Paths to statistical fluency for ecologists

Abstract

Twenty-first century ecology requires statistical fluency. Observational and experimental studies routinely gather non-Normal, multivariate data at many spatiotemporal scales.Experimental studies routinely include multiple blocked and nested factors. Ecological theories routinely incorporate both deterministic and stochastic processes. Ecological debates frequently revolve around choices of statistical analyses. Our journals are replete with likelihood and statespace models, Bayesian and frequentist inference, complex multivariate analyses, and papers on statistical theory and methods. We test hypotheses, model data, and forecast future environmental conditions. And many appropriate statistical methods are not automated in software packages. It is time for ecologists to understand statistical modeling well enough to construct nonstandard statistical models and apply various types of inference -estimation, hypothesis testing, model selection, and prediction -to our models and scientific questions. In short, ecologists need to move beyond basic statistical literacy and attain statistical fluency. 2 In a nutshell: 23! Ecologists need to use nonstandard statistical models and methods of statistical inference to 24 test models of ecological processes and to address pressing environmental problems. 25 ! Such statistical models of ecological processes include both deterministic and stochastic 26 parts, and statistically-fluent ecologists will need to use probability theory and calculus to fit 27 these models to available data. 28 ! Many ecologists lack appropriate background in probability theory and calculus because 29 there are serious disconnections between the quantitative nature of ecology, the quantitative 30 skills we expect of ourselves and our students, and how we teach and learn quantitative 31 methods. 32 !

Replication, lies and lesser‐known truths regarding experimental design in environmental microbiology

“…A traditional course in biostatistics should emphasize the importance of replication. A progressive course in biostatistics, however, should convey that there are some situations where replication is either not feasible or not necessary, depending on the question and system (see Ellison and Dennis, 2010). Ultimately, environmental microbiologist must appreciate that replication determines a study's inference space, but that it also provides a relatively straightforward way to accept or reject hypotheses, which is an essential tool for testing theory (Prosser et al ., 2007).…”

Better Science Needed for Restoration in the Gulf of Mexico

“…Incentives to increase data sharing, such as making it a requirement of funding, permits, or publication, should be developed ( 22). Improve assessment tools for evaluation of anthropogenic impacts on populations by fostering interdisciplinary research among fi sheries science, marine ecology, and conservation biology and by funding opportunities for student training and continuing education for managers in the quantitative sciences ( 23). The Bering Sea Project is an excellent example of such a program ( 24).…”

Towards an integration of scale and complexity in marine ecology

“…These Hereu et al 2008). They do require some more statistical savvy on the part of ecologists (Ellison and Dennis 2009), but the potential gains are enormous. This was done by stretching a net across a cove, excluding fish from one side (right in diagram) of the experiment, and creating an urchin-only area inside the fence.…”