See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/232690499

Effects of Organic and Conventional Practices on Weed Control in a Perennial

Cropping System

Article in Weed Science · September 2009

DOI: 10.1614/WS-06-171 · Source: OAI

CITATIONS READS
41 236

3 authors, including:

Kerri Steenwerth Lissa Veilleux

USDA/ARS at Davis, CA & University of California, Davis 4 PUBLICATIONS 143 CITATIONS
50 PUBLICATIONS 3,855 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Kerri Steenwerth on 13 June 2014.

The user has requested enhancement of the downloaded file.

Weed Science 2007 55:352–358

Effects of Organic and Conventional Practices on Weed Control in a Perennial

Cropping System
Kendra Baumgartner, Kerri L. Steenwerth, and Lissa Veilleux*
Vineyard weed communities were examined under the influence of an organic weed control practice, soil cultivation with
a Clemens cultivator, and applications of the herbicide glyphosate. Experimental treatments (winter–spring glyphosate,
spring cultivation, fall–spring cultivation, fall cultivation–spring glyphosate) were carried out in a California wine grape
vineyard for 3 yr. Cultivation alone was not as effective as glyphosate, based on lower weed biomass in the glyphosate-only
treatment in 2 of 3 yr. However, given that two passes with the Clemens cultivator decreased weed biomass relative to one
pass, it is possible that additional passes could bring about further reductions. Pairing fall cultivation with glyphosate was as
effective at reducing weed biomass as two glyphosate applications in 2 of 3 years, suggesting that substituting a glyphosate
application with cultivation may be an effective method of reducing herbicide use in vineyards. Canonical correspondence
analysis revealed significant treatment effects on community structure. Weed composition in the spring cultivation
treatment was significantly different from that of all other treatments. Based on our findings of high relative abundance of
field bindweed and sowthistle species, which are problematic vineyard weeds that grow into the vine canopy and disrupt
canopy management practices, it is possible that either the presence of soil disturbance or the absence of herbicides favored
these species.
Nomenclature: Glyphosate; annual sowthistle, Sonchus oleraceus L. SONAL; field bindweed, Convolvulus arvensis L.
CONAR; spiny sowthistle, Sonchus asper (L.) Hill SONAS; wine grape, Vitis vinifera L. ‘Merlot’.
Key words: Clemens cultivator, perennial cropping system, tillage, vineyard, weed community.

Organic wine grapes are gaining popularity among wine- IWM in California vineyards typically involves the in-
makers and the public in Europe (Willer and Zanoli 2000). tegration of postemergence and preemergence herbicides
Their popularity in the United States is evidenced by the fact (Agamalian 1992), with less emphasis on incorporation of
that grapes represented 10% of 2002 organic commodity sales nonchemical methods. This practice reflects, in part, the
in California (Klonsky 2004), which produces 88% of U.S. paucity of published research on weed control in organic
grapes (Anonymous 2006b). U.S. acreage of organic vineyards vineyards. The few examples of weed research that pertain to
has increased substantially over the past 15 yr, and currently vineyards focus on herbicides (Kadir and Al-Khatib 2006;
represents 1.5% of the total grape acreage, 90% of which is Monteiro and Moreira 2004). Research on IWM in vineyards
located in California (Green and Kremen 2003). Rising and other perennial cropping systems lags far behind that of
acreage of organic vineyards in the United States may be annual systems (e.g., Cardina et al. 2002; Legere et al. 2005;
driven, in part, by passage of more stringent water-quality Menalled et al. 2001; Shrestha et al. 2002). As such, there is
regulations in California (Anonymous 2006a). These higher a need for research on nonchemical practices for vineyards, to
standards for water quality mitigate pollution from agricul- minimize the negative impacts of wine grape production on
tural runoff by restricting pesticide use, thereby forcing public-trust resources.
growers to use different pesticides, to limit pesticide The aim of this research was to compare the organic weed
applications, or to adopt organic practices. control practice, soil cultivation, to the conventional practice,
A growing list of herbicide-resistant weeds (Basu et al. applications of the herbicide glyphosate, in a perennial
2004) makes it clear that repeated use of a single tactic for pest cropping system in northern California. Investigation of
control not only leads to a preponderance of the most glyphosate and cultivation within a wine grape production
problematic species, but can fundamentally shift the genetic system is warranted because the deficit irrigation and
composition of their populations. Integrated weed manage- fertilization practices used purposely to devigorate the vines,
ment (IWM) emphasizes the use of multiple tactics to address as low yields are associated with high wine quality, are unique
the causes of weed problems, rather than simply reacting to among cropping systems. Infrequent irrigation and the typical
weed infestations (Buhler 2002). Cardina et al. (1999) outline absence of precipitation during the grapevine growing season
various levels of IWM, which start with individual weed mean that weed growth is restricted primarily to winter and
control practices and progress to the integration of practices. spring, which allows for minimal weed control attempts.
Liebman and Gallandt (1997) also emphasize a multistrategy Nonetheless, weed establishment is minimized on the
approach and its incorporation into the cropping system, vineyard floor beneath the vines in order to prevent weed
given that weeds are responsive not only to weed control, but shoots from growing into the vine canopy, where they
interfere with the numerous, labor-intensive, canopy man-
also to numerous facets of crop production.
agement practices. Objectives were (1) to evaluate the efficacy
of the practices in reducing weed biomass; (2) to characterize
the weed community; (3) to monitor vine yield, growth, and
DOI: 10.1614/WS-06-171.1 nutrition under the influence of the practices; and (4) to
* First and third authors: United States Department of Agriculture— determine the effects of the practices on soil biological
Agricultural Research Service, Department of Plant Pathology, University of activity. Our intent in monitoring vine and soil parameters
California, One Shields Avenue, Davis, CA 95616; second author: United States
Department of Agriculture—Agricultural Research Service, Department of
was to identify effective weed control practices that can be
Viticulture and Enology, University of California, One Shields Avenue, Davis, integrated into the cropping system without impacting wine
CA 95616. Corresponding author’s E-mail: [email protected] grape production.

352 N Weed Science 55, July–August 2007

Table 1. Weed control practices associated with experimental treatments.
Practice dates
Treatment Practice 2003 2004 2005
Fall–spring cultivation Cultivation November 27, 2002 November 24, 2003 December 3, 2004
Cultivation May 16, 2003 April 20, 2004 May 7, 2005
Fall cultivation–spring glyphosate Cultivation November 27, 2002 November 24, 2003 December 3, 2004
Glyphosate (5.6 kg ai ha21) May 22, 2003 April 27, 2004 May 13, 2005
Spring cultivation Cultivation May 16, 2003 April 20, 2004 May 7, 2005
Winter–spring glyphosate Glyphosate (2.8 kg ai ha21) February 22, 2003 January 31, 2004 February 11, 2005
Glyphosate (5.6 kg ai ha21) May 22, 2003 April 27, 2004 May 13, 2005

Materials and Methods biomass was timed in between the last weed control practices
and the start of the dry season, and was based on visual
The experiment was conducted in a commercial wine grape observation of peak weed height (June 4, 2003; May 12,
vineyard in the Napa Valley of northern California from 2003 2004; and May 31, 2005). Weed biomass was collected from
to 2005. The vineyard was established in 1996 with Merlot four randomly placed, 0.6-m2 quadrats per treatment per
(clone 314) on 110R rootstock (V. berlandieri Planch. 3 V. block (two at the base of vine trunks, two between adjacent
rupestris Scheele). Vine spacing was 1.8 by 1.8 m, with east– vines), to give a total of 80 quadrats per year. Positioning half
west row orientation. Vines were trained as unilateral cordons of the quadrats at the base of the vine trunks accommodated
to a vertical shoot positioning trellis system. The 0.84-m-wide the fact that the Clemens cultivator is directed away from this
section of soil in the vineyard row, where treatments were section of the vineyard floor so as to avoid damage to
carried out, was level with the soil in between the rows grapevine roots. Weeds were sorted by species, dried (70 C,
(vineyard middles); vines were not elevated on berms. The 7 d), and weighed. Volunteer grape seedlings and cover crop
vineyard was on Bale soil (fine–loamy, mixed, thermic seedlings were considered as weeds. Our use of species applies
Cumulic Ultic Haploxeroll). to more than one species in the cases of filaree species
There were four treatments: winter–spring glyphosate, (Erodium sp.) and sowthistle species (Sonchus sp.). Several
spring cultivation, fall–spring cultivation, and fall cultivation– quadrats with plants that shared characteristics of more than
spring glyphosate (Table 1). Glyphosate1 was applied with
a tractor-mounted, 1.2-m-wide, boom sprayer with two fan-
type nozzles directed beneath the vines on both sides of the
tractor. Cultivations were done with a Radius Weeder2
(Clemens cultivator), which consists of a 0.3 by 0.1-m metal
blade positioned perpendicular to the direction of tractor
movement. When inserted slightly below the soil surface, it
severs weed shoots from their roots. An automatic articulating
arm directs the cultivator around vine trunks and trellis system
posts. Because the Clemens cultivator mounts to one side of
the tractor, each cultivation required two passes per row.
Glyphosate is a common herbicide in wine grape vineyards,
and is typically applied twice per season (once at budbreak,
once after removing trunk suckers in late spring). Cultivation
is a common weed control practice in organic vineyards,
where the use of pesticides is forbidden. Frequency of
cultivation varies depending on the type of cultivator, but is
typically infrequent in summer, as the resulting clouds of dust
settle on the leaves, leading to spider mite infestations.
Treatments were arranged in a randomized complete block
design with five blocks (0.27 ha per block). Weed control
practices were applied to three adjacent vineyard rows; data
were collected from the center row. A no-till cover crop of
zorro fescue (Vulpia myuros var. hirsuta Hack.) was main-
tained in the vineyard middles. The cover crop was reseed-
ed in October 2002 with a seed drill (10 kg ha21) and
mowed every June. Temperature and precipitation were
recorded by the nearest California Irrigation Management
Information System (CIMIS) weather station (Oakville
Station No. 77; Figure 1).
We anticipated that a combination of infrequent, drip
irrigation at the study site (85 kl ha21 applied once per week,
July to October) and rare precipitation during the growing
season would restrict informative weed measurements to early Figure 1. Average monthly rainfall and air temperature during study years (2003
in the growing season. Collection of aboveground weed to 2005) and 15-yr averages (1991 to 2005).

Baumgartner et al.: Vineyard weed practices N 353

one of three Erodium species—broadleaf filaree [Erodium
botrys (Cav.) Bertol.], redstem filaree [E. cicutarium (L.)
L’Her. ex Ait.], and whitestem filaree [E. moschatum (L.)
L’Her. ex Ait.]—necessitated combining all Erodium biomass
measurements into Erodium sp. The same situation applied to
Sonchus plants that shared characteristics of annual sowthistle
and spiny sowthistle.
Petioles and soil for analyses of mineral composition were
collected at full bloom (June 5, 2003; June 1, 2004; and May
26, 2005). Within each replicate row, 100 petioles were
collected by a standard sampling procedure (Winkler et al.
1965), pooled, dried (70 C, 7 d), ground, and analyzed for
total nitrogen (N), total phosphorus (P), total potassium (K),
zinc (Zn), and boron (B) (DANR Laboratories, University of
California, Davis, CA). Soil samples were collected from four
random locations with a 4.6-cm-diameter auger to a depth of
15 cm, pooled, dried (70 C, 7 d), ground, and analyzed for
NH4-N, NO3-N, exchangeable P (X-P), exchangeable K (X-
K), exchangeable sodium (X-Na), exchangeable calcium (X-
Ca), exchangeable magnesium (X-Mg), cation exchange
capacity (CEC), organic matter (OM), and pH. To determine
cumulative effects of treatments on soil biological activity, we Figure 2. Total weed biomass for the treatment by year interaction. Error bars
measured net nitrification and N mineralization, potential N represent 95% confidence intervals for mean total weed biomass; treatment means
mineralization, and potential microbial respiration (Robertson with overlapping confidence intervals are not significantly different. Treatment
et al. 1999) on May 26, 2005. Fruit clusters were harvested abbreviations are as follows: fall–spring cultivation (F–SC), fall cultivation-spring
glyphosate (FC–SG), spring cultivation (SC), and winter-spring glyphosate (W–
(September 19, 2003; September 28, 2004; and October 25, SG), respectively.
2005) from six adjacent vines per row. Dormant canes were
weighed (November 27, 2003; November 22, 2004; and
December 12, 2005) from the same vines. CANOCO,4 with axis scores centered to interspecies distances
Analyses of variance (ANOVAs) were used to determine the and biplot scaling (Leps and Smilauer 2003). Automatic
effects of treatment and year on total weed biomass; vine forward selection with Monte Carlo permutation tests was
mineral nutrients; soil chemical, physical, and biological used to determine the significance of the treatments.
properties; grape yields; and pruning weights. Biomass from Treatment centroids and canonical coefficients for the species
the four quadrats per treatment per block were averaged. are presented in biplots. Proximity of a species score to
ANOVAs were performed with the use of the MIXED a treatment centroid signifies that the species had the highest
procedure in SAS,3 with Kenward-Roger as the denominator relative abundance in that treatment.
degrees-of-freedom method (Littell et al. 1996). Year was
considered a repeated measure, block and block interactions
were random effects, and treatment, year, and treatment by Results and Discussion
year were fixed effects. To satisfy the assumption of
homogeneity of variance, the following transformations were Treatment Efficacy. Total weed biomass varied significantly
applied: log10 transformations to total weed biomass, petiole among treatments, but relative differences were not consistent
Zn, soil NH4-N, soil NO3-N, and potential microbial among years (treatment by year interaction significant at P ,
respiration; square-root transformations to vine pruning 0.0001). In 2003, mean value comparisons for each treatment
weights and soil X-Na; reciprocal square-root transformations indicated that winter–spring glyphosate was most effective in
to soil X-K, net N mineralization, and net change in soil N reducing weed biomass (Figure 2). In 2005, both glyphosate
pools; and rank transformations to soil X-Ca, soil X-Mg, soil treatments, winter–spring glyphosate and fall cultivation–
CEC, and soil pH. For main or interaction effects that were spring glyphosate, were equally effective and had significantly
significant (P , 0.05), differences among treatment means lower weed biomass than cultivation alone. In 2004, however,
were assessed by comparison of 95% confidence intervals, all treatments were equally effective, except for fall cultiva-
such that means without overlapping intervals were consid- tion–spring glyphosate. Total weed biomass declined by 10-
ered significantly different (Westfall et al. 1999). Reverse- fold from 2003 to 2004, and remained relatively low in 2005
transformed means and 95% confidence limits are presented, (Figure 2). Lower weed biomass in all treatments in 2004 may
for ease of interpretation. be attributable to a combination of low precipitation (88 cm
Canonical correspondence analysis (CCA) was used to in October 2003 to June 2004) and unseasonably high
evaluate treatment effects on weed community structure (ter temperatures in late winter (Figure 1). Based on mean value
Braak 1987). Analysis was based on aboveground biomass of comparisons, this was followed by statistically significant
weed species present in $ 8% of the samples. Species omitted increases, albeit slight, in weed biomass in 2005, but only for
from the analysis were present in fewer than 4 of 60 total fall–spring cultivation and spring cultivation (Figure 2); weed
samples collected over the course of the entire experiment and biomass in the glyphosate treatments remained low. Regard-
were, thus, uninformative. Treatments were treated as less, all treatments had 10-fold lower weed biomass in 2005
independent variables, species biomass as dependent variables, than in 2003, despite similar rainfall in both years (112 and
and years and blocks as covariables. CCA was performed in 113 cm, respectively; Figure 1).

354 N Weed Science 55, July–August 2007

Table 2. Weed species in a California vineyard at peak biomass in 2003–2005.
Species presence (+)/ absence (2)
Fall cultivation–spring Winter–spring
Fall–spring cultivation glyphosate Spring cultivation glyphosate
Species Common name 2003 2004 2005 2003 2004 2005 2003 2004 2005 2003 2004 2005
Anagallis arvensis L. Scarlet pimpernel + 2 2 2 + 2 + 2 + 2 2 +
Brassica rapa L. Birdsrape mustard + 2 2 2 2 + 2 2 + 2 2 2
Bromus carinatus H. & A. California brome 2 2 2 2 2 2 2 2 + 2 2 2
Bromus hordeaceus L. Soft brome 2 2 + 2 2 + 2 2 2 2 2 2
Calendula arvensis L. Field marigold 2 2 2 2 2 2 + 2 + 2 2 +
Chenopodium album L. Common lambsquarters 2 + 2 2 2 + + 2 2 2 2 2
Convolvulus arvensis L. Field bindweed 2 + 2 2 + 2 2 + + 2 2 2
Cyperus sp. Sedge 2 2 2 2 2 2 2 2 2 2 2 +
Epilobium brachycarpum C. Presl Panicle willowherb + 2 + + + + + 2 2 + + 2
Erodium sp. Filaree + 2 + 2 + + + 2 + + + +
Festuca idahoensis Elmer Idaho fescue 2 2 2 2 2 2 2 + 2 2 2 +
Geranium carolinianum L. Carolina geranium + 2 + 2 + + 2 + + 2 + +
Gnaphalium purpureum L. Purple cudweed + 2 2 2 2 2 2 2 2 2 2 2
Hordeum brachyantherum Nevski Meadow barley 2 2 + 2 2 2 2 2 + 2 2 2
Kickxia spuria (L.) Dumort. Female fluvellin + 2 + 2 + + 2 + + 2 2 2
Lactuca serriola L. Prickly lettuce 2 + 2 2 2 2 2 2 + 2 2 2
Lythrum hyssopifolia L. Loosestrife 2 2 + 2 + 2 + 2 2 2 2 2
Medicago polymorpha L. California burclover 2 2 + 2 2 2 2 2 + 2 2 +
Picris echioides L. Bristly oxtongue 2 2 2 2 2 2 2 2 + 2 2 2
Plantago lancelota L. Buckhorn plantain 2 2 2 2 2 2 2 2 + 2 2 2
Raphanus raphanistrum L. Wild radish 2 + 2 2 + + 2 2 2 2 2 2
Rumex crispus L. Curly dock 2 2 2 2 + + + 2 + + + +
Senecio vulgaris L. Common groundsel 2 2 2 2 2 2 2 2 2 2 + +
Sonchus sp. Sowthistle + 2 + 2 + + + 2 + 2 2 +
Veronica persica Poir. Persian speedwell 2 2 2 2 2 2 2 2 + 2 2 2
Vitis vinifera L. ‘Merlot’ Volunteer grape + 2 + + + + 2 2 + 2 + +
Vulpia myuros var. hirsuta Hack. Volunteer zorro fescue + + + + + + + + + + + +

Zorro fescue was especially dominant in 2003 (Figure 2), position at the origin of the biplot (Figure 3). The three most
8 mo after seeding the vineyard middles with this cover crop. common species in 2003, 2004, and 2005, respectively, were
Significant annual changes in total weed biomass may be (followed by ranges of relative proportions of biomass per
related, in part, to germination of the cover crop in the rows. block in parentheses): zorro fescue, panicle willowherb
Zorro fescue is a strong competitor in California’s annual (Epilobium brachycarpum C. Presl), and sowthistle (Sonchus
grasslands, due to its rapid germination after the first rains and sp.) (75 to 100%); zorro fescue, Carolina geranium
its ruderal nature (Brown and Rice 2000). Our finding of low (Geranium carolinianum L.), and curly dock (Rumex crispus
weed biomass in all treatments in 2005, despite high rainfall, L.) (24 to 63%); and filaree (Erodium sp.), Carolina
suggests that either the climate of the 2004 rainy season had geranium, and sowthistle (15 to 65%).
persistent impacts on subsequent weed establishment or that The weed community associated with spring cultivation
the dwindling biomass of zorro fescue reduced its contribu- was distinct from that of all other treatments, based on its
tion to total weed biomass over the course of the study. opposite position on axis 1, which was the only axis that
Our finding of low weed biomass with two glyphosate significantly explained community differences among treat-
applications per year indicates that this herbicide is more ments (P 5 0.05; Figure 3). Spring cultivation had the
effective than cultivation at reducing total weed biomass. highest relative abundances of California burclover and
However, given that two passes with the Clemens cultivator sowthistle species. Scarlet pimpernel (Anagallis arvensis L.)
further decreased weed biomass relative to one pass (Figure 2), and field bindweed were also most abundant in spring
it is possible that the level of control achieved with two cultivation, albeit at much lower biomass than California
glyphosate applications may be matched through additional burclover and sowthistle species. Scarlet pimpernel was
cultivation passes. Pairing fall cultivation with glyphosate was relatively abundant in both spring cultivation and fall–spring
as effective at reducing weed biomass as two glyphosate cultivation, which is reflected by the proximity of this species’
applications in 2004 and 2005, suggesting that substituting biplot score to both treatments.
a glyphosate application with cultivation, instead of using two Based on the high relative abundance of scarlet pimpernel,
glyphosate applications per year, may be an effective method field bindweed, and sowthistle species resulting from spring
of reducing herbicide use in vineyards. cultivation in 2 of 3 study years, it is possible that either the
presence of soil disturbance or the absence of herbicides shifts
Weed Communities. CCA revealed significant community the vineyard weed community to these species. Field
differences among treatments. The species present in the bindweed and sowthistle species are considered problematic
communities fell into one of three categories: (1) ubiquitous in vineyards because they grow into the vine canopy and
among treatments (e.g., zorro fescue); (2) sporadically present interfere with harvest (Lanini and Bendixen 1992). It is
in a given treatment [e.g., California brome (Bromus carinatus possible that the high relative abundance of field bindweed in
H. & A.)]; or (3) dominant in certain treatments [e.g., cultivated rows is due to dispersal of its rhizomes by the
California burclover (Medicago polymorpha L.)]. Zorro fescue Clemens cultivator, unlike in previous studies that reported
was present in all treatments and years (Table 2); hence its a decrease in this species’ frequency with deeper cultivation by

Baumgartner et al.: Vineyard weed practices N 355

 Although treatment centroids from both fall cultivation
treatments grouped close to meadow barley (Figure 3), this
species was absent from fall cultivation–spring glyphosate
(Table 2).
Based on high relative abundances of filaree species and
curly dock in winter–spring glyphosate in all study years
(Figure 3), it is possible that repeated glyphosate use shifts the
vineyard weed community to these species. Of the three, curly
dock is most problematic in vineyards because of its perennial
nature and tall shoots (Lanini and Bendixen 1992). Given that
filaree and curly dock germinate in the rainy season
(DiTomaso and Healy 2007), their dominance in the
glyphosate-only treatment, coupled with their contrasting
low biomass in fall cultivation–spring glyphosate, suggest that
they were well established before the winter application of
glyphosate in the glyphosate-only treatment. Our findings are
consistent with those of Young (2004), who found that
broadleaf filaree and curly dock were only partially controlled
by glyphosate a week after treatment, possibly due to their
dominant tap roots, which may enhance their tolerance to
glyphosate and/or increase their susceptibility to damage by
the Clemens cultivator.
There was a significant effect of the treatment by year
interaction on weed species richness (P 5 0.02), but mean
Figure 3. Species–treatment biplot from canonical correspondence analysis of comparisons showed no significant differences among treat-
weed communities in the four treatments. The circle, triangle, square, and ments (data not shown). Neither species diversity nor evenness
diamond represent fall–spring cultivation (F–SC), fall cultivation–spring
glyphosate (FC–SG), spring cultivation (SC), and winter–spring glyphosate varied significantly among treatments (P 5 0.7 and P 5 0.07,
(W-SG), respectively. Bayer codes represent the following species: Scarlet respectively) or years (P 5 0.07 and P 5 0.5, respectively).
pimpernel (ANGAR), field bindweed (CONAR), panicle willowherb (EPIPC), We expected the treatments to influence species diversity,
filaree species (EROSP), Carolina geranium (GERCA), meadow barley richness, and evenness differentially, given that herbicides
(HORBR), female fluvellin (KICSP), California burclover (MEDPO), curly
dock (RUMCR), sowthistle species (SONSP), volunteer grape (VINVI), and have been shown to have greater effects on diversity than
volunteer zorro fescue (VLPMY). Axis 1: P 5 0.05, l 5 0.21, species– tillage in other systems (Legere and Samson 2004; VanGessel
environment correlation 5 0.75; axis 2: P 5 0.14, l 5 0.15, species– et al. 2004). The lack of significant treatment effects may be
environment correlation 5 0.61 (sum of all canonical eigenvalues is 0.38). attributed to the low number of weed control attempts in our
treatments. Also, the Clemens cultivator is less physically
disruptive to soil than tillage implements (e.g., mouldboard or
mouldboard plows that likely buried the rhizomes (Froud- chisel plows) used in other studies (Critchley et al. 2006;
Williams 1988). Our finding of a high relative abundance of Legere et al. 2005).
sowthistle species in spring-cultivated rows and a concomitant
low relative abundance in the glyphosate-only treatment is Impacts on Production. There were no significant differ-
supported by similar results from annual cropping systems by ences in yield due to treatment (P 5 0.1), year (P 5 0.5), or
Critchley et al. (2006) and Puricelli and Tuesca (2005). their interaction (P 5 0.4). Average yield across years and
Sensitivity of California burclover to glyphosate has also been treatments was 6.5 kg vine21 (n 5 12). Pruning weights
documented (Wallace et al. 1998). varied significantly among years (P 5 0.003), with the highest
The weed community associated with fall–spring cultiva- measured in 2005, at 0.7 kg vine21 (n 5 4), and the lowest in
tion was more similar to that of fall cultivation–spring 2004, at 0.6 kg vine21 (n 5 4). There were no significant
glyphosate than to spring cultivation (Figure 3). Fall differences in pruning weights due to treatment (P 5 0.1) or
cultivation was associated with high relative abundances of the treatment by year interaction (P 5 0.7).
panicle willowherb and volunteer grape, in the years these two Petiole K was the only mineral nutrient that was
species were present (Table 2), regardless of the type of significantly affected by the treatments, based on ANOVA
practice (cultivation or glyphosate) that followed in spring. (Table 3). Vines in fall–spring cultivation had the lowest
Female fluvellin [Kickxia spuria (L.) Dumort.] and volunteer concentrations of total K compared with that of winter–spring
grape were also associated with fall cultivation, albeit at much glyphosate (15.5 mg vs. 17.4 mg). However, overlapping
lower biomass than panicle willowherb. Panicle willowherb is 95% confidence intervals among all four treatment means
a problematic vineyard weed because of its height. In contrast, signified that they were not significantly different (data not
volunteer grape rarely becomes established in California shown). Petiole N, K, and Zn varied significantly among
vineyards, which is likely due to their susceptibility to years, but not treatments (Table 3). Changes in petiole P and
Daktulosphaira vitifoliae (Fitch) (grape phylloxera), a wide- B over time were not consistent among treatments, hence the
spread pest that necessitates grafting wine grape cultivars on significant treatment by year interaction, but means compar-
phylloxera-resistant rootstocks (Granett et al. 2001). Meadow isons following ANOVA showed no significant differences
barley (Hordeum brachyantherum Nevski), present only in among treatments within years (data not shown). Annual
2005 (Table 2), was most abundant in fall–spring cultivation. means averaged across treatments (n 5 4 per year) ranged

356 N Weed Science 55, July–August 2007

Table 3. Analysis of variance of grapevine (petiole) and soil mineral nutrition treatment, it seems that incorporation of weed biomass
parameters, and soil physical properties under the influence of four weed control
treatments, from 2003–2005.
influenced changes in soil N availability, thereby enhancing
soil biological processes. No change in soil microbial
P values respiration due to cultivation was observed, despite past
Parameters Treatment (T) Year (Y) T by Y reports of negative effects of tillage (Calderón et al. 2000).
Petiole nutrients This may be due to the shallow soil disturbance from the
Total N 0.4349 , 0.0001 0.0988 Clemens cultivator, whereas more intensive tillage implements
Total P 0.5865 , 0.0001 0.0173 bring about significant reductions in microbial respiration
Total K 0.0304 , 0.0001 0.0565 (Franzluebbers et al. 1999).
B 0.5598 , 0.0001 0.0187
Zn 0.4985 , 0.0001 0.1276
Soil nutrients Management Implications. Given that vine yield, growth,
NH4-N 0.0122 , 0.0001 0.3564 and nutrition were unaffected by the high weed biomass in the
NO3-N 0.2487 , 0.0001 0.5302 low-efficacy, spring cultivation treatment, it seems that weed
Olsen P 0.9403 0.5533 0.4200 growth poses a minor threat to wine grape yields, at least in
X-K 0.4279 0.4643 0.9532
X-Na 0.3927 , 0.0001 0.5240 the drip-irrigated, northern California vineyard we examined.
X-Ca 0.9527 0.1143 0.1384 However, this treatment was associated with the highest
X-Mg 0.7790 0.1897 0.0128 relative abundance of field bindweed and sowthistle species.
Soil physical properties These are problematic vineyard weeds due to their habit of
CEC 0.1700 0.1967 0.0630 growing into the vine canopy, where their presence slows
OM 0.1303 0.7445 0.4272 canopy management practices and harvest, all of which are
pH 0.6596 0.0740 0.7931
typically done by hand. With respect to weed community
composition, our findings of differential species responses to
the treatments are consistent with those of past research in
from 8 to 11 mg total N, 7 to 8 mg total P, 13 to 20 mg total annual cropping systems, which show that frequency, timing,
K, 41 to 45 mg B, and 91 to 364 mg Zn g21 dry petiole. and tolerance of cultivation or herbicides interact to have
Soil NH4-N was the only soil parameter that was varied effects on weed species (e.g., Baylis 2000; Critchley et
significantly affected by the treatments, based on ANOVA al. 2006; Legere and Samson 2004; Poggio 2005). Although
(Table 3). Soil NH4-N in spring cultivation was highest, at we limited our assessment of weed biomass and composition
7.5 mg, compared to the lowest concentration of 5.8 mg in fall to the end of the rainy season, when vineyard weeds are most
cultivation–spring glyphosate. However, overlapping 95% abundant, there is a need for additional work on seasonal
confidence intervals among all four treatment means signified weed community dynamics. The lack of treatment effects on
that they were not significantly different (data not shown). yield, growth, and mineral nutrition parameters in all 3 study
Soil NO3-N, NH4-N, and X-Na varied significantly among years suggests that changes in weed biomass or species
composition that may have occurred after we sampled weed
years, but not treatments (Table 3). Changes in X-Mg over
biomass did not impact production. Nonetheless, detailed
time were not consistent among treatments, hence the
knowledge of when the most problematic vineyard weeds
significant treatment by year interaction, but means compar-
become established and senesce, information that is not
isons following ANOVA showed no significant differences currently available, is crucial for developing more sustainable
among treatments within years (data not shown). Annual weed control practices.
means averaged across treatments (n 5 4 per year) ranged
from 4 to 13 mg NH4-N, 3 to 10 mg NO3-N, 20 to 22 mg
Olsen P, 7.4 to 7.8 mmol X-K, 1.8 to 2.2 mmol X-Na, 130 to
138 mmol X-Ca, 122 to 127 mmol X-Mg, 378 to 389 mmol Sources of Materials
cation exchange capacity, and 23.8 to 24.1 mg organic matter 1
Roundup UltraMax, Monsanto Company, 800 North Lind-
g21 dry soil, and 5.7 to 5.9 pH. ANOVAs of soil microbial bergh Boulevard, St. Louis, MO 63167.
activity, assessed in 2005, showed no treatment effects on net 2
Radius Weeder, Clemens GmbH & Co. KG, Rudolf-Diesel-
nitrification and N mineralization, potential N mineraliza- Strasse 8, 54516 Wittlich, Germany.
tion, or potential microbial respiration (data not shown). 3
SAS Version 8.2 statistical software, SAS Institute, Inc., SAS
Potential N mineralization also tended to be higher in the Campus Drive, Cary, NC 27513.
cultivation treatments (15 to 34 mg NH4-N g21 7 d21) versus 4
CANOCO Version 4.5 statistical software, Plant Research
winter–spring glyphosate (8 mg NH4-N g21 7 d21). International, P.O. box 16, 6700 AA Wageningen, The Nether-
Given that neither vine yield nor growth were affected by lands.
the treatments, it seems that one or two Clemens cultivations
per year are unlikely to harm grapevines, at least under soil
conditions similar to that of our study site. Lower-petiole K in Acknowledgments
cultivated rows was within adequate levels (. 15 mg g21 at
bloom; Christensen et al. 1978). Higher NH4-N in spring We thank Dr. M. Rejmanek (Department of Evolution and
cultivated soils was likely due to the short time (1–3 wk) Ecology, University of California, Davis) for canonical correspon-
dence analysis advice and Dr. B. Mackey (Pacific West Area
between cultivation and sampling; we detected the ephemeral Statistician, USDA-ARS, Albany, CA) for advice on experimental
increase in soil NH4-N that follows incorporation of plant design and means comparisons. We also thank M. Neal, M. Klug,
material (Jackson 2000). Given that treatments with cultiva- and D. Bosch for use of the field site and for carrying out all
tion had both higher weed biomass and displayed trends vineyard practices. Dr. A. Shrestha (University of California
toward higher soil microbial activity than the glyphosate-only Cooperative Extension, Parlier, CA) and Dr. R. L. Bugg (Sustainable

Baumgartner et al.: Vineyard weed practices N 357

Agriculture Research and Extension Program, University of Kadir, S. and K. Al-Khatib. 2006. Weed control in grape after fall and spring
California, Davis) provided helpful comments on this manuscript. application of selected herbicides. Weed Technol. 20:74–80.
A. C. Euliss, X. Cheng, J. G. Warren, B. Craughwell, and S. Klonsky, K. 2004. Organic Agricultural Production in California. Pages 241–256
Schnabel helped with field sampling. Funding was provided by in J. Siebert, ed. California Agriculture: Dimensions and Issues. Oakland, CA:
Giannini Foundation, University of California Press.
USDA-ARS. Lanini, W.T. and W.E. Bendixen. 1992. Characteristics of Important Vineyard
Weeds. Pages 321–325 in D.L. Flaherty, L.P. Christensen, W.T. Lanini, J.J.
Marois, P.A. Phillips, and L.T. Wilson, eds. Grape Pest Management. 2nd ed.
Literature Cited Oakland, CA: University of California, Division of Agriculture and Natural
Resources, Publication No. 3343.
Agamalian, H.S. 1992. Vegetation Management Guidelines. Pages 326–330 in Legere, A. and N. Samson. 2004. Tillage and weed management effects on weeds
D.L. Flaherty, L.P. Christensen, W.T. Lanini, J.J. Marois, P.A. Phillips, and in barley–red clover cropping systems. Weed Sci. 52:881–885.
L.T. Wilson, eds. Grape Pest Management. 2nd ed. Oakland, CA: University Legere, A., K.L. Stevenson, and D.L. Benoit. 2005. Diversity and assembly of
of California, Division of Agriculture and Natural Resources, Publication weed communities: contrasting responses across cropping systems. Weed Res.
No. 3343. 45:303–315.
Anonymous. 2006a. The California Water Plan Update. Sacramento, CA: Leps, J. and P. Smilauer. 2003. Multivariate Analysis of Ecological Data using
California Department of Water Resources Bulletin 160-05. http://www. CANOCO. Cambridge, United Kingdom: Cambridge University Press.
waterplan.water.ca.gov. Accessed September 28, 2006. Liebman, M. and E.R. Gallandt. 1997. Many Little Hammers: Ecological
Anonymous. 2006b. Noncitrus Fruits and Nuts 2005 Summary. Washington, Approaches for Management of Crop–Weed Interactions. Pages 291–343 in
DC: United States Department of Agriculture, Agricultural Statistics Service. L.E. Jackson, ed. Ecology in Agriculture. San Diego: Academic.
Basu, C., M.D. Halfill, T.C. Mueller, and C.N.J. Stewart. 2004. Weed genomics: Littell, R.C., G.A. Milliken, W.W. Stroup, and R.D. Wolfinger. 1996. SAS
new tool to understand weed biology. Trends Plant Sci. 9:391–398. System for Mixed Models. Cary, NC: SAS Institute.
Baylis, A.D. 2000. Why glyphosate is a global herbicide: strengths, weaknesses Menalled, F.D., K.L. Gross, and M. Hammond. 2001. Weed aboveground and
and prospects. Pest Manag. Sci. 56:299–308. seedbank community responses to agricultural management systems. Ecol.
Brown, C.S. and K.J. Rice. 2000. The mark of Zorro: Effects of the exotic annual Appl. 11:1586–1601.
grass Vulpia myuros on California native perennial grasses. Restor. Ecol. Monteiro, A. and I. Moreira. 2004. Reduced rates of residual and post-emergence
8:10–17. herbicides for weed control in vineyards. Weed Res. 44:117–128.
Buhler, D.D. 2002. Challenges and opportunities for integrated weed Poggio, S.L. 2005. Structure of weed communities occurring in monoculture and
management. Weed Sci. 50:273–280. intercropping of field pea and barley. Agric. Ecosyst. Environ. 109:48–58.
Calderón, F.J., L.E. Jackson, K.M. Scow, and D.E. Rolston. 2000. Microbial Puricelli, E. and D. Tuesca. 2005. Weed density and diversity under glyphosate-
responses to simulated tillage in cultivated and uncultivated soils. Soil Biol. resistant crop sequences. Crop Protect. 24:533–542.
Biochem. 32:1547–1559. Robertson, G.P., D. Wedin, P.M. Groffman, J.M. Blair, E.A. Holland, K.J.
Cardina, J., C.P. Herms, and D.J. Doohan. 2002. Crop rotation and tillage Nadelhoffer, and D. Harris. 1999. Soil Carbon Availability: Nitrogen
system effects on weed seedbanks. Weed Sci. 50:448–460. Mineralization, Nitrification, and Soil Respiration potentials. Pages 258–271
Cardina, J., T.M. Webster, C.P. Herms, and E.E. Regnier. 1999. Development in G.P. Robertson, D.C. Coleman, C.S. Bledsoe, and P. Sollins, eds. Standard
of weed IPM: levels of integration for weed management. J. Crop Prod. Soil Methods for Long-Term Ecological Research. New York: Oxford
2:239–267. University Press.
Christensen, L.P., A.N. Kasimatis, and F.L. Jensen. 1978. Grapevine Nutrition Shrestha, A., S.Z. Knezevic, R.C. Roy, B.R. Ball-Coelho, and C.J. Swanton.
and Fertilization in the San Joaquin Valley. Oakland, CA: University of 2002. Effect of tillage, cover crop and crop rotation on the composition of
California, Agriculture and Natural Resources. weed flora in a sandy soil. Weed Res. 42:76–87.
Critchley, C.N.R., J.A. Fowbert, and A.J. Sherwood. 2006. The effects of annual ter Braak, C.J.F. 1987. Ordination. Pages 91–169 in R.H.G. Jongman, C.J.F. t.
cultivation on plant community composition of uncropped arable field Braak, and O.F.R. v. Tongeren, eds. Data Analysis in Community and
boundary strips. Agric. Ecosyst. Environ. 113:196–205. Landscape Ecology. Wageningen, The Netherlands: Centre for Agricultural
DiTomaso, J.M. and E.A. Healy. 2007. Weeds of California and Other Western Publishing and Documentation.
States. Volume 2. Geraniaceae–Zygophyllaceae. Oakland, CA: University of VanGessel, M.J., D.R. Forney, M. Conner, S. Sankula, and B.A. Scott. 2004. A
California, Division of Agriculture and Natural Resources, Publication sustainable agriculture project at Chesapeake Farms: a six-year summary of
No. 3488. weed management aspects, yield, and economic return. Weed Sci.
Franzluebbers, A.J., G.W. Langdale, and H.H. Schomberg. 1999. Soil carbon, 52:886–896.
nitrogen, and aggregation in response to type and frequency of tillage. Soil Sci. Wallace, A., R.A. Lancaster, and N.L. Hill. 1998. Application of non-selective
Soc. Am. J. 63:349–355. herbicides during flowering of pasture legumes can reduce seed yield and alter
Froud-Williams, R.J. 1988. Changes in weed flora with different tillage and seed characteristics. Aust. J. Exp. Agric. 38:583–594.
agronomic management systems. Pages 213–236 in M.A. Altieri and M. Westfall, P.H., R.D. Tobias, D. Rom, R.D. Wolfinger, and Y. Hochberg. 1999.
Liebman, eds. Weed Management in Agroecosystems: Ecological Approaches. Multiple Comparisons and Multiple Tests. Cary, NC: SAS Institute.
Boca Raton, FL: CRC. Willer, H. and R. Zanoli. 2000. Organic viticulture in Europe. Pages 23–29 in
Granett, J., M.A. Walker, L. Kocsis, and A.D. Omer. 2001. Biology and H. Willer and U. Meir, eds. Proceedings of the International Congress on
management of grape phylloxera. Ann. Rev. Entomol. 46:387–412. Organic Viticulture. Basel, Switzerland.
Green, C. and A. Kremen. 2003. U.S. Organic Farming in 2000–2001: Adoption Winkler, A.J., J.A. Cook, W.M. Kliewer, and L.A. Lider. 1965. General
of Certified Systems, Agriculture Information Bulletin No. 780. Washington, Viticulture. Berkeley, CA: University of California Press.
DC: United States Department of Agriculture, Economic Research Service. Young, S.L. 2004. Natural product herbicides for control of annual vegetation
http://www.ers.usda.gov/publications/aib780/. Accessed September 28, 2006. along roadsides. Weed Technol. 18:580–587.
Jackson, L.E. 2000. Fates and losses of nitrogen from a nitrogen-15-labeled cover
crop in an intensively managed vegetable system. Soil Sci. Soc. Am. J.
64:1404–1412. Received September 29, 2006, and approved February 14, 2007.

358 N Weed Science 55, July–August 2007