Cost and Precision Functions for Aerial Quadrat Surveys: a Case Study of Ring‐Necked Ducks in Minnesota

Abstract

Cost considerations may be as important as precision when making survey‐design choices, and the ability to accurately estimate survey costs will be essential if survey budgets become more constrained. We used data from a survey of ring‐necked ducks (Aythya collaris) to illustrate how simple distance formulas can be used to construct a cost function for aerial quadrat surveys. Our cost function provided reasonable estimates of effort (hr) and costs, and allowed us to evaluate plot‐size choices in terms of expected cost‐precision tradeoffs. Although factors influencing costs in wildlife surveys can be complicated, we believe that cost functions deserve more attention and should be routinely considered in conjunction with traditional power analyses.

Spending degrees of freedom in a poor economy: A case study of building a sightability model for moose in northeastern Minnesota

“…These concerns suggest that model‐assisted survey estimators might be worth exploring (e.g., a Bayesian approach that allows one to place a prior on ${\hat {\pi }}$ 's, effectively shrinking them away from zero). Lastly, survey costs are also an important consideration (Noyes et al 2000, Rabe et al 2002, Giudice et al 2010) and may increase if a large suite of predictor variables must be recorded at each sighting. Thus, we concluded that a sightability model based solely on VO had many advantages.…”

An aerial grid‐based sampling method to estimate nonbreeding duck abundance

“…We defined a precision goal of coefficients of variation of ≤25% to evaluate effectiveness because estimates meeting this level of precision would be most useful to detect abundance changes found in the grid‐based survey conducted by Eggeman et al (). We chose to use a grid‐based approach, rather than the traditional transect‐based method (Reinecke et al , Pearse et al ), because 1) transect‐based surveys were logistically infeasible because of the relatively narrow floodplain width in portions of the region that contained the majority of waterfowl habitat; 2) waterfowl distribution was predicted to be nonrandom across the entire region, but was possibly random within more localized areas; 3) grid‐based approaches have shown promise in other regions (e.g., Giudice et al ); and 4) available, low‐winged aircraft used for concurrent inventory surveys were not well‐suited for transect surveys and we sought to maintain as much consistency as possible between survey types. We superimposed a grid consisting of 769 individual 260‐ha cells (2.6 km 2 or 1 mile 2 ) onto the study area (Fig.…”

“…Attempts to employ probability‐based sampling techniques during the nonbreeding season for use at smaller scales or in riverine areas, which often contain discontinuous and irregular habitat, have rarely been implemented (but see Fleskes and Yee , Shirkey et al ). Grid‐based designs incorporating helicopters have been shown to work well in such habitats during the breeding season (Cordts et al , Zicus et al , Giudice et al ); yet, aside from Eggeman et al (), a grid‐based approach has not been conducted during the nonbreeding season with a fixed‐wing aircraft.…”

A spatially explicit, multi‐scale occupancy model for large‐scale population monitoring

“…When combined with the addition of environmental variables, our spatial modeling framework offers a powerful tool for ecologists and resource managers to monitor, explain, and even predict trends in species occupancy related to land‐use change and management activities. Cluster sampling designs (Thompson ) can increase sampling efficiency over large areas by decreasing transit time and costs so that the same number of samples can be obtained for a lower amount of total effort (Giudice et al ). We demonstrated the application of our framework to a common sampling method, evaluated model performance under different ecological and sampling scenarios, and showed tradeoffs in efficiency among sampling designs.…”