Effects of floral diversification on beneficial arthropods and ecosystem services in an edamame agroecosystem

Progress report for GNE21-254

Project Type: Graduate Student
Funds awarded in 2021: $14,998.00
Projected End Date: 08/31/2023
Grant Recipient: University of Maryland
Region: Northeast
State: Maryland
Graduate Student:
Faculty Advisor:
Anahi Espindola
University of Maryland, College Park
Expand All

Project Information

Project Objectives:

This proposal seeks to understand how floral diversification practices (i.e., the addition of wildflower strips and floral intercropping) within and bordering the crop, affect (1) the diversity and density of beneficial arthropod communities (natural enemies and pollinators), within the crop and in the surrounding habitat, and (2) pest control and the reproductive output of wild plant communities and crops. Specific objectives and sub-objectives are to:

Objective 1. Quantify the effect of floral additions (i.e., wildflower strips and floral intercropping) on the abundance and diversity of beneficial arthropods in designated croplands and in neighboring wild plant communities. I hypothesize that arthropod diversity and abundance will increase in plots with higher combined flower density (intercropping and wildflower strip).

Objective 2a. Assess how floral additions affect pest control and the reproductive output of crops and non-target (i.e., non-crop) wild plants in the neighboring landscape. I hypothesize that both pest control and the reproductive output of crops and wild plants will increase with proximity to floral additions.

Objective 2b. Evaluate the spatial scale (e.g. distance) at which the floral additions affect beneficial arthropod diversity and abundance, and translate to direct ecosystem services (i.e., pest control, plant reproduction) by measuring the limits of the spillover effect of floral additions on the crop and the natural habitat. I hypothesize that increases in the diversity and abundance of arthropods is local and will decrease with increasing distance from the wildflower strips. I also hypothesize that the addition of wildflower strips and floral intercropping will lead to increased competition for ecosystem services between plants in natural and managed lands, leading to a reduction of ecosystem services in those plants that are the furthest away from the wildflower strips.


The purpose of this project is to quantify the beneficial effects of enhanced floral diversification on ecosystem services within a crop and the surrounding habitat. The loss of natural and seminatural habitats associated with expanded agricultural acreage and an associated increase in monocultural plantings has resulted in reduced floral diversity in landscapes1, which has been shown to negatively affect beneficial arthropods2. Indeed, pollinators and natural enemies of crop pests require plant-derived resources which are often limited in monoculture cropping systems and may only be present for a finite period. Two practices that may be used to enhance these resources in agricultural fields include adding wildflower strips along the crop border and intercropping flowering plants within the crop field. These practices are known to help with weed suppression, and enhance biological control and pollination activity in the crop fields3,4. Flowering plants situated around the perimeter of a crop enhance and support natural enemies and pollinators by providing them food and refuge throughout the season2. Further, intercropping (i.e., growing different plant species between cultivated crop rows) has been shown to improve soil quality, suppress weeds and attract natural enemies4. Some of these intercropped plants (e.g., clover) can also attract pollinators and natural enemies, enhancing crop pollination, pest suppression, and crop yield3,5,6.

Although increasing floral resources could restore ecosystem services within a crop, little research attention has been directed at determining how these newly-added resources can affect ecosystem services in neighboring plant habitats. For example, pollinator visitation increases with increased floral density and diversity7,8. However, it is unclear as to what extent such an increase in pollinator visitation from the floral additions along crop borders and within crop rows will spillover into target crops and native plants in the surrounding habitat9

Further, floral strips may also act as pollinator sinks, concentrating pollinators and pollination services from the surrounding habitats within the floral strip8–10. This could result in competition for pollinators between crops and wild plants, negatively impact the reproductive output of plant communities in neighboring habitats10, and indirectly affect food webs11. Ultimately, natural areas are sources of pollinators and biological control agents, and changes in resource availability and floral diversity could affect this spillover from natural to managed systems9. This, in turn, could have lasting implications on pest populations and ecosystem services in cropping systems. However, to date little research has been conducted on spillover effects in agroecosystems from managed to natural areas9.

Given the current pollinator crisis12, it is important that the effects of these practices are quantified beyond the crop-field. To date, virtually no research has been done to quantify the combined effects of using wildflower strips and intercropping to enhance natural enemy efficacy and pollination within crop fields and surrounding habitats. This project proposes to address these knowledge gaps and concomitantly develop pest suppression practices that are more congenial to pollinators and natural enemies (predators and parasitoids) within and outside the crop field, thus enhancing the sustainability of the entire agroecosystem.


Materials and methods:

Experimental Design Common to all Objectives

Experiments will be conducted over two years at four experimental sites in Maryland: (1) Beltsville Agricultural Research Center, (2) Wye Research & Education Center, (3) Western Maryland Research & Education Center in Keedysville, and (4) Central Maryland Research and Education Center in Upper Marlboro. Each site will serve as a replicate and consist of three treatments (Figure 1): (i) a mowed strip followed by four rows of wild native plants placed at increasing distances from the mowed strip, and, on the other side, a field of edamame (variety Chiba Green) planted as a monoculture (control [C]); (ii) as in (i), but with a wildflower strip in place of the mowed strip (WF); and (iii) as in (ii) with a wildflower strip, and the edamame field is inter-planted with red clover, Freedom cultivar (Trifolium pratense) (WF+CL). Each treatment plot will cover an area of 92m X 15m, spaced within the farm by at least 500m to maintain treatment integrity and reduce arthropod flight overlap. Treatments will include a 15m X 15m plot of 16 edamame rows with an inter-row cropping of 76cm, one 3.5m X 15m strip (mowed [C] or seeded with wildflower mix [WF and WF+CL]) on one end of the edamame plot, and four groups of nine native wild plants placed at distances of 10m (group 1), 30m (group 2), 50m (group 3), and 70m (group 4) from the edge of the flower or mowed strip (Figure 1). Placement of a single wildflower strip between the edamame plot and wild plant community was meticulously done so as to concomitantly monitor the wildflower strip’s influence on beneficial arthropods within the edamame plot and neighboring wild plant community, and how it varies with distance.

The wildflower seed mix used for establishing wildflower strips is a consortium of 17 local provenanced flowering plants that are mostly native to the Mid-Atlantic region (Maryland Upland Mix + seven annual natives; Ernst Seeds, Inc.). The mix consists of annuals and perennials that will bloom continuously throughout the season (April-October) and have varying morphological features to ensure greater access to nectar in flowers by diverse arthropods, including pollinators and natural enemies. Deer fencing will be installed at each site to protect the plants from damage incurred by the abundant deer populations in the vicinity of the fields.

The nine wild plant species were selected based on their native status in Maryland, and their likely presence in the habitats close to the fields. They consist of annuals and perennials that bloom in tandem with the wildflower strips, and display a variety of floral traits to attract both generalist and specialist pollinators (Table 1). The plants will be purchased from Chesapeake Natives, Inc., and Earth Sangha, and repotted in 2-gallon pots to maintain soil uniformity among sites and between years. Four pots per species will be placed at each distances of the flower strips and sunk into the ground to prevent desiccation.

Figure 1. Summary of treatments for each site (n=4). Treatments will be at a minimum of 500m apart. Four groups of potted wild plants will be placed at discrete distances (10m, 30m, 50m, 70m) from the mowed/wildflower strip in each treatment. Treatment-1 (control) is a monoculture of edamame with no wildflower strip or floral intercropping; Treatment-2 is a monoculture of edamame with a wildflower strip but no floral intercropping; Treatment-3 consists of a field of edamame intercropped with red clover and a wildflower strip.

Table 1. List of native wild plants (“Wild Plants” in Figure 1) and some of their relevant traits (flower color, bloom time and flower shape).

Objective 1. Quantify the effect of floral additions (i.e., wildflower strips and floral intercropping) on the abundance and diversity of beneficial arthropods in designated croplands and neighboring wild plant communities.

For this objective, I will measure the abundance and diversity of two groups of beneficial arthropods: natural enemies and pollinators. To monitor aerial natural enemies, I will deploy yellow sticky cards (7.6 x 12.7cm, Olson Products, Inc.) every 3.75m within the edamame plot, in the center of the wildflower/mowed strip (~ 226cm from the outermost edamame row), and in each group of wild plants (groups 1-4). Cards will be placed at approximately 2/3 the height of edamame plants and set up at the beginning, middle and end of each growing season for seven-day intervals. Upon collection, the cards will be placed in plastic zipper bags and stored in the freezer for later identification.

To document pollinator visitation and measure spillover from the wildflower strips into the wild plants and edamame, I will conduct visual observations and collections of flower visitors biweekly from April to October (onset to end of wildflowers and wild plants flowering). Observers will perform 20-minute timed observations and record pollinator type (e.g., butterfly, bee, fly), specific group (e.g., halictid bees, honeybees, syrphids), and visited plant species7,8. After each 20-minute timed observation, observers will collect all insect visitors seen interacting with the flowers for a 15-minute time interval. After collection, specimens will be individually stored in 90% ethanol for later species-level identification. Each timed observation and collection will be conducted at (i) standardized transects through randomly selected rows of red clover, (ii) wildflower strips, and (iii) wild plant groups 1-4. To measure pollinator spillover from the floral additions when edamame is in bloom (June-July), additional observations and collections will be conducted from the edamame flowers, at distances of every other row. Only pollinators observed in flowers and interacting with the reproductive structures will be recorded or collected to prevent counting non-pollinators. Surveys will be conducted biweekly by trained individuals, replicated in the mornings and afternoons, and with the order of survey randomized for each sampling date.

To quantify the presence and abundance of different pollinator species between treatment plots, I will place bee bowls within the edamame plot, border strip of each treatment, and at each wild plant group. Bee bowls are small plastic cups painted in three standardized colors (fluorescent-blue, fluorescent-yellow, and white) and filled with soapy water to attract and trap suites of pollinators22. Replicate bee bowls of each color will be placed within the wildflower strips (WF, WF+CL) or mowed rows (C), every 3.75m within the edamame plot and at each wild plant group in the morning and collected after 24 hours, biweekly from April to October. After collection, the samples will be stored in 90% ethanol for later processing. This sampling will not overlap with the other pollinator samplings. I will conduct biweekly measures of floral abundance and diversity8 in the wildflower strips and wild plant groups throughout the flowering season to assess the relative amount of supplemental food resources available to arthropods.

Analyses: I will use an ANOVA to compare arthropod species abundance and richness across treatments, distance from the border strip (10m, 30m, 50m, 70m), wild plant species, sites and years. These measures will inform me of how the floral additions affect the composition and abundance of pollinator visitors and natural enemies within a targeted crop and in wild plants in the neighboring habitats.

Objective 2a. Assess how floral additions affect pest control and the reproductive output of crops and non-target (i.e., non-crop) wild plants in the neighboring landscape.

To quantify the effects of wildflower strips and floral intercropping on pest control, I will monitor stink bug (Pentatomidae) egg-mortality due to predation and parasitism. In the Mid-Atlantic, stink bugs are economically devastating to edamame, with reported damage levels of 60%23. To quantify egg-mortality, eggs will be freeze-killed and distributed per treatment using methods similar to Herlihy et. al. (24). Freeze-killed eggs will not hatch, but are susceptible to parasitism by wasps and can be consumed by predators24. Each month, I will manually distribute sheets of paper with egg-masses (from NJ Department of Agriculture) to (1) select edamame plants between planting-date and harvest, and (2) wild plants in each treatment between April and October. These egg-masses will be distributed regularly along the fields (2 sheets/4 rows) and the rows of select wild plants (2 sheets/row). Eggs will be checked weekly to determine their fate. Eggs will be classified as (1) missing, in which eggs disappear from the surface of the leaf; (2) parasitism, in which eggs are parasitized, and (3) predator, in which eggs appear shrunken, collapsed, or chewed.

To quantify the effects of floral diversity on plant reproductive output I will measure the reproductive success of edamame and the nine wild plants (Table 1) planted at discrete distances from the wildflower/mowed strip (10 meters, 30 meters, 50 meters, 70 meters) in each treatment (C, WF, WF+CL) (Figure 1).

To measure the reproductive output of edamame, I will compare crop yield between treatments. When pods are ready for harvest (between stages R6 and R7), three groups of 15 plants each will be selected to harvest. One group will encompass the outermost plants at the farthest distance from the wildflower strip, the second group will encompass the innermost plants (rows 7-9), and the third will encompass the plants that are closest to the wildflower strips (rows 1-3). Pods will be removed by hand, weighed to estimate yield, and the number of pods, empty pods, seeds per pod, and pod weight per plant will be recorded25. In addition, pods will be collected to determine insect damage and marketable yield.

To measure the reproductive success of wild plants, I will randomly mark up to 10 flowers per individual plant (four individual plants per species) at each distance and treatment. After the plant has ceased blooming, I will collect formed fruits, and count and weigh the number of seeds produced per plant. To establish a baseline to compare seed-set with, for each wild plant species I will bag up to 15 flowers per species and site. Following standard practice in plant reproductive tests, bags will be constructed from mosquito netting, placed over the flower at bud stage, and remain in place until seed and fruit collections26.

Analyses: To evaluate effects on pest-control, I will use a binomial model with a logit-link, and compare stink bug egg-fate between treatments. With these measures, I can evaluate if the floral additions enhance biological control in edamame. To evaluate the effect on plant reproductive outputs, I will do a number of analyses. For each wild plant species (192 plants per species in total), I will measure reproductive output through (i) total seed count and (ii) total seed weight per plant. Using an ANOVA, I will compare reproductive output among species, distances from the wildflower strip, treatments, sites, and years. This will inform me if the floral additions affect the reproductive output of wild plants in the surrounding landscape. For edamame, I will use an ANOVA to assess the relationship between wildflower strips, floral intercropping, and yield (pod size, seeds per pod, pod weight, pod damage). With this information, I will infer the benefits of floral diversification on crop yield and quality.

Objective 2b. Evaluate the spatial scale at which floral additions affect beneficial arthropod diversity and activity and translates to direct ecosystem services.

To evaluate the distance and spillover effect of wildflower strips and floral intercropping, I will follow a similar procedure as for Objective 2a. I will compare arthropod community composition, pest control (e.g. egg-fate), and reproductive output of wild plants in the four groups in each treatment (Figure 1). I will also compare arthropod community composition, pest control and yield within the edamame crop across three distances from the wildflower[WF, WF+CL]/mowed[C] border strip: the innermost rows adjacent to the wildflower/mowed[C] border (row 1-3), the center rows (row 7-9) and the outermost rows farthest from the wildflower/mowed[C] border (row 15-16)

Analyses: I will use ANOVAs to compare arthropod community composition, pest control and reproductive output among treatments, distances from the wildflower border strip, sites, and years, for both wild plants and edamame. These measures will allow me to quantify the spillover effect and its limits, by measuring how the floral additions affect the beneficial arthropod abundance and diversity, pest control, and reproductive outputs of wild plants and edamame.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary:

Education/outreach description:

Clientele who will benefit from this project include researchers and agricultural practitioners, such as organic and conventional farmers, extension specialists, crop advisors, land managers, Master Gardeners, and 4-H personnel. Farmers will be presented information that will allow them to adopt practices for conserving populations of natural enemies and pollinators on their farms. This will contribute to the sustainable management of insect pests as well as increased yields and profits. Although edamame is being used as the test crop, results and techniques being investigated are applicable to other cropping systems in the Mid-Atlantic. Thus, at outreach events, attendees will be made aware that the concept is transferable to other cropping systems. Project recommendations will be incorporated into commercial vegetable production guides and results, and recommendations will be presented at local (e.g., Maryland Vegetable and Fruit meetings) and regional (e.g., Mid-Atlantic Fruit and Vegetable Convention) commodity conferences attended by farmers, agricultural agents, crop advisors and other stakeholders. Additionally, project information will be presented at commercial farm walking tours and field day events held at the study site so that attendees can see the research site. Once the research concept is developed, a medium range output is to establish demonstration plots on commercial farms which will allow farmer participants to contribute to information dissemination during commercial farm walking tours and field day events.

Extension publications will also be used as an outlet to disseminate information to stakeholders. Publications will include traditional and web-based platforms, such as farmers newsletters (e.g. UME Vegetable and Fruit News), posting on UMD extension web site and writing multilingual blogs (e.g., UME HGIC Maryland Grows; UMD Extensión en Español). The Espíndola Lab has strong connections with Maryland 4-H, with whom it collaborates in the development of pollination educational packets (lesson plans and kits) used in state-wide summer camps (~2,500 campers), targeting youth ages 8-18. Materials produced are used by camp directors and instructors to educate students about pollination. Thus, a lesson plan and experimental tool-kit based on this project will be created to educate youth about pollinator observation, identification and conservation, and demonstrate connections between floral diversity, pollination, natural pest suppression and food production.

To reach a scientific audience, results will be published in scientific reports, research publications and conference proceedings and presented at regional and national scientific conferences, including the Entomological Society of America National and Eastern Branch conferences.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.