Recipient Organization
OHIO STATE UNIVERSITY
1680 MADISON AVENUE
WOOSTER,OH 44691
Performing Department
(N/A)
Non Technical Summary
Insecticide use is ubiquitous in many pollinator-dependent crops. Despite the well-known negative effects of insecticides on long-term pollinator (mostly honeybee colony) health, the near-term impact of insecticides on crop pollination and marketable yield is surprisingly unknown. Most studies evaluating pollination services and crop production tend to focus on ecological metrics of crop yield, including fruit set and seed number, rather than those governing crop marketability. Thus, to maximize the role of bees as crop pollinators, increase agronomic profitability, and protect pollinator health, we need a better understanding of how commonly used, broad-spectrum insecticides affect pollinator foraging behavior and marketable yield. The overall goal of this project is to devise novel management tools that maintain optimal levels of crop pollination without sacrificing pest control. Using the unique spatial configuration of the pollenizer system in self-incompatible crops, we will perform novel experiments on research and commercial farms growing seedless watermelon, a highly pollinator-dependent crop. Specifically, we will: 1) Determine the direct effects of insecticides on pollinator foraging behavior and marketable yield; 2) Evaluate the indirect effects of insecticide inputs on field-scale crop yield by reducing the need for pollenizers; and 3) Assess the combined interactive effects of insecticide programming and pollenizer ratios on commercial watermelon production. This project's goals and objectives closely align with the Pollinator Health program area priorities: "factors that influence the abundance, diversity and health of pollinators" and "development and evaluation of innovative tools and management practices that would likely be adopted by stakeholders to ensure healthy pollinators".
Animal Health Component
70%
Research Effort Categories
Basic
30%
Applied
70%
Developmental
0%
Goals / Objectives
The overall goal of this project is to devise novel decision tools that maintain optimal levels of crop pollination without sacrificing pest management. Specifically, we will: Objective 1: Determine the direct effects of insecticides on pollinator foraging behavior and marketable yield, Objective 2: Evaluate the indirect effects of insecticides on field-scale crop yield by reducing the need for pollenizers and Objective 3: Assess the combined (direct + indirect) interactive effects of insecticide programming and pollenizer ratios on commercial watermelon production
Project Methods
Objective 1: Determine the direct effects of insecticides on pollinator foraging behavior and marketable yield Experimental fields will be established at four total research farms at Purdue ('SWPAC', and 'TPAC') an Ohio State University (Mellinger Research Farm, Western Agricultural Research Station). Fields will be chosen based on optimal watermelon production Figure 7: Predicted distribution of seedless watermelon in two fields with plants either untreated or treated with insecticide. Tetraploid watermelons will be transplanted into the field using black plastic, drip irrigation and standard fertilizer practices. Tetraploid var. 'Liberty' will be used due to its popularity among growers, but susceptibility to hollow heart (Johnson 2011). Pollenizer var. 'SP7' will be planted in a single row on the western edge of the field. Large fields will be used (100m row length × 50m width = ~1.3 acres with 25 rows) for treatments. Each research farm (n=4) will represent one replicate of a factorial design crossing insecticide use-- untreated control and insecticide-treatment --and pollinator supplementation. Due to the high foraging distance of many pollinators (Greenleaf et al. 2007; Osborne et al. 2007; Hagler et al. 2011), paired treatment fields will be separated by at least one mile to avoid sampling from the same communities. In addition to varying insecticide use, we will also implement two pollinator supplementation treatments (with and without stocked honeybees). Data will be analyzed using generalized linear mixed models (R package 'lme4'; Bates et al. 2015). For the short-term analysis, data will be aggregated across all sites. Insecticide treatment, honeybee supplementation, distance from pollenizer, time points and their interactions will be treated as fixed effects, and site and sample (=female flower) nested within site as random effects. Response variables-- watermelon yield, hollow heart severity, incidence of fruit deformity, and pollen deposition--will be fit to appropriate distributions and treatments in each analysis compared using least squared means (P<0.05; 'emmeans', Lenth et al. 2016). For the season-long data, we will analyze the effect of insecticide treatment and distance and all interactions as fixed effects, and site and sample (=quadrant) nested within site as random effects on pollinator abundance and diversity (commercial vs. wild bees), and duration of visitation. Models will be checked for appropriateness and fit (R packages: blmeco and Dharma; Hartig 2017; Korner-Nievergelt et al. 2019)Objective 2: Evaluate the indirect effect of insecticides on field-scale crop yield by reducing the need for pollenizers Fields will be selected in the same manner as described in Obj. 1. Experimental fields will represent a factorial design in which insecticide use (=2 levels) and pollenizer ratio (=3 levels) are varied (Fig. 9). Pollenizer and tetraploid watermelons will be planted in accordance with specific pollenizer:tetraploid ratios, including the industry standard (1:3) and two experimental reductions (1:6 and 1:12). Although pollenizer ratios vary, all fields will be the same size (100 × 100m). Similar to Obj. 1, precautions will be taken to spatially isolate treated and untreated fields such that pollinator observations and samples are reflective of the treatments imposed. Commercial pollinators will not be supplemented in this trial due to their abundance potentially underestimating the effect of insecticide. Further, we believe it would be redundant to include supplementation as a treatment when it has been prioritized in Obj. 1. Pollinators will be randomly sampled in every other row. In each row, we will randomly sample quadrant areas (2 × 2m). The process will be repeated in four replications for a total of at least 12 samples per treatment/research farm. Fields with lower pollenizer ratios (e.g., 1:12) will have a higher number of sampling points due to the greater number of rows between pollenizers. Otherwise, pollinators will be monitored as described in Obj. 1. At the end of observations, bees (n=100) will be collected at midpoints within the fields (centermost tetraploid row between pollenizer plants) for pollen and pesticide residue analysis Figure 9: One replicate of a factorial evaluating the effect of pollenizer ratio (1:3, 1:6, 1:12) and insecticide treatment (untreated or treated) on pollination and marketable yield of seedless watermelon. To determine the proportion of bees carrying diploid pollen (donor pollen from pollenizer plants), pollen will be extracted from bees (using protocols described in Potter et al. 2019) and then tested for ploidy status using modified RT-PCR methods from Wang et al. (2009). Extracted pollen will remain individuated based on bee to represent 20 total samples. Diagnostic primers will be used to determine watermelon ploidy per previous report (Wang et al. 2009). From these data, we can determine the percent of bees carrying diploid pollen within the different treatments (N bees carrying diploid pollen/20 total bees). Data will be analyzed using generalized linear mixed models (R package 'lme4' Bates et al. 2015). Insecticide treatment, pollenizer ratio and their interactions will be treated as fixed effects, and site as a random effect. Pollinator abundance, diversity (wild vs. managed), and duration of visitation will be fit with appropriate distributions and examination of overdispersion and distribution of residuals (R packages: blmeco and Dharma; Hartig 2017, Korner-Nievergelt et al. 2019). Yield will be analyzed as described in Obj.1, except we will separately test field-scale and per-plant yields. Pairwise comparison will be made between treatments ('emmeans', Lenth et al. 2016).Objective 3: Assess the combined (direct + indirect) interactive effects of insecticide programming and pollenizer ratios on commercial watermelon productionWe will work with 5 growers (10 total fields) in Southwestern Indiana, a region with multiple melon farms that cultivate large (50-100 acre) fields devoted entirely to watermelon production, At each field we will establish a factorial experiment to evaluate the impact of two insecticide programs (standard vs. action threshold-based) and two pollenizer ratios (1:3 'grower standard' vs. 1:6). Growers will pre-select two fields (10-25 acres acre each); one managed using an IPPM-based insecticide program and the other with a standard (weekly) insecticide program. Each field will be bisected to feature at least four replications of each pollenizer frequency, either 1:3 or 1:6 (refer to Fig. 9). We will not alter other practices aside from insecticide application to achieve the most realistic assessment of the proposed treatments. At the end of each growing season, all growers will supply pesticide application records for experimental fields (i.e., products, rates, dates of application). All five participating growers will complete a survey at the end of each year to describe their experience implementing the insecticide × pollenizer ratio treatments and their assessment of the viability of the treatments in commercial production. From pesticide application records we will determine the average cost spent by the grower using either their standard (weekly) or the action threshold-based insecticide program. We will contact local agrichemical companies to determine regional insecticide costs. In calculating these costs, we will leverage previous data gathered by the lab (Ternest et al. 2020) to estimate this cost relative to the expense of scouting. We will also estimate marketable yield (tons/ac) from the harvested area to calculate the relative value of watermelons from the treatments using current market prices for seedless watermelon.