Effects of cover crops on a vineyard ecosystem in the Northern San Joaquin Valley
Date of Report: September 2001
Principal Investigator:
Chuck Ingels
UC Cooperative Extension
4145 Branch Center Road
Sacramento, CA 95827-3898
Tel: (916) 875-6913
Fax: (916) 875-6233
E-mail: caingels@ucdavis.edu
Co-investigators & Cooperators:
Kate Scow - Land, Air & Water Resources Dept., UC Davis
Desley Whisson - Wildlife, Fish & Conservation Biology Dept., UC Davis
Terry Prichard - Land, Air & Water Resources Dept., UC Davis
Andy Johas and Mahinder Dhaliwal (grower cooperators), Deer Creek Vineyards,
Johas & Associates Inc., Sacramento
Location of Project: Sheldon CA, Sacramento County
Commodity: Grape
Funding:
|
Matching Funds
|
|||
|
Year
|
SAREP | Amount | Source |
|
1998
|
$4,062 | $0 | |
|
1999
|
$6,032 | $7,032 | Lodi-Woodbridge Winegrape Commission |
|
2000
|
$6,032 | $6,800 | Lodi-Woodbridge Winegrape Commission |
Objectives
- To determine the effects of several popular cover crop mixes on grapevine
shoot growth, water stress, nutrient status, and fruit production and quality
- To evaluate the biomass production and nitrogen content of cover crop mixes
- To evaluate the effects of cover crops on soil microbial ecology
- To determine the economics of different cover crop systems
Summary
This trial was conducted in a Merlot vineyard in Sacramento County, planted in 1993. It consisted of 5 treatments and 4 replications in a randomized complete block design. The mixes tested were: a California native perennial grass mix, a disked green manure mix, a nontillage annual reseeding clover mix, a disked cereal mix, and a disked control. The native grass was planted in the fall of 1996 and the others were planted in 1997; annual mixes were planted again in the fall of 1998 and 1999.
Cover crop biomass differed among treatments and years, but the native grass mix suppressed weed growth the most in 1999. Large numbers of gopher mounds were found in the clover mix and very few or none were found in the other mixes. Cover crops had greater soil microbial biomass than non-cover cropped soils. The annual clover and native grass mixes supported higher microbial biomass than did the control, bell bean/vetch, and barley/oat mix.
Cover crops had little or no effect on yield or juice quality. The native grass mix, and to a lesser extent the clover mix, reduced shoot growth somewhat, which could eventually affect quality (positively or negatively). Water potential was lower (stress was greater) in the native grass and clover treatments in 1998 only. In a blind tasting, wines from cover cropped treatments were preferred over the disked control.
The presence of cover crops increased the size of soil microbial biomass relative to that in the disked control. Microbial biomass was also much greater in the cover-cropped portion of the vineyards than in the berms. The annual clover and native grass mixes supported higher microbial biomass than did the control, green manure mix, and cereal mix.
In the first year, disking without cover crops is less expensive than cover cropping. However, in subsequent years, the native grass and clover mixes are cheaper to manage because they require only one or two mowings. However, these mixes are more expensive if they require replanting, become invaded with gophers or weeds, or if they excessively reduce vine vigor.
Specific Results
The trial was conducted in a Merlot vineyard on 5BB rootstock, planted in 1993. Deer Creek Vineyard is located on Grant Line Road near the town of Sheldon in Sacramento County. The spacing is 7 x 11 ft. and the soil type is San Joaquin silt loam. Vines are trained to a bilateral cordon system with a standard T-trellis. The vineyard is drip irrigated, with sprinklers available for frost protection and cover crop germination. Drip irrigation and fertigation were applied uniformly across all treatments. Except where nontillage cover crops were used, the vineyard floor was managed by disking the row middles in the spring and using both pre- and post-emergence herbicides in 4-5 ft. strips on the berms.
The experiment consisted of 5 treatments and 4 replications in a randomized complete block design. The native grass mix was planted in the fall of 1996. Other mixes were planted in the fall of 1997, and annual mixes were also planted in 1998 and 1999. The mixes tested were:
- Native perennial grass mix: Calif. barley / meadow barley / Calif.
brome / blue wildrye - 25 lbs./planted acre; 40 lbs. N/planted acre; nontillage;
mowed as needed in the spring or once in early spring and again after reseeding
- Green manure mix: Bell bean / 'Magnus' pea / common vetch / barley
- 80 lbs./planted acre; no fertilizer; planted every late October and mowed
and disked in April or early May
- Annual reseeding clover mix: Subclover / burr medic / crimson clover
/ rose clover - 25 lbs./planted acre; nontillage; no fertilizer; mowed once
in early spring and again after reseeding
- Cereal mixture: Barley / oats - 100 lbs./planted acre; 40 lbs. N/planted
acre/year, planted every fall, mowed once in early spring, then mowed and
disked in late spring
- Disked control - disked periodically, no fertilizer
Each replication consisted of two adjacent middles, and neighboring replications were separated by one disked middle. Most plant and soil measurements were made from or adjacent to 10 contiguous vines.
- To determine the effects of several popular cover crop mixes on grapevine shoot growth, water stress, nutrient status, and fruit production and quality
Vine Parameters. Throughout the experiment, pruning weights in the native grass treatment were significantly less and the disked control often greater than those in other treatments (Table 1). A similar trend was seen in shoot lengths (1999 only) and summer trimming weights (data not shown). All cover crops, and particularly the native grass, appeared to reduce vigor to some extent. Red wine quality in vineyards with excessive irrigation and/or vigor is usually poorer than where growth is more in balance with fruit production. To the extent that cover crops reduce vigor and moderately stress vines, wine quality may be improved. However, if the stress is excessive, yields may decrease.
Grapevine petiole nitrate levels in most years were highly inconsistent and often far outside normal ranges (data not shown). However, 2001 bloom petiole nitrate levels were within a normal range; both the petiole nitrate and blade total N content of the native grass and clover treatments were lowest and that of the green manure mix was far higher (Table 2). (During the 2000-2001 season, only the unfertilized native grass mix was allowed to grow; all other rows were disked.). High-nitrogen green manure mixes have the potential to add far too much nitrogen, which may impair wine quality. Conversely, grasses may stress vines by outcompeting the vines for nitrogen.
Water potential measurements were taken in 1998 and 2000. Before the first irrigation in 1998, water potentials were slightly lower (under greater stress) in the native grass and clover treatments than in the disked control. However, water potentials did not differ in 2000 in May or June (data not shown).
Yields and Juice and Wine Quality. There were no significant differences in yield in any year of the study (Figure 1). There were also no consistent differences in juice soluble solids, pH, or titratable acidity (Table 3).
To evaluate wine quality, 50 lbs. of grapes were picked from each replication and combined for each treatment. Small wine lots were made by E&J Gallo in Modesto, and the wines were crushed without oak. The wine was blind-tasted by 11 individuals, including Gallo winemakers, viticulturists, and research staff, as well as LWWC and UC personnel. Each taster was asked to rank the five wines based on their personal preference. Because the tasting was not replicated, it is not possible to state conclusively whether there were significant differences between the wines.
Wine lots from the annual cover crop mixes were found to be similar in flavor and "mouthfeel," with wine from the green manure mix having the best fruit, the softest mouthfeel and the best balance of all the treatments (Table 4). The wine from the native grass treatment was found to be slightly thin and somewhat out of balance by some, although it was ranked first by two tasters. The disked control was noticeably thin and out of balance. However, while it was the least preferred, it was not unacceptable; none of the wines had acute vegetal taste or reduction.
Improved wine quality with the use of cover crops has been reported anecdotally (Ingels, 1998). In a four-year trial conducted in Lodi, an annual ryegrass cover crop had a devigorating effect compared to a clean cultivated vineyard floor; the effect was cumulative with little effect the first two years (T. Prichard, pers. comm.). The ryegrass also caused a significant decrease in petiole nitrogen. The devigoration was due to a synergism of both water deficits and nitrogen level. The effect on wine, when compared to a like water deficit treatment without a cover crop, was to increase tannins. Tasters who preferred tannins or were looking for a blending wine preferred the cover crop treatment wine.
- To evaluate the biomass production and nitrogen content of cover crop mixes
There were substantial differences in biomass of the cover crops among treatments and among years. In 1998, the wet El Niño spring prevented mowing until early May, after measurements were taken. In that year, the bell bean/vetch mix produced the greatest biomass, growing to about 4.5 ft. tall (Table 5). The native grass and cereal mixes grew to about 4 ft., and the native grass mix produced large quantities of seed. The clovers grew very little until March and consequently produced the least biomass, with a final height of 26 in. for the crimson clover and 13 in. for the other clovers. In 1999 and 2000, the native grass, clover, and cereal mixes were mowed to about 1 ft., about a month before measurements were taken, so these biomass values reflect the regrowth after mowing.
In the native grass mix, weed biomass was relatively high in 1998, but declined to very low levels in 1999 and 2000 (Table 6). The winter growth and spring regrowth after mowing of this mix was vigorous and dense. Weed growth in the green manure mix was relatively low in 1998 but highest in 1999. Because the clovers are allowed to reseed, grass weeds also reseeded, so by 2000, weed biomass was higher in this treatment.
The nitrogen content of annual clovers was similar that of the green manure mix, both on a percent basis and a per-acre basis (Table 7). Because the green manure mix produced far more biomass than the clover mix in 1998, the per-acre N contribution was substantially greater (194 vs. 132 lbs. N/acre). However, a portion of N contained in the clover clippings may be lost to ammonia volatilization since it is not disked. The N content of the native grass mix was consistently greater than that of the cereal mix, showing that it has a greater N demand than the cereals. The potassium content of the clovers was highest and that of the cereal mix was lowest (Fig. 2). The phosphorus content of the mixes did not differ significantly (data not shown).
Gophers were present in high numbers in 1999 only. Gopher activity was measured by counting mounds in the adjacent middles of each treatment replication for the entire 175-vine row. In each of the three months that gophers were evaluated, they showed a distinct preference for the clover cover crop (Fig. 3). No fresh activity was observed in the cereal cover crop or the disked treatment, and there was very little activity in the native grass and green manure mixes.
- To evaluate the effects of cover crops on soil microbial ecology
Background and Approach
Cover crops have been shown to increase microbial biomass and alter community composition in a variety of agronomic systems. These changes in microbial communities, in turn, can affect soil structure and nutrient cycling. One approach for evaluating changes in microbial communities is phospholipid fatty acid analysis (PLFA) of whole communities extracted directly from soils. PLFAs are integral components of cell membranes and rapidly metabolized when a cell dies in soil; therefore, they provide a measurement of living organisms. Principle types of PLFAs are defined on the basis of chain length, degree of unsaturation, and presence of constituents (e.g., methyls, hydroxyls, cylopropane rings). There are three ways in which PLFA data can be used to provide information about microbial communities: 1) total PLFA provides a measure of viable microbial biomass, 2) the entire PLFA profile can be used as a "fingerprint" of the soil community; and 3) signature lipids can be used to detect specific subgroups within the community: e.g., sulfate reducers, methane oxidizing bacteria, fungi, and actinomycetes.
Soil samples were taken from the cover crop areas and from the berms with a 2 cm corer to a depth of 15cm. Data are reported as the mean of three lab replicates (except where the third lab rep was thrown out due to high variation). These samples were analyzed for microbial community size and composition using phospholipid fatty acid (PLFA) analysis. PLFAs were extracted from the whole soil samples, fractionated, methylated, and analyzed by gas chromatography. The data were analyzed using multivariate statistics to determine relationships between the samples and to identify which fatty acids contributed to the observed relationships.
Results
Total Biomass. The total amount of PLFA (reported in nm/g dry soil) is an indicator of total microbial biomass present in the samples (plus a minor amount of plant biomass). The average total PLFA was significantly higher in the cover crop samples than in the berm samples for all treatments (Table 8). Total PLFA was highest in the clover and native grass treatments (Table 9). Average total PLFA in the disked control was the same in the berm samples (18.2) as it was in the cover crop samples (18.0). This indicates that the increased microbial biomass in the cover cropped samples is a result of the planted cover crop rather than being an artifact of location or of disking.
Microbial Community Composition. PLFA fingerprints, each of which is made up of fatty acids contributed by the dominant members of a soil's microbial community, were compared among the different soils. To compare fingerprints requires use of a multivariate statistical technique, called Correspondence Analysis (CA), which was performed on a subset of the total number of fatty acids detected in all soil samples. CA is a data analysis technique that transforms a data set containing many variables (in this case fatty acids) into a smaller set of new variables, or dimensions.
CAs of each different sampling date show that differences in microbial communities between the berm and cover crop samples is greater than differences associated with different cover crop treatments (data not shown). The differentiation is likely associated with differences in rhizosphere of cover crops versus grapevines, and due to the fact that berm samples may have had better access to irrigation water than did the cover crop samples. Differences in rhizosphere could lead to differences in soil moisture, carbon and other nutrients, competition for nutrients with plants, as well as other factors.
Among the different cover crops, the greatest differences were associated with the annual clover and native grass samples. Microbial communities associated with these two cover crops were remarkably similar to each other and grouped independently from the other cover crop samples. This similarity could be due to the fact that neither of these two treatments was disked.
Specific groups of microorganisms. Another way of looking at the PLFA data is to compare the mass of specific lipid biomarkers that are indicators of particular groups of microorganisms. Tables 8 and 9 show the relative proportions of biomarkers for some major groups of organisms in soil, as well as the average number of lipids and the total PLFA extracted. These biomarkers include the proportion of branched (Gram positive bacteria marker), cyclopropyl (aerobic bacteria markers), linoleic acid (fungal biomarker), and methyl branched (gram positive bacteria marker) for subgroups of the samples. Values shown are averages of berm and cover crop region samples from the four different sample dates (Table 8) and of samples from berm and cover crop regions of the five different cover crop treatments (Table 9). (Note: biomarker data need to be interpreted with caution; not all biomarkers are exclusive to a particular group and it is possible that some members of a particular group may not have that biomarker).
For all cover crop treatments, the proportion of aerobic bacterial markers is higher by 6 to10 percent in the cover crop samples than in the corresponding berm samples (Table 8). Conversely, the fungal biomarker component is always higher in the berm samples than in the corresponding cover crop samples. This trend is not caused by the presence of a cover crop (which possibly could contain this lipid), since the disked control also has a much higher percentage of fungal biomarker in the berm than cover crop samples (Table 9). However, some weeds did grow in the disked treatment, whereas none grew in the berms. There is no apparent difference in lipids between berm and cover crop region samples.
The proportion of fungal biomarker is consistently higher in the spring data sets than in summer data sets (Table 8). The markers for gram positive and aerobic bacteria show no apparent seasonal trends.
- To determine the economics of different cover crop systems
In the first year, the use of the native grass is substantially more expensive than the other mixes because of the cost of seed (Table 10). The cost of maintaining native grasses in future years is relatively low - mainly requiring mowing and extra fertilizer (and possibly extra water). The cost of annual clovers is lowest in the long term, assuming that the vines do not require nitrogen fertilizer and that the stand does not decline and require replanting in future years (which it often does). Seed for the two annually sown cover crops used in this trial are relatively inexpensive, however, the requirements for repeated disking make these mixes costly each year. Overall, planting seeds and repeated disking are the costliest operations involved in using cover crops.
Potential Benefits/Impacts on Agriculture
This project has shown that, under drip irrigation and a fairly wide herbicide strip, cover crops have little if any effect on fruit yield. However, consistent with other studies and with the experiences of several growers and winemakers, wine quality may be improved with cover crops. This potential benefit is increasingly important during years when wine grape supply exceeds demand and wineries preferentially use grapes that produce the highest quality wine.
This study also showed that growers should be cautious with the use of clovers, to which pocket gophers were strongly attracted. However, no effect on the vines was seen.
Soil microbial populations were shown to increase with any cover crop, particularly native perennial grasses and nontillage annual clovers. Improved soil quality is the cornerstone of sustainable agriculture.
Other benefits, such as improved pest management, water infiltration, or wheel traction were not studied. Previous research on the effects of cover crops on vineyard pest management have shown mixed results, although a comprehensive Lodi survey showed that vineyards with perennial cover crops had lower pest mite populations (C. Ohmart, unpublished data).
There were large differences in cost of various cover crop mixes both in the first year and beyond. Native grasses are very expensive to plant, but are relatively cheap to maintain in future years. Reseeding clover mixes are very inexpensive after the first year. Cereal and green manure mixes are relatively inexpensive to plant, but they must be planted each year and the soil must be cultivated periodically.
Dissemination of Findings
On April 14, 1999 a field meeting was held at the trial site, which was sponsored by UCCE, LWWC, and SAREP. Discussion topics at the meeting included effects on the vines, effects on vineyard pest management, and the preference of gophers and voles on the various cover crop mixes (attendance 60).
On November 30, 1999 I presented results of the trial at the Fertilizer Research & Education Program's annual conference in Modesto (see attached report from the proceedings).
On December 14, 1999 I also presented these results at a breakfast meeting in Lodi that was sponsored by the Lodi-Woodbridge Winegrape Commission (attendance 55).
A summary of the research was published in the August 2001 issue of the Lodi-Woodbridge Winegrape Commission's IPM/Research Newsletter.
These research results will be submitted for publication in California Agriculture.
Literature Cited
References
Klonsky, K., L. Tourte, and C. Ingels. 1992. Sample costs to produce organic wine grapes in the North Coast. UC Davis Department of Agricultural and Resource Economics.
Verdegaal, P, K. Klonsky, and P. Livingston. 1994. Sample costs to establish a vineyard and produce wine grapes. UC Davis Department of Agricultural and Resource Economics.
- Native grass mix, June 1997 (planted October 1996)
- Native grass mix, May 1998
- Native grass clippings, July 1998
- Native grass mix after mowing, 1999
- Green manure mix, March 1998
- Green manure mix finally being mowed, May 1998 (El Niño)
- Annual clover mix, with crimson clover predominating in photo, April 1999
- Gopher mounds in annual clover mix, February 1999
- Mowed barley cover crop, April 1999
- Disked control treatment, March 1999