A Landscape Ecological Perspective in Insect and Weed Population Regulation in Low-Input and Conventional Systems

Final Report for LNC91-039

Project Type: Research and Education
Funds awarded in 1991: $75,000.00
Projected End Date: 12/31/1993
Matching Non-Federal Funds: $40,394.00
Region: North Central
State: Michigan
Project Coordinator:
Dr. Douglas Landis
Michigan State University
Expand All

Project Information


[Note to online version: The report for this project includes figures and appendices that could not be included here. The regional SARE office will mail a hard copy of the entire report at your request. Just contact North Central SARE at (402) 472-7081 or [email protected].]

Studies were completed to determine the impact of agricultural landscape structure on the biological regulation of pest insects and weeds. Overall goals were to determine the impact that factors such as field size, shape, type of crop, field border composition, distance to alternative habitats etc. have on the biological control of crop pests. These studies tested the hypothesis that landscape structure plays a pivotal role in determining the effectiveness of many species of predators and parasites in agricultural systems by fostering or inhibiting their populations over time.

An analysis of landscape structure conducted on two areas of Ingham Co. MI. in 1992 quantified significant differences in physical structure between an area of high structural complexity vs. an identically sized area two miles to the north of low structural complexity. The area of high heterogeneity had significantly more and smaller fields with smaller perimeters and shorter distances to an edge from field center. Crop-hedgerow edges were significantly more abundant in the high heterogeneity area while crop-crop interfaces were more abundant in the low heterogeneity area.

Previous studies by Landis et al. (1992) have shown that wooded field edges benefit Eriborus terebrans, currently the most important parasite of the European corn borer in the Midwest. In the current study it was found that Eriborus wasps require both a source of adult food (sugar) and a moderated microclimate for optimum survival. Corn fields do not provide these resources (during the first generation of the wasp) and wasps are forced to seek these resources in alternate habitats. Our studies have shown that nectar from flowering plants or the secretions of aphids feeding on weeds in field-edge habitats provide a suitable source of adult food. Fencerows and woods surrounding corn fields also provide a cooler, more humid microclimate than a corn field and wasps in these habitats live significantly longer. These studies demonstrate that landscape structure can play an important role in determining the effectiveness of natural enemies.
Studies of weed seed predation showed that during the winter of 1991-92 showed that, vertebrates removed 6-12% of the seeds on the soil surface of crop fields in a six day period. In the spring, insects and vertebrates removed 48.5% of the seeds within 5 meters of hedgerow and 35% at 100 meters from a hedgerow in a six week period. Carabid beetles were the most abundant insect seed predator and rodents and birds the primary vertebrate seed feeders. In 1992-93, up to 40% of large weed seeds were removed over the winter and up to 66% of the small weed seeds in the spring.


Research results and discussion:
Methods and Results

Objective 1: Characterize Landscape Structure of an Established Low-Input Dairy/Crop farm to Nearby Conventional Farms (see also Appendix I).

(This objective was completed in 1992 and reported on in the 1992 annual report. A synopsis of results is included here for continuity).

Findings: Striking differences were found between the low and high heterogeneity areas (Appendix I, p. A1-3). In the area of high heterogeneity, on average, fields were significantly smaller (3.4 vs 12.4 hectares) but more numerous (139 vs 61/ four mi sq.) when compared to the area of low heterogeneity. Field perimeter was also less (775 vs. 1637 m) in the area of high heterogeneity. One general hypothesis of this study is that certain beneficial organisms such as insect predators and parasites as well as birds and rodents may be influenced by the type of edge habitat surrounding crop fields. If they move from field edges into fields, the distance from an edge to the field center may determine if they can colonize the entire field. The average distance from an edge to the center of fields in the low heterogeneity area was one third greater (101 vs 63 m) than in the high heterogeneity area. The number of edges per field, number of edge types per field and average area to perimeter ratio were not significantly different among the areas.

When the composition of the edge types in each area were compared, three aspects were different. Fields in the area of high heterogeneity had significantly more meters of field perimeter bordered by wide hedgerow and significantly less bordered by herbaceous roadsides and other crops. Crop to crop interfaces (ex. corn planted next to soybean with no fencerow or other separation) were by far the most common edge type in the area of low heterogeneity. These types of interfaces are generally felt to be of little value to beneficial insects and vertebrates.

This study demonstrates that structural characteristics of the agricultural landscape may change dramatically over relatively short distances. We did not see a difference between the structural composition of the Fogg Farm in comparison to conventional farms in the same four square mile area. This would indicate that structural differences between their farm and those to the north are due primarily to the landforms (topography, soils etc) in the immediate area and from historical landuse decisions not to Fogg's particular landuse patterns. The retention of hedgerows in the high heterogeneity area is one of the most striking aspects of the landscape structure. A high percentage of fields and roadsides are bordered by this type of edge. To the extent that hedgerows may serve as a reservoir for beneficial organisms, one may expect to see an influence of these landscape features on the pest complex and population levels in this area. The other objectives of this project are aimed at determining if these structural characteristics may influence the biotic regulation of certain insects and weeds in these systems.

Objective 2. Contrast Insect Population Regulation in Low-Imput and Conventional Cropping Systems in Relation to Farm Landscape Structure.

Results of Objective 1 (see above) indicate that prevalence of hedgerow, herbaceous roadside and crop-crop interfaces differ significantly even within a localized agricultural area. This objective determined the impact of these differences on European corn borer (ECB) population regulation.

Four corn fields in Ingham Co, MI. were selected for the study. Each field was characterized by having one wooded edge and one herbaceous roadside edge. Fields were monitored weekly throughout the growing season. ECB populations were monitored at three sample points along each edge and at three points in the field interior. Adult parasites were collected in Malaise traps which collect flying insects at pre-determined heights. Malaise trap captures were examined from each of the two edge types, the field center and the interior of the wooded edge. Previous experience had indicated that Eriborus terebrans, the most important parasite of ECB in Michigan, was likely to be more abundant in the vicinity of wooded edges. We examined two potential reasons for this observed difference. First, we examined the hypothesis that adult wasps may find food in the form of plant pollen, nectar or aphid honeydew more abundant at wooded edges. To determine what adult food sources may be important, we reared wasps on various potential food sources and determined their average lifespan. Additionally, we examined the hypothesis that the microclimate in corn fields may not be as suitable to the wasp as neighboring non-crop habitats. We measured temperature in corn fields and nearby herbaceous, fencerow and wooded edges. Adult wasps were caged in all habitats with and without food and water to determine their survival under various conditions.

Findings: During the first generation of the European corn borer, Eriborus are more abundant (Appendix V, p. A13) and parasitize more corn borers in proximity to wooded field edges vs. field interiors or herbaceous field edges. There are no differences in abundance or parasitism in the second generation. We hypothesize that lack of an adult food source (sugar from plant nectar or aphid honeydew) and higher temperatures in corn fields is reason for these differences. Wasps provided water alone survived for an average of 1.8 days, while those provided with access to flowering wild carrot (Daucus carrota) or the secretions of aphids (Aphis fabae) feeding on lambsquarter (Chenopodium album) live an average of 13.2 days (Appendix V, p. A14-15). Other flowering plants were not as suitable for wasp survival although all six species tested significantly improved the lifespan of the wasps.

In field trials, wasps caged in herbaceous habitats and in corn fields had the shortest longievity (Appendix V. p. A16-17). Those in wooded fencerows or woods lived significantly longer. There were no differences in temperature between herbacous and corn field habitats, while the fencerow was an average of 2.5 degrees C cooler and the woodlot 3.1 degrees C cooler than the corn field.

Objective 3. Characterize Weed Population Regulation and Distribution in Relation to Landscape Structure (see Appendices II - IV).

Our studies towards this objective fall into the three general categories of :

A. Vertebrate seed predation in relation to field type and residue cover (LTER experiment).
B. Vertebrate seed predation in relation to field edge type and distance to edge.
C. Post-dispersal seed predation by insects and vertebrates in relation to landscape features.

Objectives 3A and 3B were reported on in detail in the 1992 annual report and are briefly sumarized below. Objective 3C is reported on in more detail.

A. Vertebrate seed predation in relation to field type and residue cover. (Appendix II. p. A4-5)

An experiment was carried out on the Kellogg Biological station Long Term Ecological Research (LTER) to examine winter-time removal of weed and crop seeds by vertebrates from large-plot (1 hectare) field sites under different cropping practices. Corn, velvet leaf and lambs quarters seeds were placed into one of three cropping systems: ridge-till planted to wheat (36% ground cover), no-till corn stubble ( 78% ground cover) and native successional (100% ground cover). Twenty-five of each seed type were placed on sand in petri dishes at four locations within each plot. Plots were replicated four times. After six days the seeds were recovered and counted to determine removal (predation).

Findings: Over all treatments, thirty six percent of all seeds were removed in six days. Average seed removal was significantly greater in the native succession (86.9%) than in either of the agricultural plots. Seed removal in the no-till (6.7%) and the ridge-till ( 12.1%) did not differ significantly. Differences in seed removal between agricultural and successional plots were consistent with rodent trapping data (significantly more rodents in successional plots), however, birds were also observed in all treatments and are known to be seed predators. The comparatively low predation in the agricultural treatments is thought to be primarily due to the lack of crop residue cover above the ground. Lack of vertical cover makes birds and especially rodents susceptable to predation. While rodents did take seeds from even the centers of agricultural plots as evidenced by tracks and feces, their populations were higher in the successional plots where cover was greater. The practical implications of this study are that where cover is present, vertebrates can remove a high percentage of the seeds present on the soil surface.

B. Vertebrate seed predation in relation to field edge type and distance to edge. (Appendix III)
A similar study was conducted during the winter of 1992 in 24 fields in Ingham Co. MI. Two crop types, soybeans and corn and three edge types per crop, wooded, hedgerow and crop-crop interfaces were studied. In each field three transects were established running perpendicular into the field from the edge type selected. Seed plates were placed at 0, 5, 40 and 100 meters from the edge. Velvetleaf, lambsquarters and corn seeds were placed in individual dishes at each location and replaced periodically (when corn was depleted) throughout the winter.

Findings: Vertebrates, rodents and birds were the primary seed predators observed in the study. Of the three seed species offered, corn seeds were the most highly prefered. It appeared that in some fields animals were learning the location of the seed plates and returning regularly to remove primarily the corn. This resulted in questionable data in regards the other weed species offered. Stations may have had more weed seeds taken because vertebrates were "drawn" to the area. Alternatively, stations may have had fewer weed seed taken because a more prefered seed (corn) was present in the same area. On average, between 20 - 50% of the weed seeds and 50 - 90% of the corn seeds were removed from the seed plates during the exposure period (Appendix III, p. A6-7). There were no obvious effects of weed species or distance from the edge on number of seeds removed. Crop type seemed to have a slight effect on weed seed removal with generally more weed seeds removed form corn vs soybean fields. More seeds were removed from white sand than from brown sand indicating that there is a visual component to seed location (Appendix III, p. A8). Future experiments will utilize a sand color that is similar to the soil color in the field and will not utilize corn as an indicator seed.

C. Post-dispersal seed predation by insects and vertebrates in relation to landscape features. (Appendix IV)
The influence of distance from hedgerows on seed predation by vertebrates and invertebrates in spring following tillage and overwinter was examined in corn fields. We concentrated on chisel plowed fields (4 chisel plowed and 1 no-till field) to ensure that results are minimally biased by tillage practices and because chisel plowing is a common tillage practice in central Michigan. Hedgerows are being used because they are increasingly being removed from agroecosystems and we wish to know what effect, if any, their removal may have on pest populations. Experiments were conducted both after spring tillage and after fall harvest because: 1) tillage in the spring turns up seeds that will develop into weed population for the current season, and 2) seed predation by vertebrates is likely to be most intense over winter.


Seed predation after spring tillage: Four conventional till and one no till corn field having a hedgerow on the south, east or west side were selected on the Diehl farm, Dansville, MI. In each field, three transects running perpendicular from the hedgerow were set out. Within-field transects were separated from one another by 25 m and approximately centered along the hedgerow. To ensure independence, no field (or hedgerow) had more than one set of transects running into it (or from it). At 5m and 100m along each transect (6 sites per field), an invertebrate and vertebrate exclusion plot, invertebrate exclusion plot, vertebrate exclusion plot, and control plot (no exclusion) were set out. Vertebrates were excluded from plots with 50 cm x 50 cm x 12 cm (length x width x height) .25 inch hardware cloth cages. The cages were sunk ca. 5 cm into the soil. Invertebrates were excluded with 60 cm diameter circular exclosures of plastic garden edging painted with fluon (a teflon-like material that prohibits insects from crawling over vertical surfaces). The garden edging was also sunk ca. 5 cm into the soil. Both vertebrates and invertebrates were excluded by surrounding vertebrate exclusion cages with the circular invertebrates exclosures.

In the center of each plot, a 32 x 20 cm half-flat filled with steam sterilized soil was sunk into the soil so that 1 cm of the edge of the pot edge was raised above the soil surface. On the surface of the sterilized soil 25 seeds of velvet leaf (Abutilon theophrasti), lamb's quarters (Chenopodium album), pigweed (Amaranthus retroflexus), fall panicum (Panicum dichotomiflorum) and yellow foxtail (Setaria lutescens) were scattered.

To examine post-dispersal seed predation in the spring after tillage, all cages and exclosures were placed into the fields within 1 week of planting and left for 3-4 weeks. Flats containing soil and uneaten seeds were removed from the fields and labeled according to field, location within a field (5 m vs. 100 m from hedgerows) and exclusion treatment (vertebrate, invertebrate, both or neither). They were then placed in a greenhouse on the MSU campus and kept well watered to facilitate seed germination. In June and July numbers and species of germinating weeds in each flat were recorded 3 times at ca. 2 week intervals (each seedling was removed after being recorded).

After 6 weeks, all flats were moved from the greenhouse to a cold chamber (4 degrees C) for 1 month to break dormancy in the remaining ungerminated seeds. Flats were then returned to the greenhouse, the top 1 cm of soil on each flat turned over to enhance germination and the number and species of any subsequently germinating seedlings recorded.

During the course of the spring experiment 18 pit-fall traps were set out at 5 m (9 traps) and 100 m (9 traps) in the same fields are the exclosure study. Pit-fall traps were set out along the same hedgerows, but 100 m away from the exclosure study. The pit-fall trapping provide information on the abundance and diversity of potential seed eating invertebreates (mainly carabid beetles).

Over-winter seed predation: To examine overwinter seed predation the same corn fields described above were used (except one conventional till field that was not harvested due to the wet fall) and the same sites within each field. In December, at each site, 4 flats filled with sterilized soil (sterilized with methyl bromide) were sunk into the soil as described above. Two of the flats (randomly chosen) were covered with vertebrate exclusion cages and two were left uncovered. Twenty-five seeds of the same five species were scattered onto the surface of the soil of each flat. Flats containing soil and uneaten seeds were removed from the fields in mid-April just prior to plowing and labeled according to field, location within a field (5 m vs. 100 m from hedgerows) and treatment (vertebrate exclusion vs. no exclusion). They were then placed in a greenhouse on the MSU campus and kept well watered to facilitate seed germination. In May, June and July numbers and species of germinating weeds in each flat were recorded at ca. 2 week intervals (each seedling was removed after being recorded). After each census, the top 1 cm of soil on each flat was turned over to enhance germination and the number and species of any subsequently germinating seedlings recorded.


Seed predation after spring tillage: Results from the split-plot analysis of each weed species separately indicated significant treatment (no exclosure vs. invertebrate exclosure vs. invertebrate + vertebrate exclosure) effects for A. retroflexus (F3 = 11.04; P = 0.001) and C. album (F3 = 2.81; P = 0.0452) but not for the larger seeded P. dichotomiflorum or A. theophrasti ( Appendix IV, p. A10). None of the S. lutescens seeds germinated. There were no significant position (field edge vs. field interior) effects. Significant field effects were found for C. album (F4 = 3.64; P = 0.0093) and A. theophrasti (F4 = 4.46; P = 0.0028) because germination differed among fields.

We expected germination rates to be highest in the invertebrate + vertebrate exclusion treatment and this prediction was confirmed for A. retroflexus and C. album. We also expected germination rates to be highest at the edge vs. the interior and this was never the case for any of the species.

We have not yet identified Carabid beetles to species, however, there are two results that stand out from our initial survey of the capture data: 1) Carabid densities vary considerably between fields, and 2) Carabid densities were as high 100 m from hedgerows as they were 5 m from hedgerows (Appendix IV, p. A11). These results suggest that seed predation by Carabids may be evenly spread across fields and may very considerably between fields.

Over-winter seed predation: Results from the split-split plot analysis of each weed species separately indicated significant treatment (no exclosure vs. vertebrate exclosure) effects for the larger seeded species S. lutescens (F1 = 5.51; P = 0.0022) and A. theophrasti (F1 = 13.83; P = 0.004). There was borderline significance for the species of intermediate seed size (P. dichotomiflorum; F1 = 3.98; P = 0.0506) and no treatment effect for the small seeded species A. retroflexus and C. album (Appendix IV, p. A12). There were no significant position effects (field edge vs. field interior). Significant field effects were found for A. retroflexus (F3 = 4.08; P = 0.0104), P. dichotomiflorum (F3 = 9.55; P = 0.00011) and S. lutescens (F3 = 4.73; P = 0.0049) because germination differed among fields.

As expected germination rates were highest in the vertebrate exclusion treatment, however, only for the larger seeded species. This suggests that vertebrates seed predators during the winter months prefer larger seeds to small seeds. This result was the reverse of the spring post-tillage result. We also expected germination rates to be highest at the edge vs. the interior and this was never the case for any of the species. A possible explanation here is that different species of vertebrate seeds predators prefer to forage in the open (e.g., horned larks) whereas others (e.g., sparrows and small mammals) prefer to forage near cover. These different foraging strategies may simply cancel each other out in terms of seed loss from field edge vs. field interior.

Research conclusions:

Findings on individual objectives point towards the potential for increased usage of natural population regulation (biological control) of insects and weeds through an understanding of the impact of landscape structure on these processes. This would result in a decreased need to rely solely on chemical pest control which would have both economic and environmental impacts.

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

Interim results of theses studies were presented in two papers at the Entomological Society of America national meeting in Dec. 1992. Journal articles will be prepared following the 1993 field season. Cooperators are sent periodic updates of the project results.

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.