Massive floral display affects insect visits but not pollinator‐mediated pollen transfer in Rhododendron ferrugineum

Fragmentation of natural vegetation creates one of the largest threats to plant–pollinator interactions. Although fragmentation impacts on plant populations have been explored in many, mainly herbaceous, species, the response of wild mass‐flowering species is poorly known. Here, we studied 28 heathland patches dominated by the mass‐flowering shrub Rhododendron ferrugineum, each presenting different R. ferrugineum floral display sizes (total inflorescence number per patch) and patch isolation (median distance to the three nearest patches). We assessed the impacts of these two factors on (i) heathland patch visitor assemblage (considering R. ferrugineum versus surrounding community) and (ii) R. ferrugineum flower visitation rate and pollen transfer limitation (comparing seed set from emasculated to pollen‐supplemented flowers). We found that diversity and abundance of bees visiting R. ferrugineum in heathland patches significantly decreased with decreasing R. ferrugineum floral display, while overall visitor density per patch and flower visitation rate increased. Moreover, a decrease in massive floral display and increase in patch isolation resulted in reduced visitor density in the surrounding community. Even in patches with few individuals, we found disproportionate visitor abundance in R. ferrugineum compared to the surrounding community. Finally, pollen transfer limitation in R. ferrugineum was neither affected by visitation rate nor by patch attributes. By disproportionally attracting pollinators from co‐flowering species, and probably promoting geitonogamous pollen transfer, the mass‐flowering trait appears adequate to compensate, in terms of conspecific pollen transfer, for the decrease in visitor diversity and abundance and in mate availability, which usually result from population fragmentation.     

Distribution and colonisation ability of three parasitoids and their herbivorous host in a fragmented landscape

Habitat fragmentation can disrupt communities of interacting species even if only some of the species are directly affected by fragmentation. For instance, if parasitoids disperse less well than their herbivorous hosts, habitat fragmentation may lead to higher herbivory in isolated plant patches due to the absence of the third trophic level. Community-level studies suggest that parasitoids tend to have limited dispersal abilities, on the order of tens of metres, much smaller than that of their hosts, while species-oriented studies document dispersal by parasitoids on the scale of kilometres. In this study the distribution patterns of three parasitoid species with different life histories and their moth host, Hadena bicruris, a specialist herbivore of Silene latifolia, were compared in a large-scale network of natural fragmented plant patches along the rivers Rhine and Waal in the Netherlands. We examined how patch size and isolation affect the presence of each species. Additionally, experimental plots were used to study the colonisation abilities of the species at different distances from source populations.In the natural plant patches the presence of the herbivore and two of the parasitoids, the gregarious specialist Microplitis tristis and the gregarious generalist Bracon variator were not affected by patch isolation at the scale of the study, while the solitary specialist Eurylabus tristis was. In contrast to the herbivore, the presence of all parasitoid species declined with plant patch size. The colonisation experiment confirmed that the herbivore and M. tristis are good dispersers, able to travel at least 2 km within a season. B. variator showed intermediate colonisation ability and E. tristis showed very limited colonisation ability at this spatial scale. Characteristics of parasitoid species that may contribute to differences in their dispersal abilities are discussed.     

Habitat quality matters for the distribution of an endangered leaf beetle and its egg parasitoid in a fragmented landscape

Fragmentation, deterioration, and loss of habitat patches threaten the survival of many insect species. Depending on their trophic level, species may be differently affected by these factors. However, studies investigating more than one trophic level on a landscape scale are still rare. In the present study we analyzed the effects of habitat size, isolation, and quality for the occurrence and population density of the endangered leaf beetle Cassida canaliculata Laich. (Coleoptera: Chrysomelidae) and its egg parasitoid, the hymenopteran wasp Foersterella reptans Nees (Hymenoptera: Tetracampidae). C. canaliculata is strictly monophagous on meadow sage (Salvia pratensis), while F. reptans can also parasitize other hosts. Both size and isolation of habitat patches strongly determined the occurrence of the beetle. However, population density increased to a much greater extent with increasing host plant density ( = habitat quality) than with habitat size. The occurrence probability of the egg parasitoid increased with increasing population density of C. canaliculata. In conclusion, although maintaining large, well-connected patches with high host plant density is surely the major conservation goal for the specialized herbivore C. canaliculata, also small patches with high host plant densities can support viable populations and should thus be conserved. The less specialized parasitoid F. reptans is more likely to be found on patches with high beetle density, while patch size and isolation seem to be less important.     

Herbivore-induced responses and patch heterogeneity affect abundance of arthropods on plants

Abstract.  1. Plants respond to herbivore damage by inducing defences that can affect the abundance of herbivores and predators. These tritrophic interactions may be influenced by heterogeneity in plant neighbourhood.
2. In the present study, the effects of induced responses on the abundance of herbivores (flea beetles and aphids), omnivores (pirate bugs and thrips), and predators (lady beetles and spiders) on individual plants and their neighbours between and within patches composed of three tomato plants was investigated.
3. Herbivore damage was manipulated to create homogeneous patches where either all or none of the plants had defences induced by herbivore damage, and heterogeneous patches where only one of the plants was induced.
4. Arthropod abundance on plants at different scales was compared by testing between patch effects (patch level), for neighbourhood effects at the plant phenotype level (neighbourhood level), and between near and far plants (within patch position).
5. At the patch level , plants in homogeneously induced patches contained fewer flea beetles and pirate bugs, but more lady beetles, compared with homogeneously non-induced patches. There was no effect of patch type on the abundance of aphids, thrips, and spiders on plants.
6. At the neighbourhood level , induced plants in heterogeneous patches contained more flea beetles and pirate bugs compared with induced plants in homogeneous patches, indicating that the abundance of some herbivores and omnivores on induced plants varied depending on the phenotype of the other plants within the patch. Within patch position, there was no evidence that the abundance of herbivores or predators on non-induced plants was affected by proximity to an induced plant.
7. Therefore, variation in plant neighbourhood generated by induced plant responses affected the abundance of three arthropods from three feeding guilds.     

Local floral composition and the behaviour of pollinators: attraction to and foraging within experimental patches

1. Understanding how foraging decisions take place at the local scale is relevant because they may directly affect the fitness of individual plants. However, little is known about how local diversity and density affect the foraging behaviour of most pollinator groups. 2. By introducing two potted plant species (Salvia farinacae and Tagetes bonanza) into two populations of Taraxacum officinale, we investigated how plant identity, the mixtures of these plant species, and total plant density affected the attraction to and the foraging within a patch for six pollinator groups. 3. The foraging behaviour was mainly driven by the availability of the preferred plant species, and secondly by patch diversity and density. In general, dense patches and those containing the three‐species mixture were preferred by all insect groups for arrival, although muscoid and hover flies responded less to local floral composition than bees. Local diversity and density had, however, a weaker effect on foraging behaviour within patches. Site dependence in response to floral treatments could be attributable to differences between sites in pollinator assemblage and Taraxacum density. 4. Studies like ours will help to understand how foraging decisions occur at the local scale and how foraging patterns may differ between pollinators and sites.     

Landscape context outweighs local habitat quality in its effects on herbivore dispersal and distribution

Past studies with spatially structured herbivore populations have emphasized the primacy of intrinsic factors (e.g., patch quality), patch geometry (e.g., patch size and isolation), and more recently landscape context (e.g., matrix composition) in affecting local population abundance and dispersal rate. However, few studies have examined the relative importance of each factor, or how they might interact to affect herbivore abundance or dispersal. Here, we performed a factorial field experiment to examine the independent and interactive effects of patch quality (plant biomass, leaf protein, leaf phenolics) and matrix composition [mudflat or non-host grass (Bromus inermis)] on planthopper (Prokelisia crocea) emigration from host-plant patches (prairie cordgrass, Spartina pectinata). In addition, a field survey was conducted to examine the relative importance of patch quality, geography, and matrix composition on planthopper occupancy and density. In the experiment, we found that rates of emigration from low and intermediate quality patches were, on average, 21% percent higher for patches embedded in brome than mudflat. In contrast, the emigration rate was unaffected by matrix composition in nutrient-rich patches. Within matrix types, plant quality had little effect on emigration. In the survey, planthopper density and the patch occupancy rate of planthoppers increased nonadditively with increasing patch size and the percentage of the surrounding matrix composed of mudflat. This study suggests that landscape-level factors, such as the matrix, may be more important than factors intrinsic to the patches.     

Individually mark–mass release–resight study elucidates effects of patch characteristics and distance on host patch location by an insect herbivore

1. How organisms locate their hosts is of fundamental importance in a variety of basic and applied ecological fields, including population dynamics, invasive species management and biological control. However, tracking movement of small organisms, such as insects, poses significant logistical challenges. 2. Mass‐release and individual–mark–recapture techniques were combined in an individually mark–mass release–resight (IMMRR) approach to track the movement of over 2000 adult insects in an economically important plant–herbivore system. Despite its widespread use for the biological control of the invasive thistle Carduus nutans, the host‐finding behaviour of the thistle head weevil Rhinocyllus conicus has not previously been studied. Insects were released at different distances from a mosaic of artificially created host patches with different areas and number of plants to assess the ecological determinants of patch finding. 3. The study was able to characterize the within‐season dispersal abilities and between‐patch movement patterns of R. conicus. Weevils found host plant patches over 900 m away. Large patches, with tall plants, situated close to the nearest release point had the highest first R. conicus resights. Patch area and plant density had no effect on the number of weevils resighted per plant; however, R. conicus individuals were more likely to disperse out of small patches and into large patches. 4. By understanding how R. conicus locates host patches of C. nutans, management activities for the control of this invasive thistle can be better informed. A deeper mechanistic understanding of host location will also improve prediction of coupled plant–herbivore spatial dynamics in general.     

A comparison of the responses of two tropical specialist herbivores to host plant patch size

Summary The effects of host plant patch size on the abundances of two specialist herbivores (the chrysomelid beetle, Acalymma innubum and the pentatomid bug, Piezosternum subulatum) were investigated in a natural forest community in the Virgin Islands. Abundances were compared early and late in the season in different sized patches of the cucurbit host plant (Cayaponia americana) growing in open habitat (with no surrounding plant community) and forest habitat (with diverse surrounding plant community). For both herbivore species, adult abundances per patch were positively correlated with patch leaf area, but there was a significant patch size effect (i.e., correlation between herbivore density per unit plant and patch leaf area) only for beetles in the forest habitat. Both herbivore species were significantly affected by surrounding plant diversity, but in opposite ways: beetles were more abundant in open patches whereas bugs were more abundant in forest patches. Relationships between abundance and patch size in open and forest patches changed through the season for both herbivore species. These changing abundance patterns are discussed with respect to (1) increases in the diversity of the plant community surrounding host plant patches, and (2) differences in herbivore movement patterns.     

Patch occupancy of two hemipterans sharing a common host plant

Aim Two hemipteran species were chosen as a study system for the comparative analysis of patch occupancy and spatial population structure of insects sharing a common host plant. This study tested whether (1) the incidence in the host plant patches differed between the two species, and (2) the two species exhibited a different spatial population structure, i.e. were they affected differentially by isolation and area of the host plant patches. Location The porphyry landscape north of Halle (Saale) in Germany comprising 506 patches of the host plant Brachypodium pinnatum. Methods The host plant patches were surveyed for the two hemipterans. To assess the influence of patch quality on species occurrence the patches were characterized by mean cover abundance of B. pinnatum, type of subsoil, slope, exposure, and shading. The spatial configuration of the patches was considered by patch area and isolation. The influence of the habitat factors and the spatial configuration on the occupancy of the two species was analysed by logistic regression. Results Adarrus multinotatus was found in 441 patches, while Neophilaenus albipennis was found in only 90 patches. While A. multinotatus showed virtually no relationship to the habitat factors, the occupancy of N. albipennis was influenced by subsoil type, cover abundance, and shading. The effects of area and isolation on occupancy of the patches also differed between the two species. The occupancy of N. albipennis was determined largely by area and isolation, whereas in A. multinotatus no considerable effect of spatial configuration was found. Main conclusions The study revealed a marked difference between the two hemipteran species in respect of spatial population structure. Adarrus multinotatus built up a ‘patchy population’, whereas N. albipennis showed a ‘metapopulation’ structure within the same set of patches in the same landscape. Spatial population structure was found to be not only a function of spatial configuration of habitat patches, but population structure differed between the habitat generalist A. multinotatus and the habitat specialist N. albipennis.     

Resource partitioning by three species of hemipteran herbivores on the basis of host plant density

Summary Three hemipteran herbivores partition a common resource plant, Senecio smallii, on the basis of bud density and plant patch size. The monophagous herbivore, Neacoryphus bicrucis, is most abundant in larger patches where bud density is greatest. The oligophagous herbivore, Harmostes reflexulus, is most abundant in small patches of high bud density. The polyphagous species, Lygus pratensis, is most abundant in small patches where abundance is independent of bud density. The aggressive nature of the monophagous herbivore appears to mediate the responses of the other species.     

Effects of Patch Size and Floral Herbivory on Seed Set in Trichogoniopsis adenantha (Asteraceae) in Southeastern Brazil1

The effects of patch size and floral herbivory on reproductive success of Trichogoniopsis adenantha (Asteraceae) were studied in a semideciduous forest in southeastern Brazil. Plants in large patches and capitula not attacked by endophagous insects had higher seed set. The consequences of living in large or small patches for plant reproductive success are discussed.     

Regional patch dynamics of Cirsium arvense and possible implications for plant‐animal interactions

Abstract. Plant population biology considers the dynamics of plant modules within stands. However, stands themselves may have considerable regional turnover in space and time. These changes in the number, distribution and size of plant stands generate a dynamic spatial pattern with important implications for the spatial and temporal dynamics of phytophagous insects using these plants as a host. During five successive years we studied the regional distribution and patch dynamics of the creeping thistle Cirsium arvense and the distribution of associated populations of the herbivore Urophora cardui (Diptera: Tephritidae), a specialist stem gall former. The study conducted was in a 15 km² heterogeneous, agricultural area in northeastern Bavaria. The distribution of the number of plants per patch was skewed with many more small C. arvense patches than large ones. During the five years of study, there was a 50% increase in the number of C. arvense patches, and a decrease in the mean number of plants per patch (= patch size) to less than half the patch size of the first year. Whilst patch size was randomly distributed in space, patch density showed a consistent, non‐random spatial pattern. Patch density was spatially auto‐correlated, with areas of high or low patch density having a characteristic dimension of ca. 1 km. Patch size was predictable in time and appeared to be regulated by size dependent processes, with the extinction probability of a patch being negatively correlated with its size. Correlated with the decline of C. arvense patch size during the study, the occupancy and total numbers of the herbivore U. cardui had a marked decrease, suggesting that the regional distribution of the stem gall former is not only influenced by patch number but more importantly by the mean patch size. With decreasing patch sizes, U. cardui was faced with an increasingly dynamic landscape due to higher extinction rates of small patches, although the mean distance between host plant patches decreased.     

Restoration genetics of the vernal pool endemic Lasthenia conjugens (Asteraceae)

Restoration of habitat for endangered species often involves translocation of seeds or individuals from source populations to an area targeted for revegetation. Long-term persistence of a species is dependent on the maintenance of sufficient genetic variation within and among populations. Thus, knowledge and maintenance of genetic variability within rare or endangered species is essential for developing effective conservation and restoration strategies. Genetic monitoring of both natural and restored populations can provide an assessment of restoration protocol success in establishing populations that maintain levels of genetic diversity similar to those in natural populations. California’s vernal pools are home to many endangered plants, thus conservation and restoration are large components of their management. Lasthenia conjugens (Asteraceae) is a federally endangered self-incompatible vernal pool annual with gravity- dispersed seeds. Using the molecular technique of intersimple sequence repeats (ISSRs), this study assessed levels and patterns of genetic variability present within natural and restored populations of L. conjugens. At Travis Air Force Base near Fairfield, California, a vernal pool restoration project is underway. Genetic success of the ecologically based seeding protocol was examined through genetic monitoring of natural and restored populations over a three-year period. Genetic diversity remained constant across the three sampled generations. Diversity was also widely distributed across all populations. We conclude that the protocol used to establish restored populations was successful in capturing similar levels and patterns of genetic diversity to those seen within natural pools. This study also demonstrates how genetic markers can be used to inform conservation and restoration decisions.     

Spatial and Temporal Dynamics of a Restored Population of Oryza rufipogon in Huli Marsh, South China

Studying the process of population restoration is helpful for managing and preserving endangered species. A population of Oryza rufipogon (wild rice), an endangered species, was reintroduced in 1993 into Huli Marsh. We conducted a detailed survey over a 5‐year period (1997–2001) to evaluate the present status of the population and to further our understanding of its habitat requirements and the population model. The population was surveyed using 2 × 2–m quadrats in mid‐September of each year. In total, 2,683 quadrats were surveyed covering the whole O. rufipogon reserve during each survey. The population's spatial distribution was mapped, and the maps were used to examine the relationship between patch replacement and water depth. The individual number of O. rufipogon increased steadily from 1993 to 2001. The patch number, patch area, mean patch size, and largest patch size increased over this time period, and Korcak patchiness exponents decreased. On average, 83% of the patches persisted from one year to the next. There was a significant positive correlation between the initial patch size and the size the following year. The probability of patch disappearance decreased as patch size increased. Fifty‐eight percent of the patches were located at water depths between 20 and 30 cm. Water depth had no significant effect on the patch transition from O. rufipogon to other species. The loss and gain of O. rufipogon patches were statistically correlated with the patch areas in different water depths. Our results show that the population of O. rufipogon can successfully be reintroduced to the original habitat after appropriate environmental conditions have been restored. We recommend the following transplantation practices: transplant many smaller patches rather than a few larger patches, use transplant patch sizes of at least 20 m², and transplant into sites vegetated with species with different regeneration niches from the transplanted species.     

Temporal variation in the effects of habitat fragmentation on reproduction of the Mediterranean shrub <Emphasis Type="Italic">Colutea hispanica</Emphasis>

Habitat fragmentation poses a major threat to the viability of plant populations. However, the intensity of fragmentation effects may vary among years. We studied two possible effects of habitat fragmentation (patch size and isolation) on the reproduction and proportion of damaged fruits in 24 patches of the self-compatible shrub Colutea hispanica for three consecutive years with different climate conditions. We also studied the effect of fragmentation on the incidence of two main pre-dispersal seed predators, the butterflies Iolana iolas and Lampides boeticus. High between-year variability was found in number of viable seeds per fruit, number of fruits per plant, total number of viable seeds per plant and proportion of damaged fruits. In 2003, small, isolated patches had a higher fruit set and number of fruits per plant. The proportion of damaged fruits was significantly lower in isolated populations in 2003, while it was very high in all patches in 2004 and 2005. High between-year variability was also found in the proportion of fruits per plant with I. iolas eggs. In 2003 isolated patches had a lower proportion of fruits with I. iolas eggs, but no significant effect of patch size and isolation was found in 2004 or 2005. The proportion of fruits with L. boeticus eggs was similar in the three years of study, although it was slightly higher in large, non-isolated patches in 2003. Thus, the effects of fragmentation on plant reproduction cannot be generalized from one single-year survey. In contrast to the generally accepted idea that fragmentation reduces plant reproduction, plant fitness may increase in isolated patches in years with high fruit production and low seed predation.     

Population genetic diversity and structure of two rare vernal pool grasses in central California

Vernal pool ecosystems are declining throughout California, with only 10% of historic habitat remaining. This has endangered many specialist endemic plant species, leaving extant populations fragmented, isolated, and threatened or endangered. Recovery plans for the increasing number of endangered vernal pool species require information on their genetic and ecological status to guide conservation and restoration efforts. Federally threatened Neostapfia colusana (Colusa grass) and federally endangered Tuctoria greenei (Greene’s tuctoria) are two endemic vernal pool grasses of high conservation concern in central California. Remaining populations are highly fragmented due to range-wide habitat destruction. Using five polymorphic microsatellite markers for each species, we performed genetic surveys of 240 individuals from eight vernal pools for N. colusana, and 317 individuals from 13 vernal pools for T. greenei. We detected high within-population genetic diversity for both species, with average allelic diversities of 24 alleles/locus (mean Hobs = 0.68, mean Hexp = 0.71) for N. colusana, and 19 alleles/locus (mean Hobs = 0.77, and mean Hexp = 0.79) for T. greenei. Bayesian clustering and AMOVA indicated two genetically distinct population groups for N. colusana (Fst = 0.268, P < 0.0001), and three for T. greenei (Fst = 0.11, P < 0.0001). We found very slight temporal genetic structure at one N. colusana (Fst = 0.013, P < 0.05) and two T. greenei (Fst = 0.015, Fst = 0.018, P < 0.05) pools. These estimates of population genetic diversity and structure are critical measures for both species that will help inform recovery management actions.     

Interpatch movement of the red milkweed beetle,Tetraopes tetraophthalmus: individual responses to patch size and isolation

Individual movement patterns and the effects of host plant patch size and isolation on patch occupancy were examined for red milkweed beetles, Tetraopes tetraophthalmus, residing in a heterogeneous landscape. Male beetles were found to move both more often and farther between host plant patches than female beetles, and this difference affected the patterns of patch occupancy observed. Overall, unoccupied milkweed patches were smaller and more isolated than patches occupied by beetles. Patches uninhabited by females tended to be more isolated, but not necessarily smaller, than patches with female beetles, indicating that females may be affected more by patch isolation than patch size. Presence of male beetles on patches showed a stronger response to patch size than to patch isolation. Differences in movement between males and females illustrate the need for demographically based dispersal data. Comparisons of Tetraopes interpatch movement patterns between landscapes composed of patches of different size revealed that landscapes with overall smaller patches may have greater rates of interpatch movement.     

Avian seed dispersal and seedling distribution of the endangered tree species,Taxus chinensis,in patchy habitats

Background: There is limited understanding about bird dispersal behaviour and seedling distribution of endangered tree species in patchy environments, although these processes are important for plant species persistence.

Aims: We tested how patch features affected bird behaviour and seed dispersal, and thus seedling distribution of the endangered Chinese yew tree (Taxus chinensis).

Methods: In the present study, we combined field data of bird dispersal behaviour and GIS-based information to elucidate the influence of spatial features of habitat patches on bird dispersal behaviour, and the resulting effects on the seedling distribution of the endangered Chinese yew in two patchy habitats.

Results: Our results showed that the only seed source patch could attract eight bird species for dispersal at the two sites. Post-foraging movements of bird dispersers was strongly related to both topography and the relative locations of habitat patches. Yew seedlings aggregated only at the seed source and bamboo recruitment patches, which was affect by both the spatial distribution of recruitment patches and patch use by dispersers.

Conclusions: Our results emphasise that bamboo patches in both patchy environments provide the necessary conditions for germination of yew seeds, and the post-foraging behaviour of dispersers determines seed deposited in these patches. Our study highlights the importance of the dispersal behaviour of frugivorous birds in the successful regeneration and colonisation of yew populations in patchy habitats.     

Monitoring Golden-Cheeked Warblers on Private Lands in Texas

ABSTRACT A majority of North American breeding habitat for neotropical migrants exists on private lands, requiring monitoring strategies focused on habitat in these private holdings. We outline study designs and protocols using repeated Presence-Absence surveys across a gradient of patch sizes to develop a range-wide monitoring program for the endangered golden-cheeked warbler (Dendroica chrysoparia) in Texas, USA. We surveyed 200–400 point-count locations across approximately 30 private properties annually from 2005 to 2008. We used data from our surveyed patches (n = 147) and the Ψ (occupancy), p (detection), and γ = 1 - ɛ parameterization to estimate patch dynamics and associated detection probabilities for golden-cheeked warblers. Patch size had a strong association with patch occupancy, and all patches >160 ha were predicted to be occupied. We found no evidence that large golden-cheeked warbler populations located on public lands in the vicinity of our study area influenced occupancy dynamics. We conducted simulations across a range of detection probabilities to evaluate potential sample sizes for both standard- and removal-based occupancy modeling. Simulations using parameter estimates from our analysis indicated that removal-based sampling is superior to standard sampling. Based on our results, surveying golden-cheeked warbler presence in oak-juniper (Quercus-Juniperus) patches under a removal modeling framework should be considered as one alternative for range-wide monitoring programs because patch-level monitoring would be necessary to estimate proportion of range occupied. Large contiguous patches are rare across the species’ range; hence, conservation and management of the mosaic of smaller patches within a landscape context would be required for maintaining species viability. Thus, we recommend the identification of areas where smaller, contiguous patches represent a significant portion of the available habitat within the local landscape and targeting these areas for habitat maintenance and improvement.     

Local variation in plant quality influences large‐scale population dynamics

Spatial variation in ecological systems can arise both as a consequence of variation in the quality and availability of resources and as an emergent property of spatially structured interactions. We used a spatially explicit model to simulate populations of herbivore hosts and their parasitoids in landscapes with different levels of variance in plant patch quality and different spatial arrangements of high‐ and low‐quality plant patches. We found that even small variation in patch quality at a fine spatial scale decreased overall herbivore populations, as parasitoid populations on low‐quality plant patches were subsidized by those from high‐quality neighbors. On landscapes with large, homogeneous regions of high‐ and low‐quality plant patches, herbivore populations increased with variation in patch quality. Overall, our results demonstrate that local variation in resource quality profoundly influences global population dynamics. In particular, fine‐scale variation in plant patch quality enhanced biological control of herbivores by parasitoids, suggesting that adding back plant genetic variation into perennial production systems may enhance the biological control of herbivores by their natural enemies.