Evaluation of Late-Summer Cover Crops
in the SAFS Rotation
Pio Rosell, Peter Brostrom and Steve Temple
Agronomy and Range Science Department, UC Davis
The 4-year, 5-crop rotation of the Sustainable Ag Farming Systems Project (SAFS) has in its first 8 years of production demonstrated a compelling need for a fast-growing, N-fixing legume variety for use in monoculture and/or as part of a cereal/legume cover mix, in Sacramento Valley tomato-based farming systems. The specific niche (envisioned specifically for the cover crop driven treatments) would be following harvest of early processing tomatoes, and before planting early spring crops like safflower or sunflower, or corn.
Earlier tests with existing commercial varieties of blackeye cowpea (Vigna unguiculata) planted in August has showed them to be strong seedling competitors with weeds. However, they typically demonstrate short day photosensitivity that causes them to reduce vegetative growth and to commence flowering in September, after reaching only a modest height and limited biomass. The objective of the present studies are to identify legume components of a sorghum-sudan/legume mix that will: 1) optimize vegetative growth on (largely) residual soil moisture, 2) capture and cycle forward carbon and soil mineral N following the tomato crop, and 3) winter kill in a manner that allows minimal land preparation prior to establishing March/April plantings of summer cash crops the following season.
In earlier SAFS and UC Student Farm field observations, lablab (Lablab purpurea) performed very well in monoculture, and reasonably well in a mixture with sorghum-sudan hybrid and cowpea. However, the seed size of tested lablab varieties, plus the late and uncertain seed set in Sacramento Valley tests, have combined to make the prospective seed costs prohibitive for a cover crop. We have generally observed that, among the heat-loving tropical and subtropical annual legumes, those species and varieties that demonstrate extensive growth, cold tolerance, and weed competitiveness into November and early December, are often the genotypes that are most problematical with respect to seed production and cost.
1997 Trial
Following an early August, 1997 harvest of tomatoes at the SAFS companion area at UC Davis, we planted 36 Vigna and lablab varieties (mostly cowpea genotypes) on 250 square foot plots in a randomized complete block design (3reps). The legume monoculture system, which included sorghum-sudan borders every 5 feet, was designed to partially simulate the shade anticipated in a mixed stand. Check treatments included monoculture sorghum-sudan, and the mixture of sorghum-sudan and Red Ripper cover crop cowpea (used currently in the main SAFS experiment). Data on emergence and stand establishment, seedling vigor and competition with weeds, and aboveground biomass and N content were taken. No attempt was made to differentiate between mineral N captured from the soil following the tomato crop versus N fixed by the legumes tested.
The same cowpea lines were planted in June 1997 in large, non-replicated plots at the Kearney Ag Center south of Fresno CA, to evaluate biomass and ground cover potential in an environment more representative of the hot San Joaquin Valley tomato production systems. Seed production and response to attack by Lygus, spider mite, and whitefly were scored, since these factors are all important in evaluating the economic and agronomic utility of cover legumes in tomato-based farming systems. In particular, Lygus can be a serious pest on tomato, beans, and safflower, three components of the SAFS crop rotation.
The 1997 Davis trial gave excellent results (cv for biomass was 15.6, and treatment differences were very highly significant @p=.0001). Highest biomass and N yields were obtained from the sorghum-sudan monoculture, followed by the SS/Red Ripper mixture, and then by the three lablab entries. The monoculture cowpeas, which in 1997 frost-killed in advance of the lablab, were significantly lower in biomass and N yield. From the 1997 test, we were able to choose ten of the most promising cowpea genotypes to evaluate the following season as components of a SS/cowpea mixture.
1998 Trial
For the 1998 trial, ten cowpea genotypes were evaluated in sorghum-sudan/cowpea mixtures. Because of the excellent 1997 performance of the sorghum-sudan and lablab treatments, the 1998 trial included not only the SS/Red Ripper as a check, but also SS/lablab, and monoculture sorghum-sudan, monoculture Red Ripper, and monoculture lablab. Biomass and nitrogen data (Table 1) were again of very high quality. Legume biomass showed a cv of 17.7, with highly significant treatment effects, while sorghum-sudan biomass was less consistent among plots and across legume varieties. The cv for total biomass (legume plus sorghum-sudan) was 24.4, and treatment differences were significant @ p< .001. Comparisons of the data for the two seasons are interesting, especially because temperatures were quite different.
Sorghum-sudan monoculture was in the top biomass group both seasons, suggesting that under favorable mineral N conditions it has the potential to be an excellent soil builder (and N-capturer). The mixture of SS/Red Ripper (our main plot standard) also performed very well in both seasons, although the Red Ripper monoculture was the lowest treatment across seasons. Further, the Red Ripper biomass appears to be substituting for that of sorghum-sudan, suggesting that it may contribute quality but little quantity to mixture biomass above that of the SS monoculture. The SS/lablab mixture produced the greatest total biomass yield in the trial. Data show that the relatively strong effect of lablab depressing sorghum-sudan biomass was more than offset by the excellent contribution from lablab, whose growth in 1998 was especially apparent late in the fall after all cowpea genotypes were senescing, and some had completely defoliated.
Four SS/cowpea mixtures (including that with Red Ripper) demonstrated total biomass yields close to that of the sorghum-sudan monoculture and the SS/lablab mix. Those four cowpea genotypes will be evaluated in more detail in 1999, using seed harvested from a June planting nearby. Because seed cost and the potential for local seed increase have been included as selection factors over 4 seasons of work, we are confident these elite cowpea lines can be economically multiplied locally.
From these studies, we conclude that a lablab/sorghum-sudan mixture would provide the best combination of biomass and nitrogen following tomato and preceding safflower in the current SAFS rotation. We believe this is due, at least in part, to the complimentary, non-competitive growth cycles of the two species when planted in August. The weed-suppressive potential of lablab has now been observed in several cover crop trials, and may be useful in systems where cover crops and reduced tillage/weed management are key goals. The very delayed senescence of lablab foliage may also influence the nitrogen cycling dynamics and (possibly) the amount of N lost during the winter.
The set of experimental Vigna lines did not produce total biomass nor nitrogen significantly better than the Sorghum-sudan/ Red Ripper mixture currently used at SAFS, but lines like PI 293-522 and PI 339-603 were more competitive with the fast-growing sorghum-sudan than Red Ripper and other introduced Vigna selections. Those differences in growth rate and speed of row cover are important to the suppression of late summer/fall weeds in the current system.
Table 1. Legume and sorghum-Sudan contribution to total mixture biomass (kg m-2) in 1998 summer cover crop trial. (Summarized)
| Average Biomass | (kg m-2) | %N |
Total N | ||||||
| Treatment | ID | Sorghum | Cowpea | TOTAL | Sorghum | Cowpea | kg ha-1 |
||
| Sorghum | VG 834 | 6 | 0.17 | 0.19 | 0.36 | 1.64 | 3.01 | 86.05 | |
| Red Ripper | mono. | 12 | 0.25 | 0.25 | 3.69 | 91.71 | |||
| LABLAB | mono. | 15 | 0.41 | 0.41 | 3.44 | 141.90 | |||
| Sorghum + | PI 293-522 | 4 | 0.26 | 0.20 | 0.46 | 2.29 | 3.43 | 128.13 | |
| Sorghum + | Red Ripper | 1 | 0.41 | 0.17 | 0.58 | 1.86 | 3.04 | 127.52 | |
| Sorghum | mono. | 13 | 0.59 | 0.59 | 2.12 | 125.85 | |||
| Sorghum + | PI 339-603 | 9 | 0.39 | 0.21 | 0.59 | 2.45 | 3.16 | 160.10 | |
| Sorghum + | LABLAB | 2 | 0.23 | 0.40 | 0.63 | 1.86 | 3.33 | 174.92 | |
| Average | 129.52 | ||||||||
Return to the index for the
Cover Crop Research and Education Summaries