In dryland ecosystems, connectivity of bare soil patches is recognized as a key pattern attribute controlling resource conservation and structure-function relationships. Runoff redistribution from bare to vegetated patches concentrates water, enhancing vegetation growth and biomass. Conversely, a reduction in vegetation patches can lead to highly connected bare patches, triggering a threshold-like response with a large increase in hillslope runoff. Patterns of dryland vegetation are often used as indicators of landscape vulnerability, and there is increasing interest in using vegetation indices to predict connectivity-mediated abrupt shifts. One of the challenges to this goal lies in relating structural measures of connectivity – e.g., Flowlength – to functional measures, such as the source-to-sink path length of flow or the contributing source area. To address this challenge, we present a series of virtual runoff experiments on a patchily vegetated hillslope with varying length scales and spatial configurations of vegetation, and rainfall intensity and duration. Particle tracer methods are used to relate (i) the runoff coefficient to the contributing source area, and (ii) Flowlength as a measure of structural connectivity to source-sink path lengths, as a measure of functional connectivity. The results suggest a power-law relation between runoff coefficient and contributing source area that is stable across storm and vegetation characteristics. The relation between Flowlength and functional connectivity varies between storm intensities and durations, and is strongest for low-intensity storms. The results have practical implications for tracer studies, and support incorporating rainfall intensity into the use of vegetation patterns to predict landscape stability thresholds.