Avoiding the scale sampling problem: A consilient solution

Abstract

What if there was a way to solve one of the most complex and important scale challenges that ecologists face: the selection of the appropriate sampling scale resolution and extent when conducting studies. To date, the investigator conducting any study has been tasked with this challenge. Qualitatively different patterns arise as the scale resolution or extent of the observation is changed. Likewise, different patterns arise as the scales of the sampling are changed. Too often, data at one resolution are aggregated or interpretations of patterns at smaller scale resolutions and extents are extrapolated to larger extents. When this is done, either modifiable areal unit problems or extrapolation errors arise. I argue that consilient developments in movement ecology, graph (network) theory and analyses, and important technical advances in biologging can be linked synergistically to directly address the sampling scale problem, particularly in studies where understanding animal response to changing landscapes is the goal. The essence of the solution is that rather than having to decide the spatial extent and resolution to use, the issue is avoided by allowing the telemetered animal to show the relevant scales it uses as it moves during its daily, seasonal, and annual activities. The consilient result is that big data at the appropriate scales can be extracted rather than aggregated as is now the rule. Without data aggregation, no modifiable areal unit problems are involved because the dynamics are network based. Additionally, use of first passage time methodology and measurement of area‐restricted movements allow assessments of the scales of activity and the effects of environmental covariates on animal movement. I begin with a brief history of the concept of scale, describing what is considered to be the essence of the problem. I follow with a brief explanation of the developments in movement ecology, graph theory and analysis, and the technical advancements in biologging with examples from studies of wildlife species that are using these new developments. I then describe how their consilient integration can provide a different way of thinking about scale resolution and extent in ecology and suggest how ecologists can avoid the observer sampling scale complication by making a critical scale distinction and by using first passage time analyses. © 2016 The Wildlife Society.

Defining Landscapes and Scales to Model Landscape–Organism Interactions

“…Finer-resolution data across spatiotemporal scales are now widely available thanks to ever-increasing technological advancements [33], but these data often are correlated in space and time, violating statistical inference assumptions [29]. Landscapes by their very nature are autocorrelated.…”

“…Measurement error is present in any remotely collected data, including data commonly used to evaluate species-habitat relationships such as GPS radiotelemetry data and habitat variables commonly derived in a geographic information system (GIS; e.g., landcover type, digital-elevations models, normalized difference vegetation index, etc.). Measurement error can affect both precision and accuracy [29]. Researchers often assume that poor precision should not inherently lead to bias; however, past research illustrates that this assumption is not always true [44].…”

“…This can trigger the modifiable areal unit problem (MAUP), which remains an unresolved problem in landscape ecology. In short, MAUP occurs when the interpretation of results varies according to scale [28,29]. For example, a response variable can show a negative relationship with a predictor variable at a fine scale and a positive relationship at coarser scales [30].…”

Hierarchical population monitoring of greater sage-grouse (<em>Centrocercus urophasianus</em>) in Nevada and California—Identifying populations for management at the appropriate spatial scale

“…Such an alignment is a fundamental requirement for understanding population dynamics (Levin, 1992). The contribution of environmental and anthropogenic factors responsible for regulating (density dependent) or limiting (density independent) population growth often differ when measured at varying extents; understanding these differences provides key information on the mechanisms underlying changes in population abundance (Bissonette, 1997(Bissonette, , 2016Fuhlendorf and others, 2002). Thus, spatial scales at which local populations operate must be defined and investigated if population growth and declines are to be truly understood.…”

“…When applying management actions to conserve wildlife populations, failure to account for scale-specific processes (spatial and temporal) can result in significant misinterpretation of observed patterns (Sadoul, 1997;Bissonette, 2016). If the environmental or anthropogenic threat responsible for population decline is misaligned with the subsequent implementation of management actions, such actions are likely to be unsuccessful (Epifanio, 2000;Cummings and other, 2006).…”