Support for an area–heterogeneity tradeoff for biodiversity in croplands.

Rapid expansion of the human population poses a challenge for wildlife conservation in agricultural landscapes. One proposal for addressing this challenge is to increase biodiversity in such landscapes by increasing crop diversity. However, studies report both positive and negative effects of crop diversity on biodiversity. One possible explanation, derived from the "area–heterogeneity tradeoff hypothesis," is that the effect of crop diversity on biodiversity depends on a tradeoff between increasing the number of crop types in a landscape and decreasing the amount of each single crop type. This should cause positive effects of increasing crop diversity at low to intermediate crop diversity and negative effects at intermediate to high crop diversity. We also propose two factors that could change the point at which the effect of increasing crop diversity shifts from positive to negative. First, we predicted that this shift would occur at a lower crop diversity when the surrounding landscape contains less semi‐natural habitat and at a higher crop diversity when the landscape contains more semi‐natural habitat. This should increase the likelihood of detecting negative effects of crop diversity when semi‐natural cover is low and positive effects when it is high. Second, we predicted that the shift from a positive to negative effect would occur at a lower crop diversity when it is measured locally than when it is measured at greater distances from the site, making detection of negative crop diversity effects more likely when measurements are at local extents. We tested these predictions using data on the biodiversity of herbaceous plants, butterflies, syrphid flies, woody plants, bees, carabid beetles, spiders, and birds at 221 crop field edges in Eastern Ontario, Canada. We found support for an area–crop diversity tradeoff. Semi‐natural cover and measurement extent influenced the biodiversity–crop diversity relationship, with positive effects when semi‐natural cover was high and negative effects when semi‐natural cover was low and when crop diversity was measured at local extents. The results suggest that policies/guidelines designed to increase crop diversity will not benefit biodiversity in the landscapes where conservation action is most urgently needed, that is, in landscapes with high agricultural use and low semi‐natural cover. [ABSTRACT FROM AUTHOR]