COS 28-6 - The effects of woody vegetation remnants on insect herbivore distribution: Testing bottom-up and top-down effects

Tuesday, August 13, 2019: 3:20 PM
L007/008, Kentucky International Convention Center
Douglas Lawton1, Cathy Waters2,3, Marion Le Gall4 and Arianne Cease1,3, (1)School of Life Sciences, Arizona State University, Tempe, AZ, (2)Dubbo Regional Office, NSW Department of Primary Industries, Dubbo, NSW, Australia, (3)School of Sustainability, Arizona State University, Tempe, AZ, (4)Julie Ann Wrigley Global Institute of Sustainability, Arizona State University, Tempe, AZ
Background/Question/Methods

Increasing anthropogenic mediated fragmentation has had considerable impacts on many aspects of biodiversity. One of the major contributors to fragmentation is agriculture which has changed landscape biogeography, nutrient cycling, and the distribution and connectivity of animal populations. For example, Australian plague locusts (Chortoicetes terminifera; CT) prefer open grasslands and avoid tree patches, however the mechanisms underlying this aversion remain unknown. In this study, we explore how tree stand patches within grazing pastures affect CT abundance and distribution, and the potential contributions of top down versus bottom up effects. We have three hypotheses:

  1. Trees are ‘islands of fertility’ and result in nutritionally suboptimal grasses for CT
  2. CT have decreased threat of predation in open grasslands
  3. Open grasslands are warmer and CT prefer high temperatures

We selected three fields with similar life histories in Trangie, NSW, Australia. Along 50m transects from inside a tree patches to open grassland gradient, we measured CT abundance, daily temperature, invertebrate predator abundance, ground cover, and grass nutrient content.

Results/Conclusions

We found that CT are avoiding tree stands by at least 20 meters, however temperature, predator abundance, nor grass nutrient content are independently explaining this phenomenon. Grass nutrients did not vary based on the tree patches to grass land gradient. Ground temperature was the highest within the tree patches. CT abundance was not significantly impacted by proximity to meat ant hill (Iridomyrmex purpureus) nor correlated with invertebrate predator density from pitfall trapping. An overall generalized additive model revealed the most important variables were distance from tree patches (positive linear correlation), plant protein content (negative linear correlation), and a nonlinear relationship with percent bareground (abundance was highest at 20% bareground).

Understanding the impacts of the agricultural/natural vegetation fragment mosaic on the ecological and behavioral traits of grasshoppers is important for large scale population dynamics. CT clearly avoid treestands, but there are no direct connections with the variables measured in this study. However, host plant nutrient content and ground cover were important overall predictors of CT. Further studies on thermal micro-climates may reveal addition patterns. CT population management solutions should consider woody encroachment as this is likely affecting CT abundances and distributions on multiple scales.