Hochkirch, Christian Hof, Ary A. Hoffmann, Joel G. Kingsolver, Greg P. A. Lamarre, William F. Laurance, Blas Lavandero, Simon R. Leather, Philipp Lehmann, Cécile Le Lann, Margarita M. López-Uribe, Chun-Sen Ma, Gang Ma, Joffrey Moiroux, Lucie Monticelli, Chris Nice, Paul J. Ode, Sylvain Pincebourde, William J. Ripple, Melissah Rowe, Michael J. Samways, Arnaud Sentis, Alisha A. Shah, Nigel Stork, John S. Terblanche, Madhav P. Thakur, Matthew B. Thomas, Jason M. Tylianakis, Joan Van Baaren, Martijn Van de Pol, Wim H. Van der Putten, Hans Van Dyck, Wilco C. E. P. Verberk, David L. Wagner, Wolfgang W. Weisser, William C. Wetzel, H. Arthur Woods, Kris A. G. Wyckhuys, Steven L. Chown","doi":"10.1002/ecm.1553","DOIUrl":"10.1002/ecm.1553","url":null,"abstract":"<p>Climate warming is considered to be among the most serious of anthropogenic stresses to the environment, because it not only has direct effects on biodiversity, but it also exacerbates the harmful effects of other human-mediated threats. The associated consequences are potentially severe, particularly in terms of threats to species preservation, as well as in the preservation of an array of ecosystem services provided by biodiversity. Among the most affected groups of animals are insects—central components of many ecosystems—for which climate change has pervasive effects from individuals to communities. In this contribution to the scientists' warning series, we summarize the effect of the gradual global surface temperature increase on insects, in terms of physiology, behavior, phenology, distribution, and species interactions, as well as the effect of increased frequency and duration of extreme events such as hot and cold spells, fires, droughts, and floods on these parameters. We warn that, if no action is taken to better understand and reduce the action of climate change on insects, we will drastically reduce our ability to build a sustainable future based on healthy, functional ecosystems. We discuss perspectives on relevant ways to conserve insects in the face of climate change, and we offer several key recommendations on management approaches that can be adopted, on policies that should be pursued, and on the involvement of the general public in the protection effort.</p>","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":"93 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ecm.1553","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44323008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Bruce A. Menge, Jonathan W. Robinson, Brittany N. Poirson, Sarah A. Gravem
{"title":"Quantitative biogeography: Decreasing and more variable dynamics of critical species in an iconic meta-ecosystem","authors":"Bruce A. Menge, Jonathan W. Robinson, Brittany N. Poirson, Sarah A. Gravem","doi":"10.1002/ecm.1556","DOIUrl":"10.1002/ecm.1556","url":null,"abstract":"<p>Ecosystem stability has intrigued ecologists for decades, and the realization that the global climate was changing has sharpened and focused this interest. One possible early warning signal of decreasing stability is increasing variability in ecosystems over time with increasing climate variability. Determining climate change effects on community stability, however, requires long-term studies of structure and underlying dynamics, including bottom-up and top-down effects in natural ecosystems. Although relevant datasets were rare in the early years of community ecology, such information has increased in recent decades. We investigated spatiotemporal changes in mean and variability of ecological subsidies (nutrients, phytoplankton, prey colonization), performance metrics of a dominant space occupier (mussels) and its primary predator (sea stars), and sea star predation rates on mussels in relation to climatic oscillations, temperature, and disease on rocky shores. The research involved annually repeated multiyear (~1999–2018), multisite (13 sites nested within five regions along ~260 km of the Oregon coast) observations, measurements, and experiments. We analyzed associations between environmental variables and ecological performance of key elements of the sea star-mussel-dominated mid intertidal system. We found that upwelling declined in some regions, but became more variable across all study regions. Air and water temperatures oscillated, but their mean and variation increased through time, with peak values coinciding with the 2014–2016 combined El Niño and Marine Heat Wave. Ecological subsidies generally declined during the study period but increased in variability. Excepting growth rate, mussel (<i>Mytilus californianus</i>) performance (condition index, reproductive output) generally decreased and became more variable. Primarily due to a sea star wasting epidemic, reproductive output of the top predator <i>Pisaster ochraceus</i> decreased and became more variable, and predation rate on mussels decreased. Analyses indicated that the primary drivers of these changes were temperature-related environmental factors. As declining means and increasing variability of ecological performances can typify destabilizing ecosystems, and environmental trends are toward ever more stressful conditions, the outlook for this iconic ecosystem is discouraging. Immediate and rapid action to mitigate and ultimately reverse climate change likely is the only option available to prevent an irreversible shift in the future of this, and most other ecosystems.</p>","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":"93 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47491920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ingibjörg S. Jónsdóttir, Aud H. Halbritter, Casper T. Christiansen, Inge H. J. Althuizen, Siri V. Haugum, Jonathan J. Henn, Katrín Björnsdóttir, Brian Salvin Maitner, Yadvinder Malhi, Sean T. Michaletz, Ruben E. Roos, Kari Klanderud, Hanna Lee, Brian J. Enquist, Vigdis Vandvik
{"title":"Intraspecific trait variability is a key feature underlying high Arctic plant community resistance to climate warming","authors":"Ingibjörg S. Jónsdóttir, Aud H. Halbritter, Casper T. Christiansen, Inge H. J. Althuizen, Siri V. Haugum, Jonathan J. Henn, Katrín Björnsdóttir, Brian Salvin Maitner, Yadvinder Malhi, Sean T. Michaletz, Ruben E. Roos, Kari Klanderud, Hanna Lee, Brian J. Enquist, Vigdis Vandvik","doi":"10.1002/ecm.1555","DOIUrl":"10.1002/ecm.1555","url":null,"abstract":"<p>In the high Arctic, plant community species composition generally responds slowly to climate warming, whereas less is known about the community functional trait responses and consequences for ecosystem functioning. The slow species turnover and large distribution ranges of many Arctic plant species suggest a significant role of intraspecific trait variability in functional responses to climate change. Here we compare taxonomic and functional community compositional responses to a long-term (17-year) warming experiment in Svalbard, Norway, replicated across three major high Arctic habitats shaped by topography and contrasting snow regimes. We observed taxonomic compositional changes in all plant communities over time. Still, responses to experimental warming were minor and most pronounced in the drier habitats with relatively early snowmelt timing and long growing seasons (<i>Cassiope</i> and <i>Dryas</i> heaths). The habitats were clearly separated in functional trait space, defined by 12 size- and leaf economics-related traits, primarily due to interspecific trait variation. Functional traits also responded to experimental warming, most prominently in the <i>Dryas</i> heath and mostly due to intraspecific trait variation. Leaf area and mass increased and leaf δ<sup>15</sup>N decreased in response to the warming treatment. Intraspecific trait variability ranged between 30% and 71% of the total trait variation, reflecting the functional resilience of those communities, dominated by long-lived plants, due to either phenotypic plasticity or genotypic variation, which most likely underlies the observed resistance of high Arctic vegetation to climate warming. We further explored the consequences of trait variability for ecosystem functioning by measuring peak season CO<sub>2</sub> fluxes. Together, environmental, taxonomic, and functional trait variables explained a large proportion of the variation in net ecosystem exchange (NEE), which increased when intraspecific trait variation was accounted for. In contrast, even though ecosystem respiration and gross ecosystem production both increased in response to warming across habitats, they were mainly driven by the direct kinetic impacts of temperature on plant physiology and biochemical processes. Our study shows that long-term experimental warming has a modest but significant effect on plant community functional trait composition and suggests that intraspecific trait variability is a key feature underlying high Arctic ecosystem resistance to climate warming.</p>","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":"93 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/ecm.1555","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43000659","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Applying the structural causal model framework for observational causal inference in ecology","authors":"Suchinta Arif, M. Aaron MacNeil","doi":"10.1002/ecm.1554","DOIUrl":"https://doi.org/10.1002/ecm.1554","url":null,"abstract":"<p>Ecologists are often interested in answering causal questions from observational data but generally lack the training to appropriately infer causation. When applying statistical analysis (e.g., generalized linear model) on observational data, common statistical adjustments can often lead to biased estimates between variables of interest due to processes such as confounding, overcontrol, and collider bias. To overcome these limitations, we present an overview of structural causal modeling (SCM), an emerging causal inference framework that can be used to determine cause-and-effect relationships from observational data. The SCM framework uses directed acyclic graphs (DAGs) to visualize researchers' assumptions about the causal structure of a system or process under study. Following this, a DAG-based graphical rule known as the backdoor criterion can be applied to determine statistical adjustments (or lack thereof) required to determine causal relationships from observational data. In the presence of unobserved confounding variables, an additional rule called the frontdoor criterion can be employed to determine causal effects. Here, we use simulated ecological examples to review how the backdoor and frontdoor criteria can return accurate causal estimates between variables of interest, as well as how biases can arise when these criteria are not used. We further provide an overview of studies that have applied the SCM framework in ecology. SCM, along with its application of DAGs, has been widely used in other disciplines to make valid causal inferences from observational data. Their use in ecology holds tremendous potential for quantifying causal relationships and investigating a range of ecological questions without randomized experiments.</p>","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":"93 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50141322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Applying the structural causal model (\u0000 SCM\u0000 ) framework for observational causal inference in ecology","authors":"Suchinta Arif, M. MacNeil","doi":"10.1002/ecm.1554","DOIUrl":"https://doi.org/10.1002/ecm.1554","url":null,"abstract":"","PeriodicalId":11505,"journal":{"name":"Ecological Monographs","volume":" ","pages":""},"PeriodicalIF":6.1,"publicationDate":"2022-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49449955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}