Benefits - Drawbacks of Various Winter Cover Crops in Vegetable Pest Management

Final Report for LNE00-137

Project Type: Research and Education
Funds awarded in 2000: $89,202.00
Projected End Date: 12/31/2004
Matching Non-Federal Funds: $94,104.00
Region: Northeast
State: Connecticut
Project Leader:
Kimberly Stoner
Connecticut Agricultural Experiment Station
Expand All

Project Information


(Note to the reader: Appendices, graphs, and other supporting documents are available in hard copy from Northeast SARE. Send an e-mail to requesting final report materials for LNE00-137.)

In order to evaluate pest and beneficial insects and microbes in the soil and rhizosphere, we grew a set of cover crops: planted in the fall were wheat, rye, rye + hairy vetch, crimson clover, oats, and, in one year, rape; planted in the spring, following fall oats, were annual ryegrass, spring oats, spring oats + field peas, and in one year, spring-planted rape. We also left an unplanted control for both the fall and spring plantings.

The survival of the cover crops over the winter and performance the next year varied depending on the minimum winter temperatures. Wheat and rye were consistent each year, producing 1-2 tons of dry matter per acre, over 50% ground cover, and suppressing weeds. Rye + vetch also consistently did well each year, but in the spring of 2002, after an unusually mild winter (low temperature of 11 degrees F), grew more vigorously than in the cold years, with close to 100% ground cover and 4 tons of dry matter per acre. Crimson clover varied greatly, with 97% ground cover and 4.7 tons dry matter per acre after a mild winter, but 46% ground cover and only 0.4 tons/acre after a colder winter (low temperature –5 F). Oats also varied, producing a good, winter-killed mulch in spring 2001, surviving the winter and re-growing after mowing in 2002, and dying and breaking down with no dead mulch in 2003. The attempt to follow oats with spring-planted cover crops was not successful, partly due to the variable performance of the oats. Rape did better when planted in the fall than in the spring, but this was over the mild winter, so it should be tested over a series of winter conditions.

In 2001, we studied insect populations as the cover crops re-grew during the month of June after mowing in mid-May. Under those conditions, the rye + vetch cover crop was attractive to the adults of two important species of vegetable pests, the tarnished plant bug and potato leafhopper, and also produced a substantial number of tarnished plant bug nymphs. Thrips (mixed species) were also active and caught in large numbers in wheat, rye, rye + vetch, and annual ryegrass plots.

In 2002 and 2003, we studied insect populations earlier in the season, from April through mid to late May, with the idea that most farmers would be killing and incorporating their cover crops by the end of May. In this part of the season, we found much lower numbers of potential pests, with tarnished plant bugs on rape in early to mid-April (but not producing nymphs) and thrips on wheat in one year in mid-May. We also found a surprisingly high level of activity of parasitic wasps during this period, and we are studying our collections of wasps in more detail to analyze what families and genera were present.

Sorting and analysis of vacuum samples, pitfall samples, and additional sorting of sweep samples is still continuing, so additional information about this experiment may be available in the future.

We did not find consistent patterns among potentially beneficial groups of microbes in the rhizosphere of the cover crops. In general, the numbers of bacteria of all kinds were higher in the roots of the grasses.


Cover crops are used by farmers for many purposes, including prevention of soil erosion, adding organic material to soil, fixing nitrogen, and suppressing weeds. Considerable information has been produced by sustainable agriculture programs to help farmers capture those benefits, including the SARE handbook Managing Cover Crops Profitably. However, information on the role of cover crops in the life cycles of pest and beneficial organisms affecting vegetables is very limited (particularly in the eastern U.S. – there have been several good studies in the west).

Therefore, we decided to grow several of the major fall-planted and spring-planted cover crops and follow pest and beneficial organisms using several methods: sweeps and vacuum to collect insects in the crop canopy, pitfall traps to catch insects and other arthropods walking on the soil surface, and culturing microbes from soil and root cores to quantify microbes from particular groups (fluorescent pseudomonads or Trichoderma) or with specific properties (oxidizers or transformers of Mn).

Project Objectives:

Note that objectives 3 and 4 were not fulfilled, due to problems obtaining enough wireworm larvae to conduct valid experiments. (See Changes in the Plan of Work in the Appendices.)

1. Measure the density of selected insect pests and generalist arthropod predators over time in seven winter cover crops in the field. Determine what effects cover crop management might have on their survival and on movement out of the cover crop.

2. Evaluate the growth and survival of wireworms on roots of cover crops in the laboratory, and determine the effect of cover crop residues incorporate into field soil on their growth and survival.

3. Examine the effect of winter cover crops on diversity of plant-growth-promoting bacteria and deleterious bacteria in the rhizosphere of the cover crop and on the subsequent crops.

4. Isolate bacteria associated with wireworm mortality and examine their effects on plant root health.


Click linked name(s) to expand
  • Wade Elmer


Materials and methods:

Plots: We used a randomized complete block design, with 7 blocks and 10 treatments per block in the winters of ‘00-01 and ‘01-02, and 4 blocks with 6 treatments per block in the winter of ‘02-03. The 10 treatments in ’00-01 included 6 established in the first week of September, 2000 (winter rye, wheat, oats, crimson clover, winter rye + hairy vetch, and an unplanted control) and 4 in which oats were planted in the fall to winter kill, followed by a second spring cover crop (rape, oats + Austrian field peas, oats alone, and annual ryegrass) planted into the killed oats without tilling (March 1 for most plots, but April 3 for field peas and one block covered with snow in early March). In all but one block the size of each plot was 20 ft. X 60 ft., with the remaining block having larger plots of 25 ft. X 130 ft. All plots were mowed May 11-14, 2001. Then, on June 5, each block was divided into thirds (with each plot divided also) with 1/3 left to continue to grow until a final mowing on July 19; 1/3 plowed, treated with herbicide (Prowl) and planted with corn, and 1/3 chopped with a Buffalo chopper and tilled in strips planted with corn.

In the second year of the study (’01-02), the design, plot sizes, and treatments remained the same except that rape was planted in fall instead of in spring, and no oats were replanted in spring, because of high survival of the oats through the mild winter. Plantings were made September 20-25, 2001, and March 7-8, 2002. All cover crops were chopped with a Buffalo chopper on May 20, 2002, and two rows of butternut squash were planted in the crop residue June 10-12, 2002.

In the third year (’02-03), we decided to focus on fall-planted cover crops and to have a smaller number of larger plots. There were 4 blocks with 6 treatments (winter rye, wheat, oats, crimson clover, winter rye + hairy vetch and unplanted control), planted September 18-19, 2002, in plots of 60 ft. X 66 ft. (except one block with plots 41 ft. X 130 ft.) All plots were mowed May 19 and June 5, after which all fields were plowed and a new experiment was begun.

Assessment of Cover Crop Biomass and Weeds: To create a sampling area for weed and cover crop, we made frames of PVC pipe that were 0.50 m2 in area. We visually assessed % cover of weeds and cover crops on three of these frames deployed randomly per plot, and clipped all above ground biomass in 2 frames per plot (May 9, 2001) or 3 frames per plot (June 11,2001). In 2002, this amount of biomass was unreasonably large to clip, weigh, and dry, so we used a much smaller frame (0.125 m2) and took 3 samples per plot. In 2003, we used a frame of 0.37 m2 and 3 samples per plot. We had a lot of difficulty with the main crops of corn and squash because the field had no irrigation and no fences, and because we had difficulty in managing experimental methods of strip-tilling (corn) and no-till (squash) to limit weeds and cover crop regrowth. Data on the main season crops will not be presented here. Weeds were identified with assistance from Dr. Todd Mervosh and Dr. Sharon Douglas of the CT Agricultural Experiment Station.

Insect Collection: Insects walking on the soil surface were collected using pitfall traps. These traps consisted of a 16 oz. cup buried with the top level with the soil surface and with a powder funnel tightly fitted just inside, leading to a 4 oz. cup half-filled with 70% ethanol. To keep out water and slow evaporation, the trap was covered by a roof made of waxed cardboard and held up 1 in. above the soil surface by 3 nails surrounded by 1 in. spacers made from PVC pipe. The traps were collected twice weekly.

Insects in the crop canopy were collected using sweep and vacuum samples. Sweep samples were made by taking 10 sweeps per plot, angled down into the canopy. Vacuum samples were made using a Stihl leaf blower on a vacuum setting, adapted with a screen inserted into the air tube to trap insects. Each vacuum sample was timed at 10 seconds, while the sampler was moved through the crop. One vacuum sample was taken per plot for each sample date. Insect samples were frozen or stored in 70% alcohol for later sorting.

Insect sampling was carried out in different time periods in different years of the study. In 2001, samples were taken from the beginning of June through mid-July, sampling the crop canopy that was allowed to re-grow after the first mowing in mid-May. In 2002, we decided that most growers would turn under their cover crops after mowing, so we sampled from April 11 until mowing on May 19. We followed a similar pattern in 2003, sampling from April 15 to May 30.

Methods Used for Assay of Microbes in the Soil Rhizosphere: Soil was removed with a soil augur or from exhumed roots that were shaken vigorously into plastic bags. Sampled soil was kept on ice until diluted in sterile saline (8.5 g NaCl/liter). Ten-fold dilutions were prepared from the rhizosphere soil and (0.1 ml) spread onto plates of 10% Tryptic soy agar (TSA) (Difco, Inc., Detroit, MI) (dilutions 10-4-10-5 g/ml), King’s B agar (dilutions 10-3-10-4 g/ml), Trichoderman Selective agar (dilutions 10-2-10-3 g/ml), and Mn-dioxide agar (10% Tryptic soy agar, 5.0 g Mn-dioxide, and 20.0 g sucrose, 15.0 g agar) (dilutions 10-2-10-3 g/ml. Soil moisture was determined independently. Plates were incubated in the dark at 25° C. There were 2 or 3 /dilution. Colonies of heterotrophic bacteria were counted daily on the TSA plates for 3 days and summed. Total counts between 30 and 300 colonies/plates were used for estimating bacterial densities. Fluorescent pseudomonads were counted under black light after 2 days. Trichoderma colonies and Mn-transformers were counted 1 week later. Mn-oxidizers produced a blackened deposit whereas Mn-reducers produced a clear zone around the colony. Microbial densities were expressed as CFU/g soil (dry weight equivalent)

Research results and discussion:

Effects of Weather on Cover Crop Growth:

Lockwood Farm is in USDA zone 6b (average annual minimum temperature 0 to –5 degrees F.) The minimum winter temperatures varied strongly from year to year, resulting in variation in cover crop survival and growth (especially oats and crimson clover).

During the winter of 2000-2001, we had one brief cold spell in January, when the temperature went down to –3. In that year, the crops that survived the winter (rye, rye + vetch, and wheat) provided the best weed suppression, but oats suppressed weeds almost as well through vigorous fall growth followed by winter kill, leaving a dead mulch (45-60% ground cover). Crimson clover survived the winter moderately well, with 59% clover ground coverage and 14% weed coverage in May. In June of 2001, the fall-planted oat cover was more broken down, leaving more openings for weed growth, and the spring-planted cover crops did not grow rapidly enough to suppress weeds. Rye + vetch, rye, and wheat re-grew strongly after the May mowing and suppressed weeds, with less vigorous re-growth and weed suppression by crimson clover.

The winter of 2001-2002, was very mild. The lowest temperature was 11 degrees F. Crimson clover and rye + vetch grew very vigorously with high biomass (4.0 – 4.7 tons/acre dry wt.), almost complete ground cover (97-100%) and excellent weed suppression. Rye, wheat, and rape produced 1.8 – 2.1 tons/acre (dry weight) of biomass and 50-60% ground cover. Rye and wheat also provided excellent weed suppression, while rape allowed more weed growth (14% weed cover or 298 g/m2 weed fresh weight). Oats did not winter-kill, and kept re-growing after we mowed it in the spring. Adding additional spring cover crops to the oats did not significantly increase the biomass of cover or reduce the biomass of the weeds.

During the winter of 2002-2003, we had several cold spells below 0 (min. temperature = –5) and more snow and rain. The oats did not grow well, winter-killed early, no residue was visible in the spring, and there was also no weed suppression. Crimson clover mostly survived, but did not grow vigorously, with only 0.4 tons/acre dry weight and 46% clover ground cover. It also allowed weeds to grow to cover 34% of the ground area. Rye + vetch, rye, and wheat produced 1.1 – 1.5 tons/acre (dry wt.) of biomass, 64%-74% ground cover, and they suppressed weeds, keeping most species from reaching the crop canopy and reducing their biomass, and reducing numbers of plants of horseweed and shepherd’s purse.

Densities of Plant-Feeding Insects in the Plot Canopy:

The following are the results from sweep samples from all three years of sampling. We are still in the process of sorting vacuum samples and doing second-level sorting (such as sorting the parasitic wasps by family) on some groups from the sweep samples.

Because we sampled during a different period of the season in 2001 than in 2002 and 2003, we sampled a different insect complex and a different stage of plant growth. For example, the potato leafhopper, Empoasca fabae (Harris), a key pest of potatoes and beans that does not overwinter here, did not arrive in our plots until early June, and thus was found in our samples only in 2001. Rye + vetch was by far the most attractive of our cover crops to the adult potato leafhoppers, and the attraction is clearly due to the hairy vetch component of the mixture, since rye alone had very few potato leafhoppers.

Tarnished plant bug, Lygus lineolaris (Palisot de Beauvois), is another key pest of many vegetables and fruits, and is a species that overwinters here, so it was found in samples every year. Tarnished plant bug adults prefer to feed on flower buds and developing seeds, and move from crop to crop as each flowers, sets seeds, and begins to senesce, so we saw different patterns by sampling at different times during the season when different crops were flowering. In 2001, when we were taking samples in June from re-growth after the cover crops had been mowed in mid-May, rye + vetch was very attractive to the highly mobile tarnished plant bug adults, and, in late June, produced a large number of nymphs. The plots with no cover planted, allowed to grow weeds, were also attractive to adults and produced nymphs in early June. The attractiveness of the fall oats/ spring annual ryegrass and fall oats/ spring rape treatments may be partly due to the weeds that were a significant proportion of the vegetation in these plots. In 2002, when the fall-planted rape grew well over the mild winter, the rape plots were most attractive in mid-April, when the plants bloom and set seed, followed by the weedy plots with no cover planted. Rye + vetch only began attracting adults in substantial numbers in May, along with crimson clover, which blooms in May. In 2003, the attractive plots were oats (which was essentially a weed plot, since no oat mulch lasted the winter), the unplanted weedy control, and crimson clover. In this year, rye + vetch did not have adults or nymphs. Rye alone and wheat consistently had very few tarnished plant bug adults or nymphs throughout all three years.

Another important insect that is known to move from winter cover to vegetable crops is the aster leafhopper, a vector of aster yellows disease. We did not find a consistent pattern of attractiveness to aster leafhopper adults or production of nymphs among the cover crops we studied from one year to another.

Two large groups of insects that we tracked in sweep samples were aphids and thrips. In both cases, it would have been difficult to sort out the individual species. There was no apparent pattern in aphid distribution among the treatments from year to year. For thrips, the strongest pattern was an abundance in wheat. Thrips overwinter in wheat and other grasses and begin flying from those winter crops as the plants set seed and begin to senesce. The figure shows when thrips were associated with the different crops over the month of June in 2001. In 2002, nearly all the thrips were caught in the last two sweep samples (May 10 and May 16).

Densities of Beneficial Insects in the Plot Canopy:

Some of the insect predators we studied were more active in June than in earlier months, and thus only found in large enough numbers to analyze in 2001. These included hover flies (Syrphidae), which were, curiously, most abundant in the plots with the lowest numbers of aphids in 2001, and minute pirate bugs (Orius), which were most abundant in the rye + vetch. Rye + vetch in June 2001 had substantial populations of both aphids and thrips, which are preferred prey for Orius.

Wasps caught in the sweep samples are still in the process of being sorted by family, so the numbers for the overall and very diverse group are presented here. It was unexpected to catch larger numbers of wasps when we sampled earlier in the spring (2002 and 2003) than when we sampled in June (2001). Preliminary information indicated that many of these spring wasps are aphid parasitoids that we caught in the process of emerging from aphid mummies.

Ground Beetles (Carabidae) in Pitfall Traps:

Pitfall traps from 2001 have been sorted, and the ground beetles have been identified to species with the assistance of Dr. William Kring of Yale University. Carabids were abundant in the unplanted weedy plots and in several of the treatments planted with fall oats, which were also weedy by June of 2001. This is not surprising, since many of the more abundant carabid species are prominent feeders on weed seeds (e.g. several species of Harpalus and Amara).

Microbes in the Soil Rhizosphere: Microbial densities from soil cores removed from cover crop soils on 1 June revealed elevated levels of total bacteria in soil cropped to rye. Deleterious Mn-oxidizing microbes were increased in wheat, clover, rye and rye + vetch soil. Beneficial Mn-reducing microbes were increased in soil cropped to clover and wheat. However these trends were not seen on Jun 19 when soils were re-sampled. Rhizosphere soil samples removed on 26 Jul and 6 Aug showed no trends in microbial densities among cover crops. Levels of beneficial microbes, such as the fluorescent pseudomonads, and deleterious microbes, such as the Mn-oxidizing microbes were both in greater densities in rhizosphere soil stripped-tilled than rhizosphere soils under conventional management.

In June 13, 2002, total bacterial densities were generally lower than in 2001, and were increased in soil cropped to rye + vetch, but the cover crop treatments had no effect on other microbial groups. Another plot about 100 m away was cropped in 2001 to canola, crimson clover, rye or left bare. Soil cores removed in November revealed that total bacterial densities and Fluorescent pseudomonads were in greater numbers in soil planted to canola than the bare ground whereas numbers of beneficial Trichoderma spp and Mn-reducing bacteria were enhanced in the plot cropped to rye.

In addition, rhizosphere soil from nine summer cover crops was sampled for microbial densities. Oilseed radish harbored over 100 times as many fluorescent pseudomonads/g rhizosphere soil as buckwheat, but had the fewest number of Mn-reducing microbes of all the summer crops.

Discussion and Conclusions:

Cover Crop Growth and Weed Suppression: In the climate of Lockwood Farm (USDA Zone 6b), rye + vetch, rye, and wheat consistently suppressed weeds over the winter and spring, produced substantial biomass (always more than 1 ton dry weight per acre) and over 50% ground cover in early May. In the mild winter of 2001-2002, the rye + vetch responded with more vigorous growth, almost 100% ground cover, and 4 tons per acre of biomass. Crimson clover and oats varied even more dramatically in their performance from year to year. Crimson clover produced only 0.4 tons per acre biomass in 2003 compared to 4.7 tons per acre in the mild year of 2002. Fall-planted oats were intended to grow in the fall and winter-kill, leaving a dead mulch in the spring, but this was the result only in 2001. In 2002, they did not winter kill, and in 2003, they winter-killed too early, with too little growth to leave much dead mulch. The approach of planting spring cover crops after fall oats never worked well for a combination of reasons: difficulty in effective seeding of the spring crop into the plots without tillage and variation in the oats from one year to the next. Rape performed better when planted in the fall than in the spring, but should be evaluated again, since the result might be influenced by the warm winter of 2002.

The lack of predictability of performance of crimson clover and oats could be a problem for farmers planting them in our climate with a specific result in mind. Crimson clover could be a good option for a vigorous leguminous cover in a climate where the winter temperatures dependably stay above 0 degrees F, though.

Insects in Cover Crop Canopy and Pitfalls:

The strategy of keeping a rye + vetch cover going into the summer by mowing it and allowing it to re-grow, used by some farmers along with a summer fallow as a weed control strategy, has the potential to attract and build up large numbers of potato leafhoppers and tarnished plant bugs in the cover crop in June. A similar practice with wheat, rye, or rye + vetch (and possibly annual ryegrass planted in spring) could also build up populations of thrips, some species of which are direct pests of vegetables (onion thrips, Thrips tabaci Lindeman) and others which can transmit disease to vegetable crops (Frankliniella occidentalis (Pergande)). Except for a modest number of thrips in wheat in mid-May 2002, and adult tarnished plants bugs in rape in mid-April 2002 (which apparently did not reproduce there, since we didn’t find nymphs) we did not find a build-up of deleterious pests in cover crops in April to May. May is when most farmers would kill and incorporate their cover crops in our region.

The most interesting result we found among the beneficial insects in the cover crops was the higher than expected activity of parasitic wasps in the cover crops in April and May. We will be sorting these wasp collections and analyzing them further, but it seems likely that a large proportion of these wasps are aphid parasitoids, and the role of cover crops in building up populations of aphid parasitoids that can help in biological control of aphids in the subsequent vegetable crops would be a valuable topic for additional research.

Microbes in the Soil Rhizosphere: Although no consistent pattern could be established, total densities of microbes including the beneficial microbes such as fluorescent pseudomonads, Trichoderma spp. and Mn-reducing microbes, tended to be higher under gramineous crops like rye and wheat and under the legume, crimson clover, than other cover crops or in bare ground. Deleterious microbes like Mn-oxidizng microbes would sometimes increase as well under these crops, but they were also favored by canola.

Participation Summary


Educational approach:

Research results from this project were presented in a workshop at the NOFA Summer Conference (Hampshire College, Amherst, Mass.) in August, 2003 and in a poster at the National SARE Conference in Burlington VT, in September 2004.

Project Outcomes

Areas needing additional study

Having done a general survey of insects on these cover crops, I would recommend that a different approach be taken in future studies – picking a particular insect and following it through its life cycle in fields with different cover crops. Tarnished plant bug would be a good candidate as a pest moving through different cover crops, and the complex of aphid parasitoids in vegetables would be a good group of beneficial insects to follow.

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.