The previous Northeast regional weed management project is entitled “Integrating cover crops, cultivation, and herbicides to optimize weed control”.  The project has 3 objectives and substantial progress has been made in all areas of study.

Objective 1:  Cover crop management.

 An experiment from 1995-1999 evaluated cultivation as a weed control strategy in corn in conservation tillage systems with and without a rye cover crop. Rye was seeded in the fall after silage harvest and then mowed and treated with Roundup before corn planting.  Cultivation with and without additional herbicide applications were used in two tillage systems, chisel plow and no-till, to control weeds.  Data from this four-year study strongly demonstrated that the use of a rye cover crop combined with cultivation and reduced herbicide applications effectively controlled weeds in reduced-tillage corn systems.  The rye cover crop provided substantial soil cover in the fall after silage harvest and if its biomass was greater than 2000 lbs/a dry matter in the spring, it suppressed weeds in the emerging corn crop.  Controlling weeds with cultivation alone was difficult, but reduced herbicide (1/3 recommended rate) combined with cultivation provided effective weed control with little or no effect on corn yields compared to treatments that relied exclusively on herbicides to control weeds.

Objective 2:  Using cultivation and interseeded cover crops to substitute for chemical weed control.
 This research has been conducted in potatoes, cabbage (NY-FVS), and tomatoes (MD).  Red clover, lana and hairy vetch, barley and oats were evaluated in potatoes.  Lana and hairy vetch and the cereals, when interseeded 3 weeks after planting (WAP) and regulated with low rates of postemergence herbicides 8 WAP, resulted in yields similar to conventionally planted potatoes and led to an overall herbicide reduction of 70%.  When not regulated, the cover crops tended to decrease yields.  Red clover proved to be less effective in weed suppression than all the other cover crops.

 Interseeding as a means of weed control in cabbage has been conducted in several phases (NY-FVS).  Phase I evaluated three different potential cover crops: lana vetch, hairy vetch, and oats, cultivated and interseeded 10, 20, and 30 days after transplanting (DAT).  Each interseeding date was preceded by 1, 2, or 3 cultivations at the 10 day intervals.  Results from this period of study indicated that lana vetch was too aggressive/competitive and also that interseeding of all three species was not possible 10 DAT without significant yield losses.  Phase II focused on hairy vetch and oats, interseeded 20 and 30 DAT, with and without supplemental nitrogen to compensate for the competitive effect of the cover crops.  Supplemental nitrogen was sufficient to overcome the yield reductions caused by cover crop competition with both species when interseeded 30 DAT and with vetch interseeded 20 DAT.  Oats however, were still too competitive when planted 20 DAT despite supplemental nitrogen.  Cultivation alone produced yields equivalent to the chemical standard only when done 3 times.  Escaped weeds, while they did not reduce yields, did replenish the soil weed seedbank.  Phase III is in progress, with one year’s data having been collected.  It will be repeated for an additional two years.  This phase is examining the potential for mulch regulation with low doses of selective postemergence herbicides.  The preliminary study’s results indicated that suppression of vetch with 0.094 lb ai/A of clopyralid was highly successful.  Control of oats with 0.2 lb ai/A sethoxydim was less than adequate due to the stage of growth (pre-boot) at the time of application.  Inadequate control of oats led to decreased yield in 1999.

 Lana woolypod vetch was used as a companion cover crop in transplanted tomatoes (MD).  Cultivation 2 and 2 + 4 WAT provided weed-free conditions for seeding the vetch.  Two cultivations were required to prevent significant yield losses due to weed competition in 4 consecutive years.  Without cultivation, weeds reduced yields an average of 68%; a single cultivation led to 18% yield reduction and 2 cultivations to 1% yield reduction.  As a companion crop, the vetch did suppress weed growth.  When seeded at 9 kg/ha 2 WAT vetch did not reduce tomato yields in 5 years of study, and in only one year did 18 kg/ha vetch cause yield losses.  When seeded 4 WAT, (cultivated 2 times) the higher seeding rate caused yield losses in only one of five years.
 It is apparent that interseeded cover crops can be used to suppress weed growth in several vegetable crops successfully.  However, cultivation prior to interseeding is essential.  Moreover, these companion crops will require careful management with regard to need for supplemental fertilizer and possible regulation if the species is particularly competitive.

Objective 3:  Development of models that will predict weed emergence. Predicting weed emergence density as a function of tillage, cultivation, and residue management practices was the focus of research conducted at NY-E&S and ARS.  Several mulches (corn stover, rye, crimson clover, red clover, bark chips, oak leaves, aspen excelsior, and a synthetic mulch) were studied to determine their affect on emergence of lambsquarters, redroot pigweed, wild mustard, yellow foxtail, giant foxtail, and velvetleaf.  When seeds were placed on the soil surface in relatively dry years (NY-E&S), small-seeded species showed a decline in emergence with increasing levels of mulch, whereas, emergence of the large-seeded species tended to increase with low mulch rates and reach the levels of the controls at higher mulch rates.  This suggested that weed emergence was controlled by  both soil moisture for germination, and the availability of seed reserves to nourish seedlings as they grow through the mulches.    When seeds were planted at 1 cm and  soil moisture was not limiting, all species tended to decline with increasing mulch (ARS).   An empirical model was developed (Teasdale and Mohler 2000) that describes the amount of mulch required for a given amount of weed suppression in terms of mulch properties.  When conditions are favorable for seed germination, the proportion of seedlings emerging relative to the unmulched condition, E, is

   E = exp(-b * I * Vc)

Where b and c are fitted constants for a particular weed species, I is the mulch surface area/ground area, and V is the proportion of the mulch volume occupied by solid material.  This equation fitted data almost as well as much more complex forms in which separate coefficients were fitted to each type of mulch.
 A depth stratified simulation model of processes affecting weed seedling emergence is under development, but is not yet operational.  However, a weed seed distribution study was conducted to provide data for this model and it was shown that velvetleaf and common lambsquarters seedlings were 4-9 fold greater in no-till than in conventionally tilled plots in the first year.  In the second year, the weed seedling emergence pattern showed that weed density is controlled by the specific sequence of tillage following a seed rain event.
 Weed seedbank studies in MA indicated that 80 to 85% of all weed emergence occurred from a depth of 0 to 1 inch

 Emergence in response to seed depth and mulch rate was described mathematically (NY(E&S)).  Two sorts of interactions between depth and mulch rate were apparent, an exponential interaction related to the weed suppressive power of combined deep burial under mulch and soil, and a multiplicative interaction, related to enhanced germination from combined shallow burial under mulch and soil.  Significant interaction terms indicated that crop residue and seed position in the soil profile should not be treated independently in the model.

 In the same study, effects of mulch on seed survival were small and positive.  Survival of seeds during their first winter in the soil was high and only slightly affected by depth of burial.  Survival of seeds from spring to spring was low on the surface and higher at depth, with survival rates in the order velvetleaf>lambsquarters>pigweed.  Most seed mortality occurred during the growing season.  High mortality of redroot pigweed and lambsquarters seeds indicated that rotation into sod crops and fall-sown crops will likely reduce seed densities and contribute to control.  Crop rotation probably will have less effect on velvetleaf.

 Studies in NY(E&S) evaluated the ability of different crops and weeds to emerge through various amounts of dead rye straw.  Ability to emerge through the mulch was  related to the log of the seed weight.  Weeds and crop species followed the same seed size-emergence ability curve.  This showed that the use of mulch for weed control will work best when the crop species has large seeds and the dominant weed species have small seeds.

 Developing pre-plant weed seed soil tests that would predict potential weed densities was the focus of research conducted at CT.  It was hypothesized that such predictions would allow ‘fine-tuning’ of preemergence herbicides and enhance the development of IPM programs for weed management.  Following studies to determine the relationship between germination rate and temperature, predictive equations were developed to determine emergence of lambsquarters, and large crabgrass based on pre-planting soil tests (CT).  When an action threshold of one plant/m2 was adopted, the model accurately predicted the appropriate weed management response 86 and 100% for the two species, respectively.  Based on the 50 sampled acres the model identified 20 acres that did not require control for crabgrass, resulting in a savings of up to 60 lbs of herbicide (a.i.).

 Weed emergence prediction experiments evaluated the usefulness of a nonlinear poikiltherm rate equation to describe the relationship between germination and temperature, and a Weibull function to fit the cumulative seed germination for redroot pigweed, lambsquarters, and large crabgrass.  The results of these two lines of research will lead to a coupling of both models to create a predictive model.  The predicted results can be used to optimize weed control timing or can be used as input into larger population dynamics models.  The technology exists to predict which weed species will be present in a field and when they will germinate, however, extensive additions and modifications to the soil testing plan still need to be made before comprehensive IPM programs for multiple crops can become a reality.

 Data from the collected studies allow further generalization of the model that indicates the amount of mulch needed for a given level of weed suppression.  This research will contribute to weed management models that will allow growers to determine or fine-tune optimum weed management strategies.