erosion (Giller and Cadisch, 1995), suppress weeds (Exner and Cruse, 1993), and fix biological N (Giller et al.,The rising real prices of purchased inputs is driving smallholder 1994). In addition, certain legume crops can provide maize (Zea mays L.) production towards lower levels of inorganic fertilizer. Legume intercrops are a source of plant N that can be food to humans and or livestock. The integration of produced locally and offer a practical complement to inorganic fertiliz-small amounts of inorganic fertilizer N along with orers. Field experiments conducted on a loamy sand (Typic Kandiustalf) ganic material and N from annual legumes offers a stratsoil evaluated the impact of relay-intercropping two legume crops, egy to meet the N needs of smallholder farms (Wadcowpea (Vigna unguiculata L.) and sunnhemp (Crotolaria juncea dington and Heisey, 1997). L.) into smallholder maize in Zimbabwe. The objectives were to Legume green manures are not new to the region. quantify: (i) biomass and N yield of intercropped legumes, (ii) the They were heavily researched in Zimbabwe from the impact of the legumes on companion maize yield and N uptake, and 1920s to 1940s (Metelerkamp, 1988), and large-scale (iii) the response of a subsequent maize crop to legumes. Dry matter commercial farms (the second sector in a dichotomous yield ranged from 0.6 to 4.6 Mg ha Ϫ1 for cowpea and 0.9 to 2.9 Mg ha Ϫ1 agriculture) used green manures widely. There were for sunnhemp, over two years. At the most, cowpea and sunnhemp produced 154 and 82 kg N ha Ϫ1 , respectively. Companion maize grain informal reports that some smallholders used some yields were not reduced when the legumes were relay-intercropped green manures such as sunnhemp to maintain soil fertilinto maize fertilized at 0 to 60 kg N ha Ϫ1 . However, maize yields were ity (Hikwa et al., 1998). This practice continued until reduced 18 to 31% when maize ϩ legume intercrops were fertilized the cost of inorganic fertilizers fell in the 1950s and at 120 kg N ha Ϫ1 . In the subsequent year, maize grain yields were green manures became uneconomical (Tattersfield, increased by 8 to 27% following maize ϩ legume when no fertilizer N 1982). Rising real prices of inorganic fertilizers in the was applied, compared with maize following maize. Legumes reduced 1990s and concerns about the sustainability of current fertilizer needs of a subsequent maize crop by 36 kg N ha Ϫ1 . Intersmallholder cropping systems have renewed interest in cropped annual legumes and small amounts of inorganic fertilizer green manures and related legume technologies (Hikwa offers a strategy to meet the N needs on smallholder farms.

The lack of a definitive method to assess winter hardiness in alfalfa (Medicago sativa L.) remains a challenge in the north‐central region of the USA where winterkill of alfalfa can be severe. The reliability of fall dormancy ratings for describing alfalfa cultivar susceptibility to winter injury and the role of snow depth in moderating temperatures near the plant were investigated at Chatham, MI on a Chatham Stony loam (Typic Haplorthod). Four cultivars were selected with a range of fall dormancy ratings: ‘Nitro’, ‘Magnum IV’, ‘Saranac’, and ‘Vernal’. The cultivars were planted in 1993–1994, 1994–1995, and 1995–1996 seasons in plots over which 0‐, 10‐, and 20‐cm winter snow depths were maintained. Temperatures were monitored for each plot, and stand counts were made each fall and spring to assess winter injury. Nitro suffered the most winterkill across snow cover treatments. The total yield range was 0 to 9 Mg ha−1 in the absence of a snow cover and 0.4 to 12 Mg ha−1 for a snow depth of at least 10 cm, except in 1996. Extreme minimum canopy‐level (6 cm) temperatures for 10‐cm snow depth averaged over three winter seasons were 12.1°C higher than the 0‐cm snow cover treatment, which translated into higher yields. The results suggest that snow cover of 10 cm adequately protects alfalfa from winter injury. Cultivars within the same fall dormancy rating did not necessarily perform similarly, suggesting the need to develop other methods for assessing winter survival.

Intercropping annual forage legumes with row crops has been proposed as a strategy to control erosion, suppress weeds, and contribute biological N to companion or subsequent crops. This experiment evaluated the effect of planting date of two annual medic species [burr and snail medic, Medicago polymorpha L. and M. scutellata (L.) Mill.] on yield and N accumulation of both medic and interseeded corn (Zea mays L.). Medics were either clear‐seeded (i.e., monocropped) or interseeded into corn at five planting dates from 7 May to 13 July in each of two years at East Lansing, MI, on a Capac loam (fine‐loamy, mixed, mesic Aerie Endoaqualf). Clear‐seeded medic produced up to 3 Mg ha−1 of dry matter at 60 d after planting. In the interseeding system, medic dry matter yields were only 50% of yield in a clearseeded system. Average aboveground medic N concentration was 50 and 20 kg ha−1 in clear‐seeded and interseeded systems, respectively. Interseeding at corn planting produced the greatest medic dry matter, but also was associated with the greatest reduction in corn yield. Corn yields were not reduced by interseeding medic 28 d after com planting. Interseeding medic into corn was associated with a positive yield response (3–25%) of subsequent corn grain. Medics reduced fertilizer needs of corn in the subsequent year by 37 kg N ha−1.

The commercial production of cranberries relies on abundant water resources for frost protection, soil moisture management, and harvest and winter flooding. Given water resource demands and regulations in southeastern Massachusetts, we sought to quantify the annual water requirement for the commercial production of cranberries. Based on 2 yr of monitoring across five sites, the mean water requirement for cranberries was 2.2 (±0.6) m yr −1 (one standard deviation in parentheses). On average, the 3 mo maximum area threshold of 3.15 ha was within~20% of the value currently used to establish water permits for renovated cranberry farms in Massachusetts. Variation in the water requirement was primarily related to differences in the harvest and winter floods, which combined for two-thirds of the annual water requirement. The water requirement for the winter flood exhibited the greatest annual variation (54%), which was relatively low for the harvest flood (20%). Environmental variation was significantly related to water requirements for the winter flood, as well as seasonal irrigation, and should be carefully considered in agricultural water use regulations.

Grass–legume mixtures have the ability to supply more consistent forage yields across a wide range of environments throughout the grazing season than do grass monocultures. The suitability of diverse grass species in binary mixtures with birdsfoot trefoil (Lotus corniculatus L.) in rotational stocking systems has not been extensively studied. The objective of this study was to evaluate binary mixtures of five cool‐season grasses with the birdsfoot trefoil cultivar Norcen for herbage mass, botanical composition, and cattle (Bos taurus) grazing preference under a rotational stocking. Experiments were established at Lake City and Chatham, MI, in 1994. Binary mixtures were grazed for 2 yr with beef or dairy cows three times yearly at predetermined periods from spring to fall. Total herbage dry mass production ranged from 3 to 10 Mg ha−1 yr−1 over two years and locations. The grass fraction in binary mixtures was 327 to 946 g kg−1 in swards over two years and locations. Perennial ryegrass (Lolium perenne L.) failed to persist at Lake City, probably due to less consistent snow cover. Birdsfoot trefoil fraction was highest in binary mixtures with smooth bromegrass (Bromus inermis Leyss) and timothy (Phleum pratense L.). Binary mixtures with orchardgrass (Dactylis glomerata L.) and tall fescue (Festuca arundinacea Schreb.) produced the highest herbage biomass but were less preferred by grazing animals while binary mixtures with timothy and smooth bromegrass were associated with the highest apparent herbage utilization at both locations (84–100%).