Abstract
Abstract.1. Absences in distributional data may result either from the true absence of a species or from a false absence due to lack of recording effort. I use general linear models (GLMs) and species distribution models (SDMs) to investigate this problem in North American Odonata and present a potential solution.2. I use multi‐model selection methods based on Akaike’s information criterion to evaluate the ability of water–energy variables, human population density, and recording effort to explain patterns of odonate diversity in the USA and Canada using GLMs. Water–energy variables explain a large proportion of the variance in odonate diversity, but the residuals of these models are significantly related to recorder effort.3. I then create SDMs for 176 species that are found solely in the USA and Canada using model averaging of eight different methods. These give predictions of hypothetical true distributions of each of the 176 species based on climate variables, which I compare with observed distributions to identify areas where potential under‐recording may occur.4. Under‐recording appears to be highest in northern Canada, Alaska, and Quebec, as well as the interior of the USA. The proportion of predicted species that have been observed is related to recorder effort and population density. Maps for individual species have been made available online (http://www.odonatacentral.org/) to facilitate recording in the future.5. This analysis has illustrated a problem with current odonate recording in the form of unbalanced recorder effort. However, the SDM approach also provides the solution, targeting recorder effort in such a way as to maximise returns from limited resources.
Highlights
Datasets of distributional data have been used extensively for a number of purposes including global change biology (Hickling et al, 2006), conservation (Polasky et al, 2001), pest management (Worner & Gevrey, 2006) and fisheries (Perry et al, 2005)
Present diversity Species richness in North American Odonata does not conform to the usual latitudinal gradient that is seen as a general global pattern
Exhaustive variable selection resulted a model containing the terms Potential evapotranspiration (PET), Global Vegetation Index (GVI) and mean annual precipitation as well as the area which was included by force
Summary
Datasets of distributional data have been used extensively for a number of purposes including global change biology (Hickling et al, 2006), conservation (Polasky et al, 2001), pest management (Worner & Gevrey, 2006) and fisheries (Perry et al, 2005). Insect conservation ecology has been hampered by the relative under-reporting of insect sightings (Dunn, 2005), making distributional data patchy. Attempts at collecting distributional data generally take one of two forms. The first form of recording is the standardised census. Standardised datasets have been used extensively in macroecological research (Harrington et al, 2007; Hill et al, 2002). With standardised methods the data from such surveys are relatively straightforward to analyse and trends can be obtained with a degree of certainty. The second, and historically more common, form of recording is the collection of reports of sightings. Due to the stochastic nature of record submission and the heavy reliance on the enthusiasm of the public both to observe and report specific taxa, these datasets tend to be extremely biased in time and space. There is, an interest both in quantifying variation in recorder effort and in directing future recording effort in such a way as to maximise returns from limited resources (Murdoch et al, 2007)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
