The ability to collect and process geographic information systems (GIS) data is becoming easier and allows rapid large-scale analysis. These enhanced capabilities may benefit many scientific fields, especially ecology studies. Many studies employ GIS analyses in order to determine the effects of human land use on organisms in local ecosystems. This requires choosing a spatial scale, which is often arbitrary and may lead to inconsistencies between studies, specifically when looking at highly variable components of an ecosystem. In an extreme example, two independent research projects could find opposite effects simply by having different spatial scales, even if they are accurately measuring the same phenomenon. The importance of choosing spatial scales is highlighted in this presentation by examples of GIS quantification of impervious land cover and forest cover for six study sites at three relatively common radii: 300, 500, and 800m. These land attributes were correlated with wild bee abundance from 2016 to 2019 as an example of how a study's results can vary based on scale.
Results/Conclusions
There were notable changes in both impervious land cover and forest cover across the three scales with greater changes in urban sites compared to rural areas. Correlations between bee abundance with impervious land cover were negative with the strength of these correlations increasing as scales became larger. For example, correlations between bee abundance and impervious land cover at 300m, 500m and 800m increases incrementally (R2 = 0.09, 0.1, and 0.12 respectively). Correlations between forest area and bee abundance also increased incrementally with increasing scale (R2 = 0.17, 0.15, and 0.21 respectively). Therefore, how does the investigator decide how many increases in spatial scale and to what extent are legitimate to achieve a better correlation? Incongruous conclusions from the same analysis at different spatial scales could be dependent on landscape uniformity as is typical of a rural setting compared to landscape heterogeneity in urban gradients. Therefore, these considerations become more important in situations where GIS are used to study urban ecology.