Increasing crop field size does not consistently exacerbate insect pest problems.

Increasing diversity on farms can enhance many key ecosystem services to and from agriculture, and natural control of arthropod pests is often presumed to be among them. The expectation that increasing the size of monocultural crop plantings exacerbates the impact of pests is common throughout the agroecological literature. However, the theoretical basis for this expectation is uncertain; mechanistic mathematical models suggest instead that increasing field size can have positive, negative, neutral, or even nonlinear effects on arthropod pest densities. Here, we report a broad survey of crop field-size effects: across 14 pest species, 5 crops, and 20,000 field years of observations, we quantify the impact of field size on pest densities, pesticide applications, and crop yield. We find no evidence that larger fields cause consistently worse pest impacts. The most common outcome (9 of 14 species) was for pest severity to be independent of field size; larger fields resulted in less severe pest problems for four species, and only one species exhibited the expected trend of larger fields worsening pest severity. Importantly, pest responses to field size strongly correlated with their responses to the fraction of the surrounding landscape planted to the focal crop, suggesting that shared ecological processes produce parallel responses to crop simplification across spatial scales. We conclude that the idea that larger field sizes consistently disrupt natural pest control services is without foundation in either the theoretical or empirical record.

Phylogenetic escape from pests reduces pesticides on some crop plants.

Pesticides are a ubiquitous component of conventional crop production but come with considerable economic and ecological costs. We tested the hypothesis that variation in pesticide use among crop species is a function of crop economics and the phylogenetic relationship of a crop to native plants because unrelated crops accrue fewer herbivores and pathogens. Comparative analyses of a dataset of 93 Californian crops showed that more valuable crops and crops with close relatives in the native plant flora received greater pesticide use, explaining roughly half of the variance in pesticide use among crops against pathogens and herbivores. Phylogenetic escape from arthropod and pathogen pests results in lower pesticides, suggesting that the introduced status of some crops can be leveraged to reduce pesticides.

The causes and consequences of pest population variability in agricultural landscapes.

Variability in population densities is key to the ecology of natural systems but also has great implications for agriculture. Farmers' decisions are heavily influenced by their risk aversion to pest outbreaks that result in major yield losses. However, the need for long-term pest population data across many farms has prevented researchers from exploring the drivers and implications of pest population variability (PV). Here, we demonstrate the critical importance of PV for sustainable farming by analyzing 13 years of pest densities across >1300 Spanish olive groves and vineyards. Variable populations were more likely to cause major yield losses, but also occasionally created temporal windows when densities fell below insecticide spray thresholds. Importantly, environmental factors regulating pest variability were very distinct from factors regulating mean density, suggesting variability needs to be uniquely managed. Finally, we found diversifying landscapes may be a win-win situation for conservation and farmers, as diversified landscapes promote less abundant and less variable pest populations. Therefore, we encourage agricultural stakeholders to increase the complexity of the landscapes surrounding their farms through conserving/restoring natural habitat and/or diversifying crops.

A double-edged sword: parental care increases risk of offspring infection by a maternally vectored parasite.

Parental care can protect offspring from predators but can also create opportunities for parents to vector parasites to their offspring. We hypothesized that the risk of infection by maternally vectored parasites would increase with the frequency of mother-offspring contact. Ammophila spp. wasps (Hymenoptera: Sphecidae) build nests in which they rear a single offspring. Ammophila species exhibit varied offspring provisioning behaviours: some species enter the nest once to provision a single, large caterpillar, whereas others enter the nest repeatedly to provision with many smaller caterpillars. We hypothesized that each nest visit increases the risk of offspring parasitism by Paraxenos lugubris (Strepsiptera: Xenidae), whose infectious stages ride on the mother wasp (phoresy) to reach the vulnerable Ammophila offspring. We quantified parasitism risk by external examination of museum-curated Ammophila specimens-the anterior portion of P. lugubris protrudes between the adult host's abdominal sclerites and reflects infection during the larval stage. As predicted, Ammophila species that receive larger numbers of provisions incur greater risks of parasitism, with nest provisioning behaviour explaining ca 90% of the interspecific variation in mean parasitism. These findings demonstrate that parental care can augment, rather than reduce, the risk of parasite transmission to offspring.

Landscape simplification increases vineyard pest outbreaks and insecticide use.

Diversifying agricultural landscapes may mitigate biodiversity declines and improve pest management. Yet landscapes are rarely managed to suppress pests, in part because researchers seldom measure key variables related to pest outbreaks and insecticides that drive management decisions. We used a 13-year government database to analyse landscape effects on European grapevine moth (Lobesia botrana) outbreaks and insecticides across c. 400 Spanish vineyards. At harvest, we found pest outbreaks increased four-fold in simplified, vineyard-dominated landscapes compared to complex landscapes in which vineyards are surrounded by semi-natural habitats. Similarly, insecticide applications doubled in vineyard-dominated landscapes but declined in vineyards surrounded by shrubland. Importantly, pest population stochasticity would have masked these large effects if numbers of study sites and years were reduced to typical levels in landscape pest-control studies. Our results suggest increasing landscape complexity may mitigate pest populations and insecticide applications. Habitat conservation represents an economically and environmentally sound approach for achieving sustainable grape production.

Crop pests and predators exhibit inconsistent responses to surrounding landscape composition.

The idea that noncrop habitat enhances pest control and represents a win-win opportunity to conserve biodiversity and bolster yields has emerged as an agroecological paradigm. However, while noncrop habitat in landscapes surrounding farms sometimes benefits pest predators, natural enemy responses remain heterogeneous across studies and effects on pests are inconclusive. The observed heterogeneity in species responses to noncrop habitat may be biological in origin or could result from variation in how habitat and biocontrol are measured. Here, we use a pest-control database encompassing 132 studies and 6,759 sites worldwide to model natural enemy and pest abundances, predation rates, and crop damage as a function of landscape composition. Our results showed that although landscape composition explained significant variation within studies, pest and enemy abundances, predation rates, crop damage, and yields each exhibited different responses across studies, sometimes increasing and sometimes decreasing in landscapes with more noncrop habitat but overall showing no consistent trend. Thus, models that used landscape-composition variables to predict pest-control dynamics demonstrated little potential to explain variation across studies, though prediction did improve when comparing studies with similar crop and landscape features. Overall, our work shows that surrounding noncrop habitat does not consistently improve pest management, meaning habitat conservation may bolster production in some systems and depress yields in others. Future efforts to develop tools that inform farmers when habitat conservation truly represents a win-win would benefit from increased understanding of how landscape effects are modulated by local farm management and the biology of pests and their enemies.

Bugs scaring bugs: enemy-risk effects in biological control systems.

Enemy-risk effects, often referred to as non-consumptive effects (NCEs), are an important feature of predator-prey ecology, but their significance has had little impact on the conceptual underpinning or practice of biological control. We provide an overview of enemy-risk effects in predator-prey interactions, discuss ways in which risk effects may impact biocontrol programs and suggest avenues for further integration of natural enemy ecology and integrated pest management. Enemy-risk effects can have important influences on different stages of biological control programs, including natural enemy selection, efficacy testing and quantification of non-target impacts. Enemy-risk effects can also shape the interactions of biological control with other pest management practices. Biocontrol systems also provide community ecologists with some of the richest examples of behaviourally mediated trophic cascades and demonstrations of how enemy-risk effects play out among species with no shared evolutionary history, important topics for invasion biology and conservation. We conclude that the longstanding use of ecological theory by biocontrol practitioners should be expanded to incorporate enemy-risk effects, and that community ecologists will find many opportunities to study enemy-risk effects in biocontrol settings.

Disease, contagious cannibalism, and associated population crash in an omnivorous bug, Geocoris pallens.

Disease and cannibalism are two strongly density-dependent processes that can suppress predator populations. Here we show that California populations of the omnivorous predatory bug Geocoris pallens are subject to infection by a pathogen, as yet unidentified, that elicits elevated expression of cannibalism. Laboratory experiments showed that the pathogen is moderately virulent, causing flattened abdomens, elevated nymphal mortality, delayed development, and reduced body size of adult females. Infection furthermore increases the expression of cannibalism. Field populations of Geocoris spp. declined strongly in association with sharp increases in the expression of egg cannibalism by adult G. pallens. Increased cannibalism was accompanied by a strongly bimodal distribution of cannibalism expression, with some females (putatively uninfected) expressing little cannibalism and others (putatively infected) consuming most or all of the eggs present. Highly cannibalistic females did not increase their consumption of Ephestia cautella moth eggs, suggesting that the high cannibalism phenotype reflected a specific loss of restraint against eating conspecifics. Highly cannibalistic females also often exhibited reduced egg laying, consistent with a virulent pathogen; less frequently, more cannibalistic females exhibited elevated egg laying, suggesting that cannibalism might also facilitate recycling of nutrients in eggs. Elevated cannibalism was not correlated with reduced prey availability or elevated field densities of G. pallens. Geocoris pallens population crashes appear to reflect the combined consequences of direct virulence-adverse pathogen effects on the infected host's physiology-and indirect virulence-mortality of both infected and uninfected individuals due to elevated cannibalism expression by infected individuals.

Ecoinformatics (Big Data) for Agricultural Entomology: Pitfalls, Progress, and Promise.

Ecoinformatics, as defined in this review, is the use of preexisting data sets to address questions in ecology. We provide the first review of ecoinformatics methods in agricultural entomology. Ecoinformatics methods have been used to address the full range of questions studied by agricultural entomologists, enabled by the special opportunities associated with data sets, nearly all of which have been observational, that are larger and more diverse and that embrace larger spatial and temporal scales than most experimental studies do. We argue that ecoinformatics research methods and traditional, experimental research methods have strengths and weaknesses that are largely complementary. We address the important interpretational challenges associated with observational data sets, highlight common pitfalls, and propose some best practices for researchers using these methods. Ecoinformatics methods hold great promise as a vehicle for capitalizing on the explosion of data emanating from farmers, researchers, and the public, as novel sampling and sensing techniques are developed and digital data sharing becomes more widespread.

Modest Pollen Limitation of Lifetime Seed Production Is in Good Agreement with Modest Uncertainty in Whole-Plant Pollen Receipt.

We recently introduced a model that predicts the degree to which a plant's lifetime seed production may be constrained by unpredictable shortfalls of pollen receipt ("pollen limitation"). Burd's comment in this issue criticized our analysis, first by arguing that the empirical literature documents much higher levels of pollen limitation than our model predicts and then suggesting that the apparent discrepancy stemmed from our (1) underestimating the costs of securing a fertilized ovule and (2) assuming too little unpredictability in whole-plant pollen receipt. We reply as follows. First, the empirical literature must be consulted carefully. Burd relies on pollen supplementation experiments performed on parts of plants or on whole plants but during only one reproductive season for polycarpic perennials; in both cases, resource reallocation often leads to gross overestimates of pollen limitation. We comprehensively review pollen limitation estimates that are free of these estimation problems and find strong agreement with our model predictions. Second, although cost estimates for different components of seed production are imprecise, errors are likely to be small relative to the >1,000-fold differences observed across plant species, the primary focus of our article. Finally, contrary to Burd's argument, pollen receipt by entire plants is much more predictable than that by individual flowers because the flower-to-flower variation "averages out" when summed across many flowers. Our model uses parameter values that are in broad agreement with the empirical record of modest plant-to-plant variation in pollen receipt and thus predicts the generally modest pollen limitation that is observed in nature.

Does an 'oversupply' of ovules cause pollen limitation?

Lifetime seed production can be constrained by shortfalls of pollen receipt ('pollen limitation'). The ovule oversupply hypothesis states that, in response to unpredictable pollen availability, plants evolve to produce more ovules than they expect to be fertilized, and that this results in pollen limitation of seed production. Here, we present a cartoon model and a model of optimal plant reproductive allocations under stochastic pollen receipt to evaluate the hypothesis that an oversupply of ovules leads to increased pollen limitation. We show that an oversupply of ovules has two opposing influences on pollen limitation of whole-plant seed production. First, ovule oversupply increases the likelihood that pollen receipt limits the number of ovules that can be fertilized ('prezygotic pollen limitation'). Second, ovule oversupply increases the proportion of pollen grains received that are used to fertilize ovules ('pollen use efficiency'). As a result of these opposing influences, ovule oversupply has only a modest effect on the degree to which lifetime seed production is constrained by pollen receipt, producing a small decrease in the incidence of pollen limitation. Ovule oversupply is not the cause of the pollen limitation problem, but rather is part of the evolutionary solution to that problem.

Cannibalism amplifies the spread of vertically transmitted pathogens.

Cannibalism is a widespread behavior. Abundant empirical evidence demonstrates that cannibals incur a risk of contracting pathogenic infections when they consume infected conspecifics. However, current theory suggests that cannibalism generally impedes disease spread, because each victim is usually consumed by a single cannibal, such that cannibalism does not function as a spreading process. Consequently, cannibalism cannot be the only mode of transmission of most parasites. We develop simple, but general epidemiological models to analyze the interaction of cannibalism and vertical transmission. We show that cannibalism increases the prevalence of vertically transmitted pathogens whenever the host population density is not solely regulated by cannibalism. This mechanism, combined with additional, recently published, theoretical mechanisms, presents a strong case for the role of cannibalism in the spread of infectious diseases across a wide range of parasite-host systems.

The epidemiology and evolution of parasite transmission through cannibalism.

Cannibalism is a widespread behavior, and evidence is abundant for transmission from infected victims to susceptible cannibals in many parasite-host systems. Current theory suggests that cannibalism generally impedes disease spread, because each victim is usually consumed by a single cannibal. Thus, cannibalism merely transfers pathogens from one individual to another without spreading infections to additional hosts. This assumes that cannibalism is the only mode of transmission and that the host population is homogenous. However, host developmental stages are a key determinant of both cannibal-victim and host-pathogen interactions. We suggest that multiple modes of pathogen transmission can interact through host stage structure. We show theoretically that cannibalism can enhance disease spread by consistently transferring infections from low quality to high quality hosts that are more infectious via horizontal transmission. We review empirical evidence for the generality of key conditions required for this process, and analyze the implications for the evolution of transmission through cannibalism. More generally, our theory promotes the consideration of multiple transmission pathways when studying parasite-host systems, and advances a useful intuition for assessing whether or not such pathways may be mutually augmentative.

Evolutionary and ecological consequences of multiscale variation in pollen receipt for seed production.

Variation in resource availability can select for traits that reduce the negative impacts of this variability on mean fitness. Such selection may be particularly potent for seed production in flowering plants, as they often experience variation in pollen receipt among individuals and among flowers within individuals. Using analytically tractable models, we examine the optimal allocations for producing ovules, attracting pollen, and maturing seeds in deterministic and stochastic pollen environments. In deterministic environments, the optimal strategy attracts sufficient pollen to fertilize every ovule and mature every zygote into a seed. Stochastic environments select for allocations proportional to the risk of seed production being limited by zygotes or seed maturation. When producing an ovule is cheap and maturing a seed is expensive, among-plant variation selects for attracting more pollen at the expense of producing fewer ovules and having fewer resources for seed maturation. Despite this increased allocation, such populations are likely to be pollen limited. In contrast, within-plant variation generally selects for an overproduction of ovules and, to a lesser extent, pollen attraction. Such populations are likely to be resource limited and exhibit low seed-to-ovule ratios. These results highlight the importance of multiscale variation in the evolution and ecology of resource allocations.

Parental optimism versus parental pessimism in plants: how common should we expect pollen limitation to be?

Many organisms exhibit parental optimism, producing more of the initial stages of offspring (e.g., eggs, embryos) than they can usually mature. For plants, parental optimism may be linked to the risk of seed production being limited by pollen receipt (pollen limitation). Here we elaborate a stochastic model of pollen limitation developed by Haig and Westoby ( 1988 ) and Burd ( 2008 ) and link it with published data on the magnitudes of prepollination costs versus postpollination costs of seed production in 80 plant species. We demonstrate that parental optimism should be expected when prepollination costs of seed production are small relative to postpollination costs. This was observed for most (62 of 80) of the plant taxa surveyed. Under parental optimism, plants overinvest in securing fertilized ovules, and consequently pollen limitation is predicted to be uncommon. However, for a sizable minority of plant species (18 of 80), prepollination costs approach or exceed postpollination costs. For these species, parental pessimism is instead optimal. Parents initiate fewer zygotes than they can usually mature, and pollen limitation is predicted to be severe. Because the relative magnitudes of prepollination and postpollination costs vary by more than 1,000-fold across plant taxa, parental outlook (optimism vs. pessimism) and levels of pollen limitation are predicted to vary widely.

A test of balanced fitness limitations theory: Pollen limitation in plants.

When reproductive success is determined by the relative availabilities of a series of essential, non-substitutable resources, the theory of balanced fitness limitations predicts that the cost of harvesting a particular resource shapes the likelihood that a shortfall of that resource will constrain realized fitness. Plant reproduction through female function offers a special opportunity to test this theory; essential resources in this context include, first, the pollen received from pollinators or abiotic vectors that is used to fertilize ovules, and, second, the resources needed to provision the developing seeds and fruit. For many plants realized reproductive success through female function can be readily quantified in the field, and one key potential constraint on fitness, pollen limitation, can be assessed experimentally by manually supplementing pollen receipt. We assembled a comparative dataset of pollen limitation using only studies that supplement pollen to all flowers produced over the plant's reproductive lifespan. Pre- and post-pollination costs were estimated using the weight of flowers and fruits and estimates of fruit set. Consistent with expectations, we find self-incompatible plants make greater pre-pollination investments and experience greater pollen limitation. However, contrary to theoretical expectations, when variation due to self-compatibility is accounted for by including self-compatibility in the statistical model as a covariate, we find no support for the prediction that plants that invest more heavily in pre-pollination costs are subject to greater pollen limitation. Strong within-species, between-population variation in the expression of pollen limitation makes the quantification of mean pollen limitation difficult. We urge plant ecologists to conduct more studies of pollen limitation using whole-plant pollen supplementation to produce a richer comparative dataset that would support a more robust test of the balanced limitations hypothesis.

Leveraging satellite observations to reveal ecological drivers of pest densities across landscapes.

Landscape ecologists have long suggested that pest abundances increase in simplified, monoculture landscapes. However, tests of this theory often fail to predict pest population sizes in real-world agricultural fields. These failures may arise not only from variation in pest ecology, but also from the widespread use of categorical land-use maps that do not adequately characterize habitat-availability for pests. We used 1163 field-year observations of Lygus hesperus (Western Tarnished Plant Bug) densities in California cotton fields to determine whether integrating remotely-sensed metrics of vegetation productivity and phenology into pest models could improve pest abundance analysis and prediction. Because L. hesperus often overwinters in non-crop vegetation, we predicted that pest abundances would peak on farms surrounded by more non-crop vegetation, especially when the non-crop vegetation is initially productive but then dries down early in the year, causing the pest to disperse into cotton fields. We found that the effect of non-crop habitat on pest densities varied across latitudes, with a positive relationship in the north and a negative one in the south. Aligning with our hypotheses, models predicted that L. hesperus densities were 35 times higher on farms surrounded by high versus low productivity non-crop vegetation (EVI area 350 vs. 50) and 2.8 times higher when dormancy occurred earlier versus later in the year (May 15 vs. June 30). Despite these strong and significant effects, we found that integrating these remote-sensing variables into land-use models only marginally improved pest density predictions in cotton compared to models with categorical land cover metrics alone. Together, our work suggests that the remote sensing variables analyzed here can advance our understanding of pest ecology, but not yet substantively increase the accuracy of pest abundance predictions.

Limits to the reproductive success of two insect parasitoid species in the field.

Debates on the relative importance of different factors in limiting the realized fitness of insect parasitoids and herbivores have continued for decades. One major reason for the duration of these debates is the paucity of empirical evidence regarding the reproductive success of minute insects under field conditions. We used a novel technique to estimate lifetime reproductive success in two Anagrus spp. parasitoids, whose hosts are eggs of leafhoppers that feed on grape leaves. Females were collected soon after they died naturally, and the number of eggs in their ovaries was counted. We used these data to estimate the lifetime oviposition success of individual females. We found that more than 10% of females from the field exhausted their entire supply of eggs before they died. The lifetime reproductive success of females was positively related to their body size and was higher at field sites with more abundant hosts, although we could not rule out a causal role for other site-specific factors. In addition, we found that females from habitats rich with hosts emerged with more eggs, suggesting that they might be adapted to local conditions. The results are consistent with theoretical predictions from models considering the risk of egg limitation.

Costs of Lygus herbivory on cotton associated with farmer decision-making: an ecoinformatics approach.

Because the farmer is typically excluded from the experimental research setting, experimental research may face challenges in evaluating pest management tactics whose costs and benefits hinge on farmer decision-making. In these cases an ecoinformatics approach, in which observational data collected from the commercial farming setting are "mined" to quantify both biological variables and farmer behavior, can complement experimentation as a useful research tool. Here I analyze such an observational data set to characterize associations between early- (June) and mid-season (July) Lygus hesperus Knight populations and farmer decisions to apply plant growth regulators and defoliants. Previous experimental work suggested the hypothesis that Lygus herbivory, by inducing abscission of young flower buds, might generate increased use of plant growth regulators and defoliants. Cotton's ability to compensate for loss of flower buds may, however, increase as plants grow. On upland cotton, June Lygus populations were associated with increased use of plant growth regulators, as expected, but this relationship was not observed for July Lygus populations. June Lygus populations were not associated with the use of defoliants, whereas, surprisingly, July Lygus populations were associated with decreases in defoliant use. In contrast to these positive and negative associations observed on upland cotton, on Pima cotton Lygus populations exhibited no associations with use of either plant growth regulators or defoliants. These results suggest that cotton responses to Lygus herbivory, as demonstrated in previously published experimental studies, can translate into economically meaningful changes in farmer decisions to apply agricultural chemicals.