I investigated mechanisms of parasitoid coexistence in a spatially structur
ed host-multiparasitoid community (harlequin bugs [Murgantia histrionica] a
nd two specialist parasitoids [Trissolcus murgantiae and Ooencyrtus johnson
ii]), I tested both local and metapopulation hypotheses. The local hypothes
is, intraguild predation, predicts coexistence if the inferior larval compe
titor is superior at finding unparasitized hosts. Hence, the superior larva
l competitor should be absent from patches of low host productivity. The me
tapopulation hypothesis, dispersal-competition trade-off, predicts coexiste
nce if the inferior competitor is a superior disperser. Hence, the superior
larval competitor should be absent from patches isolated by distance. Mani
pulative experiments demonstrate that coexistence does not require a disper
sal advantage to the inferior larval competitor. Field surveys show that pa
tches from which the superior larval competitor is absent are not the most
isolated, but the least productive. In a natural experiment, loss of the su
perior larval competitor was not associated with habitat loss or fragmentat
ion that increases distance among occupied patches, but with a large reduct
ion in host productivity. Taken together, these results strongly suggest th
at parasitoid coexistence occurs via local interactions rather than spatial
processes, This study provides the first empirical evidence of the role of
spatiotemporal variation in host productivity on parasitoid coexistence. T
he results have implications for multiparasitoid food webs in patchy enviro
nments and offer practical insights regarding the release of multiple paras
itoid species in pest control.