Native Grass Species for use as Perennial Cover Crops in San Joaquin Valley Vineyards
Final Report - August 1999
Principal Investigator:
Michael Costello, Farm Advisor, Fresno County
Costello Agricultural Research & Consulting
P.O. Box 165
Tollhouse, CA 93667
(559) 855-2847, michaelcostello@mindspring.com
Other Investigators:
Dan Munk, Farm Advisor, Fresno County
Kurt Hembree, Farm Advisor, Fresno County
Chuck Ingels, Farm Advisor, Sacramento County
Location of project: San Joaquin Valley
Commodities: winegrapes
Funding:
FY 1997-98: $15,000
Table of Contents:
Introduction
Materials and Methods
Results and Discussion
Summary
Acknowledgements
Figures: 1, 2, 3, 4, 5, 6
Tables: 1, 2, 3
Literature Cited
Introduction
Managing vegetation in California grape vineyard middles is commonly achieved by clean cultivation or mowing. Each of these methods has its benefits and drawbacks. Cultivation keeps weeds down, but has the undesirable side effect of creating dusty conditions, which decreases photosynthesis and can decrease vine resistance to mite attacks. In addition, water penetration is commonly a problem through sandy loam and fine sandy loam soils when clean cultivated: clean cultivation disturbs cover crop root channels and increases the rate of organic matter decomposition (Land, Air and Water Resources-Cooperative Extension Joint Infiltration Committee 1984, Gulick et al. 1994). On the other hand, maintaining a vegetative cover improves soil structure and water infiltration (Prichard 1998), reduces dust, allows equipment to be used in wet conditions and cuts down on the sunburning effect of reflected light. The contribution of cover crops to improved vine water status and dust reduction can play an important role in spider mite management (Flaherty et al. 1971).
Mowed vegetation can be either a planted cover crop or resident (weedy) vegetation (Ingels and Klonsky 1998). Cover crops are usually perennials, either legumes such as white or strawberry clover, or grasses such as ryegrass, fescue or orchardgrass. Resident vegetation is often comprised of (but not limited to) grasses such as foxtail (Setaria spp.), crabgrass (Digitaria sanguinalis) and cupgrass (Eriocholoa gracilis). Management of these ground covers consists of periodic mowing (every 2-3 weeks). Cover crops have the advantage of crowding out weeds, whereas over time management of resident vegetation usually leads to increased presence of undesirable weeds such as bermuda grass (Cynodon dactylon) and purple nutsedge (Cyperus rotundus). However, all of these ground covers use more water than that lost from cultivated ground and make unavailable some portion of the soluble soil nitrogen. They therefore have the potential to be too competitive with the vines, causing decreased vigor and possibly lower yields over time (Costello and Daane 1997). The presence of cover crops from budbreak till mid-April can also increase the risk of frost injury by preventing full absorption of solar radiation by the vineyard floor.
An alternative to resident vegetation management or clean cultivation is the use of perennial native California grasses. Native grasses evolved in California and so are well adapted to our climatic conditions. Their major advantage is in having a phenology opposite to that of the grapevine, that is, they begin growing with the first fall rains (just as the vines are going into dormancy), and go dormant in the summer. As a result, they provide the advantages of a perennial cover without the disadvantage of excessive competition. Several winegrape growers in the Lodi-Woodbridge region have planted native grasses such as California brome (Bromus carinatus), blue wildrye (Elymus glaucus) and Idaho fescue (Festuca idahoensis) for purposes of dust control, weed control and accessibility of equipment under wet conditions (Howe 1994; Ingels 1998). Daane & Costello (1994) monitored pests, natural enemies and measured vine water status in a vineyard with an established cover crop of California brome, blue wildrye and meadow barley (Hordeum brachyantherum). They found lower late-season water stress in vines with native grass cover compared to clean cultivation, and suggested that this was due to higher water infiltration rates. no research has been conducted on native grass culture in the central San Joaquin Valley.
Native grasses fall into two broad categories with respect to phenology: those that go dormant with shorter days (daylength obligate dormancy) or those that go dormant with decreasing water availability (moisture obligate). This study will look at the possibilities of using daylength obligate native grasses in drip irrigated vineyards, and moisture obligate native grasses in flood irrigated vineyards.
Materials and Methods
The main experimental site was at Kearney Agricultural Center (KAC) in Parlier, where the cover crop treatments were established in the fall of 1996. A second study site was established in the fall of 1997, at a dry-on-the-vine raisin vineyard in west Fresno owned and operated by the Seibert family. The Seibert site was set up primarily to look at the effect of native grass cover crops on spider mite populations, but in the 1998 season almost no mites were found at this site. Therefore, all the results shown are from the KAC site.
At KAC, two separate experiments with native grasses as cover crops were planted in the fall of 1996: one in a one-acre drip irrigated block of a seven year old Barbera vineyard, the other in a one-acre flood irrigated block of a seven year old Grenache vineyard. Soil type at the site was a Hanford fine sandy loam. At each site, four native grass treatments were compared to blando brome (Bromus hordeaceus), and a clean cultivated control. For the drip irrigated block two of the original treatments included California melic (Melica californica), which did not establish. As a result the treatments were: 1) Resident vegetation, 2) Nodding needlegrass (Nassella cernua), 3) California barley prostrate (Hordeum brachyantherum ssp. californicum) and 4) California barley prostrate + nodding needlegrass. For the flood irrigated block treatments were: 1) purple needlegrass (Nassella pulchra), 2) nodding needlegrass, 3) California brome (Bromus carinatus) + blue wildrye (Elymus glaucus) and 4) California brome. The California brome/blue wildrye treatment originally included Meadow barley (Hordeum brachyantherum ssp. brachyantherum), but it did not establish. The blando brome treatment in the flood irrigated block was replaced by a naturalized stand of yellow foxtail in the summer. The vines were on 7 x 10 ft spacings, plot size was five rows by 6 vines and treatments were replicated three times in a randomized complete block design. At the drip irrigated site, vines were irrigated at full evapotranspiration (ET) throughout the season, whereas at the flood irrigated site the vineyard was irrigated about every three weeks.
In each of the KAC sites, soil moisture status was measured from the nodding needlegrass and clean cultivated treatments, taking bi-weekly readings with a neutron probe (Model 503DR, Campbell Pacific Nuclear, Martinez, CA), with one neutron probe tube placed within the vine row (in-row) and one between the vine rows (middles) in each nodding needlegrass and clean cultivated plot. Readings were taken every foot from one to five feet from May through September. Vine water status was estimated bi-weekly from these same treatments using a pressure bomb (PMS Instruments, Corvalis, OR), taking five readings per plot between the hours of 1100 and 1600. The spring and early summer of 1998 were unusually rainy, and at the flood irrigated site, irrigations did not begin until July. Grape yield was be estimated by harvesting and weighing the fruit from four half-vines per plot. In winter two vines per plot were pruned, and the brush weighed as an estimate of vine vigor.
Results and Discussion
Key weeds at the drip irrigated site were soft chess (Bromus mollis), spotted catsear (Hypochoeris radicata), and foxtail barley (Hordeum leporinum). These weeds constituted a high proportion of the resident vegetation treatment at this site. Key weeds at the flood irrigated site were ryegrass (Lolium perenne), cudweed (Gnaphalium purpureum), cupgrass (Eriochloa gracilis), zorro fescue (Vulpia myuros), spotted catsear and soft chess. Native grass species which did not establish well were California oniongrass at the drip irrigated site and meadow barley and blue wildrye at the flood irrigated site.
Figures 1-4 show soil moisture data by irrigation method and location of the neutron probe tubes, either within the row or between-rows (the middles). At the drip irrigated site, there was a pattern of higher in-row soil moisture for the nodding needlegrass treatment at 1-4’ (Fig. 1). However, at all soil depths between-rows, soil moisture levels were higher for clean cultivation (Fig. 2). Whereas one might expect the needlegrass treatment to use more water than clean cultivation, which is what was found between-rows, it is a bit puzzling why in-row soil moisture was higher in the needlegrass plots. One suggestion was that the vines in the needlegrass treatment did not grow as well, and therefore had a smaller canopy that used less water. However, the pruning weight data do not bear out this theory, as there were no significant differences in pruning weights (Table 2). If anything, it appears that vine vigor in the nodding needlegrass treatment was somewhat higher than clean cultivation. Somehow the presence of the needlegrass allowed more water to be absorbed into the soil at the drip emitters, although it is difficult to understand how a cover crop several feet away could have influenced this. Overall, water use was slightly higher with the needlegrass cover crop at the drip irrigated site. Over the entire period of monitoring (May-September), average soil moisture was 1.15 percentage points higher in-row in the nodding needlegrass treatment, and 2 percentage points higher between rows in the clean cultivated treatment (Table 1). Overall, soil moisture averaged 13.88 % (± 0.33) in the nodding needlegrass plots, and 14.31 % (± 0.32) in the clean cultivated plots.
At the flood irrigated site, early season water use was considerably higher in the needlegrass treatment. Soil moisture was higher in the clean cultivated treatment early in the season, but from July 15 on, it was higher in the nodding needlegrass treatment at depths of 4 and 5’ (Figs. 3 and 4). The first irrigation was on July 10, followed by irrigations on July 27 and August 25. It is clear that with each irrigation the needlegrass cover allowed water to penetrate deeper, as soil moisture at the four and five foot levels were consistently higher than with clean cultivation. It appears that the increased storage due to the cover crop balanced the water used by it, and overall water use was about equivalent. Over the entire period of monitoring, average soil moisture was about equivalent between treatments in-row, and was 1.1 percentage points higher in the nodding needlegrass treatment in the middles (Table 1). Overall, soil moisture averaged 10.54 % (± 0.28) in the nodding needlegrass plots, and 10.27 % (± 0.22) in the clean cultivated plots.
Although there was considerable variation in soil moisture status between needlegrass and clean cultivated treatments, vine water status did not begin to differ between treatments until late August, and only on the very last sampling date (which was post-harvest) was it significantly different (Figs. 5 & 6). On that date at both the drip and flood irrigated sites, vine water stress was lower in the nodding needlegrass treatment by about 10%.
At the drip irrigated site, average yield of Barbera was highest in the nodding needlegrass and Blando brome treatments, intermediate with clean cultivation, nodding needlegrass/California barley blend and resident vegetation, and lowest with the pure stand of California barley (Table 2). However, there was a great deal of variation among the samples, and the only statistically significant difference was found between the highest yield (nodding needlegrass) and lowest (California barley). Differences in pruning weights were also not significant, although the pattern of weights among the treatments suggests that there was a relationship between vigor and yield at this site (Table 2). It is interesting to point out a pattern at the drip irrigated site. The highest yield and vigor was found with a pure stand of nodding needlegrass, whereas vines grown with California barley prostrate had the lowest yield and were among the lowest in vigor. The blend of nodding needlegrass and California barley prostrate fell in between in terms of yield. This suggests that California barley prostrate is more competitive than nodding needlegrass. Unfortunately, we did not have the resources to monitor water use in the California barley prostrate treatment, but it is possible that this grass is more water consumptive than the needlegrass.
At the flood irrigated site, average yield of Grenache was highest in the clean cultivated treatment, with yield in the blando brome, purple needlegrass and nodding needlegrass in the intermediate range, and California brome/blue wildrye blend and pure stand of California brome at the lowest level (Table 3). Again, there was a lot of variation among the samples, and no statistically significant differences were found. Pruning weights were quite similar and not statistically significant.
Summary
Perennial native grasses were tested as cover crops in two vineyards at the Kearney Agricultural Center in Parlier. In a drip-irrigated vineyard of Barbera, the cover crops which established were nodding needlegrass, California barley prostrate and a blend of the two. An attempt to establish oniongrass failed. In a flood irrigated block of Grenache, the cover crops which established were purple needlegrass, nodding needlegrass, California brome and a blend of California brome and blue wildrye. Meadow barley was seeded but failed to establish. Seasonal monitoring was done for in-row and between-row soil moisture and plant water status, at harvest yield was measured, and in winter pruning weights were taken as an estimate of vine vigor. At the drip irrigated site, overall soil moisture status was higher in-row for the nodding needlegrass treatment at 1-4’, but soil moisture levels were higher between-rows for clean cultivation. At the flood irrigated site, soil moisture in the nodding needlegrass treatment was lower until the first irrigation. With each irrigation (July 10, July 27 and August 25), soil moisture status at the deeper levels was elevated in the needlegrass treatment. Very few significant differences were found among cover crop treatments in yield or pruning weight. The pattern of average yields at the drip irrigated site were: highest in the nodding needlegrass and Blando brome treatments, intermediate with clean cultivation, nodding needlegrass/California barley blend and resident vegetation, and lowest with the pure stand of California barley, and average pruning weights closely followed this pattern. The pattern of average yields at the flood irrigated site were: highest in the clean cultivated treatment, intermediate in the blando brome, purple needlegrass and nodding needlegrass treatments, and lowest in the California brome/blue wildrye blend and pure stand of California brome.
Acknowledgements
Thanks to Scott Stewart of Conservaseed (Rio Vista, California) for project inspiration and for providing the native perennial grass seed for both study sites. Thanks to Jonathan Wroble, Juliet Schwartz, Kimberly Miyasaki and Jose Cantu for help with data collection.
Figures 1-6
TABLES 1-3
Table 1. Season-wide average percent soil moisture (± standard error of the mean) for nodding needlegrass and clean cultivated treatments, 1998.
| Irrigation system and variety | Neutron probe location | Cover crop treatment | Average season-wide soil moisture (% w/w) ± standard error |
| Drip irrigation, Barbera | In-row | Nodding needlegrass | 16.22 ± 0.36 |
| Clean cultivation | 15.07 ± 0.40 | ||
| Between-rows | Nodding needlegrass | 11.54 ± 0.48 | |
| Clean cultivation | 13.55 ± 0.48 | ||
| Flood irrigation, Grenache, | Between rows | Nodding needlegrass | 11.40 ± 0.43 |
| Clean cultivation | 10.50 ± 0.35 | ||
| In-row | Nodding needlegrass | 9.66 ± 0.31 | |
| Clean cultivation | 10.05 ± 0.27 |
Table 2. Average yield (kg fruit/vine) and pruning weight (kg prunings/vine) of Barbera, drip irrigated site, 1998 season.
| Treatment | Yield (kg/vine ± standard error) | Pruning weight (kg/vine ± standard error) |
| Nodding needlegrass | 26.05 ± 2.50 a | 2.63 + 0.17 |
| Blando brome | 25.20 ± 2.38 ab | 2.36 ± 0.17 |
| Clean cultivation | 21.14 ± 2.10 ab | 2.32 ± 0.17 |
| Nodding needlegrass/Cal barley prostrate | 21.01 ± 2.06 ab | 2.20 ± 0.19 |
| Resident vegetation | 20.4 ± 2.25 ab | 2.08 ± 0.19 |
| Cal barley prostrate | 16.9 ± 1.84 b | 2.13 ± 0.22 |
| ANOVA | 0.039 | 0.374 |
Table 3. Average yield (kg fruit/vine) and pruning weight (kg prunings/vine) of Grenache, flood irrigated site, 1998 season.
| Treatment | Yield (kg/vine ± standard error) | Pruning weight (kg/vine ± standard error) |
| Clean cultivation | 27.14 ± 1.82 | 3.91 ± 0.15 |
| Blando brome (yellow foxtail) | 25.86 ± 1.31 | 3.34 ± 0.31 |
| Purple needlegrass | 25.50 ± 1.82 | 3.81 ± 0.17 |
| Nodding needlegrass | 24.03 ± 2.88 | 3.79 ± 0.27 |
| California brome/blue wildrye | 22.34 ± 3.12 | 3.63 ± 0.18 |
| California brome | 20.09 ± 2.00 | 3.41 ± 0.22 |
| ANOVA | 0.325 | 0.444 |
Literature Cited
Costello, M.J. and K.M. Daane. 1997. Effect of ground covers on grapevine growth and vineyard microclimate. Proceedings of the Plant and Soil Conference, American Society of Agronomy, California Chapter, Visalia, CA.
Daane, K.M. and M.J. Costello. 1994. An assessment of leafhoppers and their natural enemies in the Lodi-Woodbridge winegrape region, with particular reference to within-vineyard cultivation of native grasses. Crop Year 1994 Report to the Lodi-Woodbridge Winegrape Commission.
Gulick, S.H., D.W. Grimes, D.S. Munk and D.A. Goldhammer. 1994. Cover-crop-enhanced water infiltration of a slowly permeable fine sandy loam. Soil Sci. Soc. Am. J. 58: 1539-1546.
Flaherty, D., C. Lynn, F. Jensen and M. Hoy. 1971. Influence of environment and cultural practices on spider mites in southern San Joaquin Valley Thompson Seedless vineyards. California Agriculture Nov. 1971.
Howe, K. 1994. Lodi's dynamic duo cuts chemicals by 80%. Farmer to Farmer 6: 1-3 & 10.
Ingels, C. and K. Klonsky. 1998. Vineyard cover crops and their uses: Historical and current uses. In C. Ingels, P. Christensen and G. McGourty (eds.), Cover Cropping in Vineyards: A Growers Handbook, pp. 93-106. University of California Division of Agriculture and Natural Resources Publication 3338.
Ingels, C. 1998. Implementation of cover cropping in vineyards: Grower practices. In C. Ingels, P. Christensen and G. McGourty (eds.), Cover Cropping in Vineyards: A Growers Handbook, pp. 93-106. University of California Division of Agriculture and Natural Resources Publication 3338.
Land, Air and Water Resources-Cooperative Extension Joint Infiltration Committee. 1984. Water penetration problems in California soils. Tech. Rep. Dept. Land, Air and Water Resour., University of California, Davis.
Prichard, T. L. Effects of cover cropping on soil and water management: Water use and infiltration. In C. Ingels, P. Christensen and G. McGourty (eds.), Cover Cropping in Vineyards: A Growers Handbook, pp. 93-106. University of California Division of Agriculture and Natural Resources Publication 3338.