Goodwin, B.J. 2000. Landscape connectivity: The interaction between insect movements and landscape spatial structure. Ottawa-Carleton Institute of Biology, Carleton University, Ottawa, Ontario, Canada, 200 pp.

I assessed a patch predictor of insect densities by summarizing tests of the Resource Concentration Hypothesis (RCH) using meta-analysis. Support was weak overall and very heterogeneous between studies. Analysis of moderator variables indicated RCH holds when: (1) resource concentration is manipulated as plant diversity or patch size; (2) plants are high in secondary metabolites detected by insects; (3) insects are adults; (4) insects detect host-plants from afar; (5) insects tend to remain in host-plant patches; (6) host-plant patches are artificial or have conspicuous edges; and (7) host-plant patches are accessible. This suggests that patch predictors of population levels are insufficient and interpatch movement behaviour needs to be considered.

I investigated environment-movement interactions with a goldenrod beetle (Trirhabda borealis Blake, Chrysomelidae) in mosaic microlandscapes of goldenrod, cut vegetation, and cut vegetation with camouflage netting patches. Beetles in: (1) goldenrod patches exhibited short bursts of fast, meandering movements interspersed with long periods of immobility, (2) cut patches exhibited longer bouts of slow, directed movements with short periods of immobility, and (3) netting patches exhibited short bursts (same as goldenrod) of fast, meandering movements with short periods of immobility between (same as cut). Using mark and release experiments I determined that T. borealis did not respond to patch edges, detect goldenrod from afar, nor exhibit density-dependent emigration. Movements in goldenrod patches maintain contact with food while movements in cut patches expeditiously cross inhospitable areas. Netting patch structural complexity both hindered movements and presented a resource (anti-predator defense) worth maintaining contact with.

I investigated the impact of landscape structure on landscape connectivity using a combination of simulation and empirical experiments. Landscape connectivity was based on individual movements, and was measured in six different ways. Different aspects of landscape structure influenced these measures, suggesting that: (1) connectivity is a poorly defined concept, and (2) the same landscape may have different connectivity values when different measures of connectivity are used. Two general predictions that held over all measures of connectivity were: (1) increasing isolation decreases connectivity and (2) matrix elements were insignificant in comparison to habitat elements. These simulation results were supported in the empirical tests.