Indirect Measurement of Leaf Area Index in California North Coast Vineyards

The performance of the LI-COR LAI-2000 Plant Canopy Analyzer (PCA) for indirect measurement of leaf area index (LAI) was evaluated in vineyards of Californiaʼs North Coast region. Twelve plots were established, representing vineyards of differing trellis, cultivar, and planting density. Mean LAI ranged from 0.5to 2.25-m2 leaf area per m2 ground area by direct measurement (defoliation). Indirect LAI derived by a standard two-azimuth, diagonal-transect measurement protocol was significantly related to direct LAI (r = 0.78, P ≤ 0.001). However, the PCA underestimated direct LAI by about a factor of two. Narrowing the instrumentʼs conical field of view from 148° to 56° increased indirect LAI by 13% to 60% in vertically trained plots, but still resulted in substantial underestimation of direct values. Use of this PCA protocol in vineyards should therefore be accompanied by direct measurement for calibration purposes. Leaf area index (LAI; one-sided leaf area normalized by ground area) is a common descriptor of vegetation canopy structure that is related to light interception, productivity, and evapotranspiration (Welles and Norman, 1991). Within vineyards, LAI can be used to monitor carbon balance and assess fruit quality potential (Ollat et al., 1998). A number of methods have been developed for LAI measurement (Welles, 1990). The most direct measure, as used in this study, involves physical removal of leaves from the plant and derivation of their cumulative area with an electronic meter. While accurate on a per plant basis, this approach is destructive and labor intensive. Viticulturists often make management decisions on the basis of dormant period pruning weights (prior seasonʼs wood production), which are empirically related to LAI (Johnson et al., 2003). Optical instruments such as the LI-2000 Plant Canopy Analyzer (PCA) (LI-COR, Inc., Lincoln, Nebr.) are designed for rapid, nondestructive LAI estimation in agricultural and natural plant canopies. Remote sensing offers a means to vastly increase spatial coverage with respect to that achievable by ground-based methods alone. Airborne imagery has been used to map relative LAI differences within individual vineyards (Hall et al., 2002; Johnson et al., 2001; Wildman et al., 1983). Commercial winegrape growers in Californiaʼs North Coast viticultural region (Napa, Sonoma, Lake, and Mendocino counties) are using digital imagery for various purposes, such as harvest preparation, vineyard redevelopment, and identification of problems related to irrigation, nutrition, disease, and pest infestation (Carothers, 2000). More recently, multispectral aircraft and satellite imagery has been used to map vineyard canopy density in absolute terms as leaf area (m) per vine or per meter of row (Dobrowski et al., 2002; Johnson et al., 2003). For greatest accuracy, such maps require collection of supporting, ground-based LAI data. Prior studies have used several measurement protocols to estimate vineyard LAI with varying degrees of success (Grantz and Williams, 1993; Ollat et al., 1998; Patakas and Noitsakis, 1999; Sommer and Lang, 1994). The current study was conducted to evaluate use of a standard two-azimuth, diagonal-transect, PCA protocol for LAI estimation in commercial winegrape vineyards of Californiaʼs North Coast. A “whole farm” research approach (National Research Council, 1997) was pursued whereby LAI differences were driven by actual farming practices and grower choices concerning planting density, trellis system, and cultivar. Materials and Methods Study sites.Study plots were located within the Tokalon and Huichica “ranches” of the Robert Mondavi Winery (Oakville, Calif.). The 500-ha Tokalon ranch is located in the Napa Valley (lat. 38°25 ́N, long. 122°25 ́W) and planted mainly to red grape cultivars on clay loam soils. The 300-ha Huichica ranch is located 22 km south-southeast of Tokalon in the Carneros region (lat. 38°14 ́N, long. 122°22 ́W). Huichica produces red and white grape cultivars on clay soils with variable topography. Six plots were established within separate vineyards at each ranch (Table 1). Of these 12 plots, seven plot canopies were vertically trained, one was a split canopy, and five were untrained. Vertical canopy systems essentially produce “walls” of leaves and may be thought of as having an “I” cross-section, while split canopies have a “Y” cross-section (Fig. 1). Plot centers were situated at least 15 m from the nearest edge to avoid contamination by light scattered laterally into the sensor field-of-view from outside the vineyard. Indirect LAI.The PCA calculates LAI from observations made with a “fish eye” optical sensor as fully described in Welles and Norman (1991) and LI-COR (1992). Briefly, the Table 1. Study plots located at Robert Mondavi Vineyards (Oakville, Calif.). Leaf area index (LAI) by direct measurement (defoliation) also shown. Age Vine space Row space LAI Plot Trellis Cultivar (years) (m) (m) (m·m)