CALIFORNIA AGRICULTURE, MAY-JUNE 2000 7

This article and the one to follow reportinvestigations in which scientists havemade a novel application of geographic in-formation systems (GIS) to study agricul-tural expansion on a county scale. Thefirst article describes the use of GIS to mapSonoma County vineyards and identifychanges in the pattern of vineyard expan-sion between 1990 and 1997. The secondarticle describes a model — built by ex-tracting variables from the GIS — whichwas used to identify sites for possible fu-ture vineyard expansion in SonomaCounty. The map and model became im-portant sources of information to SonomaCounty’s diverse interests, as they devel-oped plans and policies to guide the cur-rent wave of vineyard expansion. — ED.

Mapping vineyard expansionprovides informationon agriculture and the environmentAdina M. Merenlender

Vineyards are expanding rapidly inCalifornia’s coastal counties due toa booming wine market. This changein land use has led to public debateover natural habitat removal, over-production of wine grapes, loss ofagricultural diversity, and changingscenery. Using a geographic infor-mation system (GIS) to map andanalyze vineyard expansion inSonoma County, approximately11,600 acres of new vineyards wereidentified from 1990 through 1997.The total acreage was calculated tobe at least 48,000 acres in 1997, 20%more than reported by county agri-cultural officials . Compared to vine-

yards established before 1990, ahigher percentage of vineyardsplanted between 1990 and 1997were located on hillsides thatsupported oak woodlands. Oakwoodlands support a majority ofthe region’s biodiversity, provideecosystem goods and services,and have aesthetic value. Thisresearch was designed to docu-ment the effects of recent vine-yard expansion on SonomaCounty’s hardwood rangelandsand to provide tools for informeddiscussion and decision-makingby land-use planners, farmers,residents and environmentalists.

California wines are popular na-tionally and internationally, result-ing in increased demand for high-quality wine grapes, record-highwine bottle prices and increases inwine-grape planting. Over thepast decade, shipments of winesthat sell for $7 a bottle and up haveincreased 157% (Press Democrat,Nov. 1, 1999). The California wineboom has been an economic bless-ing for many rural landowners,who can reap significant value bydeveloping their land into vine-yards. Money from the Californiawine industry also makes its wayinto many facets of the localeconomy by providing jobs and in-

8 CALIFORNIA AGRICULTURE, VOLUME 54, NUMBER 3

In 1996 the Integrated HardwoodRange Management Program estab-lished a geographic information sys-tems (GIS) lab at the UC Hopland Re-search and Extension Center. GIS is acomputer mapping and analysis toolthat allows the integration of largeamounts of spatial and nonspatial in-formation. We used a GIS to integratea map of Sonoma County’s vineyardswith other digital information.

The digital vineyard maps usedfor this project were originally com-piled by the Sonoma County GrapeGrowers Association and put intodigital format by Circuit Rider Pro-ductions, Inc. to produce high-qualitymaps for tourists and wine enthusi-asts. Vineyards were identified in themajor appellation areas of SonomaCounty using 1990 aerial photo-graphs, and more recent vineyard de-velopments were mapped from infor-mation provided by the grapegrowers themselves through 1997(fig. 1).

Other digital information that wasused includes elevation, slope and as-pect data derived from U.S. Geologi-cal Survey (USGS) Digital ElevationModels (DEM); roads and streamsfrom the U.S. Census Bureau; farm-land and urban areas from the Cali-

fornia Department of ConservationFarmland Mapping and MonitoringProgram; and vegetation from the Cali-fornia Department of Forestry and FireProtection 1990 hardwoods layer. Addi-tional base layers were also obtainedfrom the Sonoma County InformationSystems Department GIS Division.

This integrated GIS was intersectedwith maps showing the location ofvineyards developed before and after1990 with elevation and slope derivedfrom the DEM. We also used this GISto extract all of the variables used inthe vineyard expansion model. Areasthat met the designated regulatory lev-els in the Sonoma County VineyardErosion and Sediment Control Ordi-nance were identified by intersectingexisting vineyards with the DEM andsoils data. To assess the potential fu-ture impact of this ordinance, we alsointersected the areas identified by thevineyard expansion model as beingthe most likely to be suitable for futurevineyard expansion. This GIS holdspromise of being a powerful analysistool that we hope to improve and useto continue this research to examineland-use changes, quantify the envi-ronmental consequences and providepolicy and planning analysis.

— A.M.M.

How the geographic information systemused here was built

creasing tourism. Many local govern-ments promote continued vineyarddevelopment and encourage regionalwine-making to improve economicvitality.

Statewide, wine grape acreage hasapproximately doubled since 1990(California Agricultural StatisticsService 1999). In some places thisnew vineyard acreage is beingplanted in upland areas that supportoak woodlands and forests. Conver-sion of undeveloped land to vine-yards involves the clearing of nativeupland and riparian vegetation. Thistype of conversion has the potentialto affect natural resources — increas-

ing hillside erosion, impacting en-dangered species or impeding wild-life migration. In addition, increasedvineyard development may lead tooverproduction of wine grapes, lossof local agricultural diversity andchanging scenery.

As a first step toward addressingthese issues, we used a geographicinformation system (GIS) to map andquantify the extent of hillside devel-opment and rates of natural habitatconversion (see sidebar, below).Thiseffort demonstrated that publishedstatistics on vineyard acreage are in-complete and could be improved bybetter mapping.

Sonoma County findings

Sonoma County, which containsnine incorporated cities, is in the heartof the North Coast wine country and isone of the fastest growing counties inCalifornia. The U.S. Census Bureau re-ported a population increase of 282%from 1960 to 1995 (Census Bureau1998). In addition to being one ofCalifornia’s premium wine-grapegrowing regions, Sonoma County’sapproximately 1 million acres includea substantial amount of hardwoodrangeland (135,599 acres) and forests(561,468 acres) (ABAG 1996). Addi-tionally, over 90% of the land is in pri-vate ownership.

A GIS was developed to analyze theamount and pattern of vineyard devel-opment in Sonoma County’s major ap-pellation areas, legally recognizedwine-grape growing areas with specificgeographic designations (fig. 1). Wefound that 11,663 acres of new vine-yards were planted from June 1990through June 1997, bringing acreage ingrapes to at least 48,000 acres in 1997.This was 20% higher than the SonomaCounty Agricultural Commissioner’s es-timate at the time. This larger figure wasa surprise to most county residents, andwas reported as front-page news withthe headline, “New Sonoma Countymaps add 8,000 vineyard acres” (PressDemocrat, June 1, 1998).

The GIS data layers used also in-cluded physiographic data such asslope, aspect and elevation (fig. 2). Theintegration of these data allowed us todemonstrate that upland areas are in-creasingly targeted for vineyard devel-opment. For example, in SonomaCounty we estimate that 25% of thevineyards developed since 1990 wereon slopes greater than 10 degrees (18%slope) and 42% were above 328 feet(100 meters) in elevation (figs. 3 and4). By comparison, less than 6% of thevineyards established prior to 1990were on slopes greater than 10 de-grees, and only 18% were above 328feet. There are several reasons that areprobably driving the move upslope:(1) flat land is developed; (2) whenavailable, flat land is more expensive;and (3) the microclimate found on hill-sides can result in higher quality winegrapes.

CALIFORNIA AGRICULTURE, MAY-JUNE 2000 9

Fig. 1. Vineyards mapped within the major appellation areas of Sonoma County, includ-ing Alexander Valley, Russian River Valley, Chalk Hill, Dry Creek Valley, Knights Valley,Los Carneros, Sonoma Mountain and Sonoma Valley.

The vineyards mapped in figure 1were intersected with a 1990 digitalvegetation map, allowing us to esti-mate the amount of undeveloped landconverted to agriculture from 1990through 1997. The 1990 vegetationmap was derived from Thematic Map-per Satellite imagery provided by theCalifornia Department of Forestry andFire Protection and was assumed torepresent the vegetation cover typesthat existed in 1990. The vegetationanalysis showed that approximately1,631 acres of dense hardwood forest(greater than 50% tree cover), 278 acresof conifer, 367 acres of shrubland and7,229 acres of oak grassland savannawere converted to vineyards between1990 and 1997 in Sonoma County. Itshould be noted that these are slightunderestimates, because vineyardsoutside the appellation areas in figure1 had not been mapped and thereforecould not be included in our analysis.

These GIS results demonstrate thatSonoma County has substantiallymore vineyard acreage than previ-ously reported. Vineyards planted inthe last 10 years typically occur athigher elevations and on steeperslopes than earlier plantings, resultingin the conversion of approximately9,505 acres of undeveloped land.

Problems in estimating acreage

Based on our 1997 acreage esti-mates, the Sonoma County Agricul-tural Commissioner reported 51,500acres of vineyards in the 1999 crop re-port (Sonoma County AgriculturalCommissioner 1999). They added anestimated additional acreage of 3,500to the 1997 figure of 48,000, itself 20%greater than the estimate that year.This estimate comes closer to the trueacreage for the county than the 43,316acres reported by California Depart-ment of Food and Agriculture (1998).Likewise, the Santa Barbara PlanningDepartment (1999) estimated har-vested wine-grape acreage at morethan 18,000 acres for 1998, as com-pared to the Fruit and Nut Crop Re-port figure of 10,799 harvested acres(UC Fruit and Nut Research 2000).

Much of the difference betweenlocally derived estimates of har-vested acreage and published esti-

mates is due to the fact that growersmay not report reliably or may notyet be accounted for in the system.The counties with the most rapidvineyard development rates are of-ten the farthest behind in reporting,and consequently have less reliabledata.

Other counties with expanding pre-mium wine-grape acreage include SanLuis Obispo with 14,988 acres reported(CDFA 1998) but actually closer to23,000 planted acres (Deputy Agricul-tural Commissioner Robert Hopkins,personal communication, 1999), andMendocino County with 14,264 acresreported (CDFA 1998) but closer to17,000 actual acres (Agricultural Com-missioner Dave Bengston, personalcommunication 1999). These discrep-ancies in wine-grape acreage figuresindicate that a better system for map-ping and monitoring agriculturalchange is needed. Aerial photographyand other remote sensing technologiescombined with GIS can be used tomap and monitor agricultural areasand land use change as demonstratedfor Sonoma County. This approachprovides more accurate statistics onthe amount of vineyard planted andcould assist growers in determiningwhether current rates of productionmay exceed demand.

Advantages of mapping

Recently, environmental impactsand landscape change due to in-creased vineyard planting have beencontentious issues in California. Fed-eral courts have ruled that althoughlandowners have the title to their land,they do not own the water, wildlife orother resources outright, and are re-sponsible for managing these resourcesfor the public trust (Araiza 1997). Moreinformation is needed to understandthe consequences of vineyard expan-sion on a landscape scale as well aswhat site-specific vineyard develop-ment and management techniques canbe used to minimize negative impactsto agriculture in general, neighboringlandowners and the environment.

Advantages to agriculture. Thewine-grape industry is highly suscep-tible to economic downturns due tolack of information on future demandand how many tons of wine grapeswill be produced when vineyardsplanted today begin bearing fruit (onaverage vines come into full produc-tion after 6 years). Quantifying theamount of bearing and nonbearingacreage is important for farmers, win-eries, financial lenders and land-useplanners to determine grape produc-tion, reduce the chance of over supply,and assess the environmental impacts

10 CALIFORNIA AGRICULTURE, VOLUME 54, NUMBER 3

30,000

25,000

20,000

15,000

10,000

5,000

00-10% 10-20% 20-30% 30-40% 40-50% >50%

83.7%

55.8%

11.6%23.4%

3.4%13.3%0.9% 5.0% 0.3%1.7% 0.1% 0.8%

Existing vineyards (pre-1990) New vineyards (post-1990)

Percent slope class

Nu

mb

er o

f ac

res

30,000

25,000

20,000

15,000

10,000

5,000

00-100 100-200 200-300 300-400 400-500 >500

81.0%

57.8%

13.4%22.6%

2.8% 6.9% 1.5% 6.5% 0.6% 3.8% 0.7% 2.3%

Existing vineyards (pre-1990)New vineyards (post-1990)

Elevation class (meters)

Nu

mb

er o

f ac

res

Fig. 2. Vineyards in Sonoma County with elevation classes. Vineyards mapped from1990 aerial photos are in red; those mapped by the grape growers as newly developedsince 1990 are in blue.

Fig. 3. Percent of total vineyard acreagebefore and after 1990 in slope classes.

Fig. 4. Percent of total vineyard acreagebefore and after 1990 in elevation classes

of more vineyard development. Cur-rently in Sonoma County, premiumwine-grape growers can expect as highas $3,600 net revenue per acre of vine-yard land in full production (PressDemocrat, Nov. 1, 1999). A downturn

in wine-grape revenues could morequickly bankrupt heavily leveragedgrowers than growers without heavydebt loads. However, wine industryexecutives forecast a favorable marketfor at least another 3 years (Wine In-

dustry Financial SymposiumSurvey 1999), in the belief thatincreasing numbers of peoplewill appreciate good-qualitywines (Stuller and Martin 1994)and be willing to pay more than$7 per bottle. Economic optimismin the wine market prevails, de-spite the fact that wine-grapejuice supply fluctuations in thepast have led to large price dropsand plowing up of vineyards,such as occurred in the CentralValley (Newsweek, Nov. 17, 1995).

In addition, the loss of diver-sity in agricultural systems canhave consequences for Califor-nia agriculture. For example, inSonoma County, reductions incrop diversity are due both tomarket value decreases in or-chard commodities such asapples, and to the increasedvalue of wine grapes. Todaythere are no prune orchards andless than 5,000 acres of apples;formerly these were the leadingfarm commodities in thecounty. The average price of

apples was $130/ton in 1999, repre-senting a 17% decline in price since1998. Similarly, in Mendocino Countypear and cherry orchards are beingconverted to wine grapes, and in LakeCounty both walnuts and pears are be-ing replaced.

Moving toward a single agricul-tural crop across the North Coast is atodds with the principles of sustainableagriculture, and could lead to overreli-ance on a single industry. Proponentsof sustainable agriculture promote in-creasing biodiversity in conventionalfarming systems to reduce insect pestsand disease, and to reduce the use ofinsecticides, fungicides, herbicides andfertilizers (Bugg and Pickett 1998). Inpart this can be mitigated by plantinga diverse suite of native cover crops inand around the vineyard.

Addressing residents’ concerns.The expansion of vineyards, especiallyon hillsides, can result in changingscenery and deforestation, which canaffect water resources. Some CentralCoast residents have expressed dissat-isfaction with the landscape after graz-ing land was converted to vineyards(Los Angeles Times, March 23, 1999).Some landowners have also claimedreduced flow from wells after vine-yard development occurred on neigh-boring properties. (Water use, alwaysa contentious issue in California, isusually settled in the courts.)

Habitat conversion. Oak wood-lands — including both sparsely cov-ered rangeland and denser oak forests— have been used primarily for live-stock grazing and wood products butare increasingly targeted for vineyarddevelopment. These natural areasalso support valuable game speciessuch as deer and wild turkey. Prop-erties containing oaks or adjacent tooak woodlands often have highermarket value than similar propertieswithout oaks (Standiford et al. 1987).Ecologically, oak woodlands are themost diverse ecosystems in Califor-nia, providing critical habitat for ap-proximately 2,000 plants, more than100 birds, 60 mammals, 80 amphib-ians and reptiles, and 5,000 insectspecies (Pavlik et al. 1991).

This habitat has already undergonefragmentation due to suburban expan-

CALIFORNIA AGRICULTURE, MAY-JUNE 2000 11

Brochure explores vineyardexpansion issues

Vineyards in an Oak Landscapegives practical information on howto assess land-management deci-sions and balance agricultural pro-duction with natural-resource con-servation. The brochure discussesthe unique values of oak woodlandsin California, the threats facing oakwoodlands and how to care for ex-isting oaks.

Along with information on oaks,the brochure outlines current infor-mation on disease, economic incen-tives and management practicies.To obtain a $4.50 copy of publica-tion number 21577, call DANR pub-lications at (800) 994-8849. —A.M.M. and Julia Crawford

sion, thinning for range improve-ment and firewood harvesting. Infact, over a million acres ofCalifornia’s oak woodlands havebeen converted since 1950 (Bolsinger1988). To complicate matters, severaloak species are not regenerating ad-equately to ensure long-term speciessurvival (Bolsinger 1988). Vineyardexpansion in coastal counties repre-sents one of the latest threats toCalifornia’s coastal oak woodlands.

Although many vineyards areplanted in treeless areas, the conver-sion of woodlands and forests is exten-sive at some sites. The Santa BarbaraCounty Planning Department reportedthat the amount of vineyard land hasdoubled to 20,000 acres since 1996,leading to the loss of more than 2,000oak trees, a larger number than devel-opment of rural subdivisions was re-sponsible for removing in the past 10years. Even in less well known wine-grape-growing areas such as San LuisObispo County, an estimated 2,000acres of rangeland are being convertedto vineyards each year.

Converting oak woodlands to vine-yards has discrete and identifiable ef-fects, including the loss of vegetationcover, displacement of wildlife, soildisturbance and habitat fragmentation(Garrison 2000). The California De-partment of Fish and Game (CDFG)has listed 15 species that may be pref-erentially affected by vineyard devel-opment in coastal California (Garrison2000). In addition to habitat loss andfragmentation, impacts on creeks andwetlands have been the focus of theregulatory agencies. The CDFG listsincreased sedimentation, introductionof nonnative fish species as a result ofincreasing the number of reservoirs,and loss of wetland habitat as issues ofconcern for watersheds with vineyarddevelopment.

Although our research focus hasbeen on oak woodlands, some NorthCoast vineyards have replaced conifer-dominated forests that include red-wood, Douglas fir and tan oak trees.However, very little data on this trendexists for Sonoma County because ithas occurred so recently and is not re-flected in the available maps and sta-tistics on vineyard acreage.

Research needs

Integrating spatial information onagricultural areas with other informa-tion about the landscape such as to-pography, hydrography and vegeta-tion allows us to determine whichwatersheds may be disproportionatelyaffected by vineyard development.

As we have demonstrated here, GISis the best tool to do this, however,there are many areas in need of im-provement. We would like to improveour mapping capabilities across alarger geographic area so that we canmonitor the North Coast’s rapidlychanging landscape. An essential im-provement would be to use remotesensing data for the mapping andmonitoring of vineyard expansionacross a larger geographic region andto assess changes in development pat-terns at regular time intervals.

Additional information on soils, cli-mate, geology and water would alsobe of great benefit in analyzing vine-yard development patterns. Higherresolution topographic information isbecoming available and will improveour analysis capabilities as well. Acombination of these improvementsand further analysis of the relation-ships between land use and environ-mental conditions should prove valu-able to those interested in agriculturalland-use planning and environmentalprotection.

A.M. Merenlender is Cooperative Exten-sion Specialist, Integrated HardwoodRange Management Program, Depart-ment of Environmental Science, Policyand Management, UC Berkeley. UC Sus-tainable Agriculture Research and Exten-sion Program funded this research. TheSonoma County Grape Growers Associa-tion, Circuit Riders Production, Inc.,Glenn McGourty, Robert Hopkins, DaveBengston and Abe Leider provided valu-

A vineyard on the edge of an oak wood-land in Sonoma County, where oaks aregradually being removed and replacedwith grapevines.

▼

12 CALIFORNIA AGRICULTURE, VOLUME 54, NUMBER 3

Modeling vineyard expansion,potential habitat fragmentationEmily Heaton ❏ Adina M. Merenlender

We used a statistical modelingtechnique called logistic regres-sion analysis, and a geographicinformation system (GIS), to mapareas of possible future vineyardexpansion in Sonoma County,based on data about vineyard de-velopment from 1990 through1997. The goal of this researchwas to develop a model thatwould improve our understandingof vineyard expansion patterns ata landscape scale (for instance,including an entire county). Theapproach involved identifyinglandscape characteristics thatwere associated with vineyard de-velopment and mapping the areaswith similar characteristics thatwere undeveloped in 1997. Weused the results to map wherehabitat removal and fragmentationcould result from vineyard expan-sion. This method, although stillunder development, is designedfor county- or regional-scaleanalysis to assist land-use plan-ners, natural resource protectionagencies and land conservationprograms in protecting valuableenvironmental resources whilesustaining a vital agriculturaleconomy.

While recognizing that the wine in-dustry represents a significant portionof California’s agricultural economy —

$1.3 billion was paid for Californiawine grapes in 1999 (CASS 2000) — itis also important to preserve the eco-logical integrity of the natural re-sources upon which we all depend forservices ranging from basic sustenanceto aesthetic pleasure. Mapping andanalyzing land use at the county scalecan facilitate community efforts to bal-ance economic growth and environ-mental protection. Land-use analysisand models can also aid planners inprioritizing undeveloped and agricul-tural land for possible protectionthrough conservation easements andother methods.

Recently, vineyard development isexpanding into rangelands, wood-lands and forestlands (forests that in-clude commercially harvested andregulated species such as Douglas fir),resulting in deforestation. We used re-sults of our model to examine the po-tential for habitat loss and fragmenta-tion. Model results can also be used toexamine the possible effects of newregulations, such as those that restrictfuture vineyard development on steepslopes (see sidebar, p. 19).

Modeling approach

Our analyses have sought to iden-tify areas that are most susceptible tofuture vineyard development inSonoma County. Our primary as-sumption is that areas that share im-portant characteristics with recentlyplanted land are more likely to be de-veloped into vineyards in the future.The results of our modeling efforts arestrongly influenced by the develop-ment trends mapped for SonomaCounty from June 1990 through June1997. The model does not specify sitesthat will necessarily support the suc-cessful propagation of wine grapes,since some newer vineyards may havebeen planted in less productive places.Furthermore, site-specific characteris-tics, such as microclimate and some

This is the first time that areas poten-tially suited to agriculture have beenmodeled using GIS for an entirecounty in California. Predictions madeby the model are derived from the cur-rently available data, which does notinclude such information as microcli-mate and soil profiles for specific sites.Scientists hope to add more informa-tion in the future. — Ed.

able data. Colin Brooks worked on thecomputer mapping and analysis. CharlesCooke, David Newburn, Emily Heaton,Kerry Heise and two anonymous review-ers provided valuable comments on themanuscript.

References[ABAG] Association of Bay Area Gov-

ernments. 1996. Existing Land Use in 1995:Data for San Francisco Bay Area Cities andCounties. Oakland, CA. 200 p.

Araiza WD. 1997. Democracy, distrust,and the public trust: Process-based consti-tutional theory, the public trust doctrine, andthe search for a substantive environmentalvalue. UCLA Law Review, 1997 Dec, v. 45N2:385-452.

Bolsinger CL. 1988. The Hardwoods ofCalifornia’s Timberlands, Woodlands, andSavannas. US Forest Service PNW-RB-148. 148 p.

Bugg RL, Pickett CH. 1998. Habitatmanagement to enhance biological control:A concept and its applications. In: PickettCH, Bugg RL (eds.). Enhancing BiologicalControl: Habitat Management to PromoteNatural Enemies of Agricultural Pests. Ber-keley, CA: UC Press. p 1-23.

California Agricultural Statistics Service.1999. May 25, 1999 California Grape Acre-age Report. Calif. Department of Food andAgriculture.

[CDFA] California Department of Foodand Agriculture. 1998. California GrapeAcreage Annual Reports.

Census Bureau. 1998. Estimates of thePopulation of Counties by Age, Sex andRace/Hispanic Origin: 1990–1997. Wash-ington DC.

Garrison, B. A. 2000. Draft Report: AStrategy for Conserving Oak Woodlands inVineyard Landscapes. California Dept. ofFish and Game. 36 p.

Pavlik BM, Muick PC, Johnson S, Pop-per M. 1991. Oaks of California. Los Olivos,CA: Cachuma Press. 184 p.

Santa Barbara Planning Department.1999. Santa Barbara Vineyard AcreageReport.

Sonoma County Agricultural Commis-sioner. 1999. Sonoma County AgriculturalCrop Report 1999.

Standiford RB, Diamond N, Passof PC,LeBlanc J. 1987. Value of oaks in rural sub-divisions. In: Proceedings of the Sympo-sium on Multiple Use of California’s Hard-wood Resources. USDA Forest ServiceGen. Tech. Report PSW-100. p 156-60.

Stuller J, Martin G. 1994. Through theGrape Vine: The Real Story BehindAmerica’s $8 Billion Wine Industry. NewYork: HarperCollins. 365 p.

UC Fruit & Nut Research and Informa-tion Center. 2000. Fruit and Nut Crop Re-port. Department of Pomology, University ofCalifornia.

Wine Industry Financial SymposiumSurvey. 1999. Universal Capital Corpora-tion Vineyard Group, Napa, CA.