Madrono, Vol. 45, No. 1, pp. 1-11, 1998 560 YEARS OF VEGETATION CHANGE IN THE REGION OF SANTA BARBARA, CALIFORNIA Scott A. Mensing NV Department of Geography, University of Nevada, Reno, 89557 Abstract Pollen evidence from two sites in the Santa Barbara region show evidence of vegetation changes following European settlement in California. In the Santa Barbara coastal region, oak woodland popula- tions (dominated by Quercus agrifolia) remained stable during the pre-European period; however, in the last century woodland densities have increased. At higher elevations along the oak woodland/pine forest ecotone, pines are becoming dominant. Reduction in fire frequency has probably been the main factor contributing to density increases. The pollen record does not show any evidence of an expansion of chaparral over the last 200 years; however, there is weak evidence for an increase in coastal-sage scrub since the early 1800's. The transformation ofthe California grassland appears to have begun particularly early with the invasion of Erodium cicutarium in the region even before the first Spanish settlement in California. With the settlement of Spanish Missionaries, be- 1994). Radiocarbon dates from the top three meters ginning in 1769, the California landscape has been of core however are influenced by old carbon ef- radically altered by human-caused environmental fects which suggest that the bulk sediment dates change. The exact nature of these changes is not may be 1200 years too old. The chronology for the always readily apparent, since descriptions of the historic period therefore was extrapolated from ex- pre-European vegetation are sketchy at best, and by otic pollen types. Oak pollen was only a minor the time reliable botanical records were gathered, component of the pollen sum and is not discussed much of the landscape had already been altered. by the authors. The vegetation we see today represents the dynam- In this paper I use pollen evidence from two sites ic result of two centuries of response to various in the Santa Barbara region to reconstruct vegeta- changes, including changes in fire regime, intro- tion history for the last 560 years. The sites include duction of livestock, invasion of alien species, and the Santa Barbara Channel, just off the coast of land clearing for agriculture and urban develop- Santa Barbara, and Zaca Lake in northern Santa ment. Understanding how vegetation has changed Barbara County (Fig. 1). Oak woodland, chaparral, in response to these impacts can provide valuable coastal-sage scrub, and grassland comprise the ma- information for present-day conservation efforts. jority ofthe vegetation in the region. This area was Ofparticularconcern recently has been the affect one of the earliest settled by Spanish missionaries of human-caused environmental changes in oak and provides an opportunity to identify the effects woodlands (Muick and Bartolome 1987; Bolsinger ofEuropean impacts on vegetation. Repeat photog- 1988). To date, no high resolution paleoecologic raphy is used to help illustrate changes in the last studies have been conducted that document envi- century. ronmental change in regions dominated by Califor- nia oak woodland. A variety of methods have been Study Area used to reconstruct past change in California oak woodlands, including age structure studies (White Santa Barbara Basin. The Santa Barbara Basin 1966; Vankat and Major 1978; Anderson and Pas- (34°11'-34°16'N; 120°01 -120°05'W), is located quinelli 1984; McClaran 1986; McClaran and Bar- between mainland California and the Channel Is- tolome 1989; Mensing 1992), historical records lands (Fig. 1). The center of the basin has a max- (Mayfield 1981; Rossi 1980), analysis of aerial imum depth of ca. 590 m and the bottom waters photographs (Brown and Davis 1991), and resur- are normally anoxic. Because of the absence of a veying of permanent plots (Holzman 1993). These bottom fauna seasonal differences in sediment den- studies have contributed significant information on sity are preserved as varves, and these have been oak woodland history; however, they are generally used to establish a high resolution chronology of restricted to the last 200 years and reveal little the last 560 years (Soutar and Crill 1977; Schim- about the pre-European period. melman et al. 1990). Pollen analysis from an estuarine sediment core The source region for pollen deposited in the on Santa Rosa Island in the Santa Barbara Channel Santa Barbara Basin is the coastal plain and Santa has documented the invasion of exotic taxa and Ynez Mountains. Quercus agrifolia appear domi- changes in native vegetation following the settle- nant on north-facing slopes, in canyons, and in me- ment of the island in the late 1800's (Cole and Liu sic sites along the coastal marine terrace. West and MADRONO 2 [Vol. 45 kilometers Ebtonuoge Fig. 1. Location map of the Santa Barbara Basin core site (SB), and Zaca Lake (ZA). south facing slopes are dominatedby chaparral, and at the turn of the century (Blakely personal com- coastal-sage scrub. Along the coastal plain intro- munication). Ornamental conifers planted near the duced grasses and herbaceous plants are common lake include Sequoia sempervirens and Cedrus de- (Ferren 1985). Pinus muricata grows in the Santa odara (Peterson 1980). Ynez Mountains northwest of Santa Barbara al- though its distribution is restricted (Griffin and Methods Critchfield 1976). Santa Barbara Basin cores SABA 87-1 and 88-1 Zaca Lake. Zaca Lake (34°36'36"N; were recovered by researchers from Scripps Insti- 120°02'17"W, elev. 730 m) is in the San Rafael tute ofOceanography with a Soutar Box Corer and Mountains, within the Los Padres National Forest, a Kasten Corer (Schimmelman et al. 1990). All approximately 50 km northwest of Santa Barbara. cores were initially sampled at near annual resolu- Two massive Quaternary landslides blocked Zaca tion, but the pollen analysis was based on sub-sam- Creek to form the lake (Hall 1981). The lake is ples which represented consecutive, ca. 5 year in- steep-sided and slopes to a flat bottom 13 meters in tervals. An equal weight of sediment was taken depth. Surface area is 6.9 ha. and the maximum from each annual sub-sample. Fifty-nine samples length is 350 m. were analyzed from the period 1425 to 1985. The lake lies at the transition zone between oak Two sediment cores were recovered from Zaca woodland at lower elevations and pine forest at Lake in May, 1992 using a modified square-rod higher elevations. Quercus agrifolia, P. coulteri, P. Livingstone piston corerfitted with a 5 cmdiameter ponderosa, and P. sabiniana are co-dominants plastic tube liner; a 910 cm core (core C) and an around the lake. South-facing slopes are character- overlapping 865 cm core (core D). While still in ized by Ceanothus spp., Arctostaphylo spp., Yucca the plastic tube, cores were X-radiographed at the spp., Salvia spp., and Artemisia californica. Can- University of California Museum of Paleontology yons include small stands of Q. douglasii, Q. chry- to record stratigraphy and density changes. Mag- solepis, and Calocedrus decurrens. Small patches netic susceptibility and Gamma-ray analyses were of exotic pines remain as a legacy of tree planting carried out at the United States Geologic Survey 1998] MENSING: 560 YEARS OF VEGETATION CHANGE 3 laboratories in Menlo Park. Sediment samples (0.5 comes increasingly variable. Beginning in 1870, cc) were then removed for pollen analysis. Quercus steadily increases from 20% to 42%, twice Standard techniques were used to concentrate as high as the average during the pre-European pe- pollen (Faegri and Iversen 1975). A known quantity riod. Pinus remains below 10% through most ofthe of Lycopodium spores was introduced as a control record, and shows virtually no change. Although to calculate absolute pollen concentration and ac- Rhamnaceae and Rosaceae show high variability, cumulation rate (Stockmarr 1971). A minimum of no long term trends appear over the 460 year pe- 400 pollen grains were counted for each level. For riod. Artemisia averages 7% and shows little vari- Zaca Lake, aquatic and riparian pollen types were ability for the first 400 years of the record, then, counted but excluded from the pollen sum. from 1820 to 1985 it increases to an average of 10%. Results Asteraceae averages 20% from 1435 to 1700, but then begins to decline, dropping to only 7% percent Santa Barbara Basin in 1970. Poaceae declines at a slow but fairly con- Chronology. The Santa Barbara Basin varve stant rate through most of the record, but clearly chronology has been corroborated by radiometric increases between 1945 and the present. The Po- dating, cross-correlation with tree-rings and corre- lemoniaceae are primarily insect pollinated, con- lation with hydrological data (Soutar and Crill sequently only small quantities of pollen reach the 1977; Koide et al. 1972; Krishnaswami et al. 1973; Santa Barbara Basin. Polemoniaceae is present in Hulsemann and Emery 1961). The chronology used virtually every level between 1425 and 1795, av- here is that of Schimmelman et al. (1990), and eraging nearly 1% of the pollen sum. In the last Schimmelman et al. (1992). Varve counts were two centuries, Polemoniaceae is commonly absent, made using high quality X-radiographs and age as- Brassicaceae is infrequent prior to 1825, but in- signments were checked against distinctive marker creases substantially in the modern period, most layers of known events such as El Nino periods, likely due to the introduction of European Brassi- floods, and oil spills. The estimated precision ofthe caceae. Erodium first appears in the pollen record time scale is ±1 year for 1900 to 1987, ±2 years in 1755 and is continuously present after that date. from 1900 to 1840, ±5 years from 1840 to 1750, and ±10 years at the 1425 level. Zaca Lake Chronology. The Zaca Lake chronology was de- Taxonomy. Taxonomic nomenclature follows veloped using core D (0-0.5 m depth) and core C Hickman (1993). Forty-eight pollen and spore types m (0.5-2.75 depth). Both cores clearly record a were identified (Mensing 1993). Percentage abun- dance of the nine most important types is shown in complex stratigraphy of laminations, dark silty lay- Figure 2. Quercus probably represents Q. agrifolia, ers, and dense clay layers described in earlier stud- by far the dominant tree species in the region. Ad- ies (Caponigro 1976; Peterson 1980). X-radio- ditional, but less important sources may include Q. graphs were used to correlate core stratigraphy with lobata, which is important in the Santa Ynez drain- that described by Caponigro and Peterson. The chronology was developed using core-to-core cor- age, Q. durata and Q. dumosa which are found in association with chaparral, and Q. tomentella from relation, radiocarbon dating, and the first appear- ance oftwo exotic pollen types {Erodium and Ced- the Channel Islands. Pinus would primarily be P. rus). muricata, P. sabiniana, P. coulteri, and P. ponder- The base of the core section analyzed gave a ra- osa. Following Heusser (1978), the taxonomically diocarbon age of2510 ± 70 BP (Beta-55301) (Cal- dciefafeicualrte cgormobuipneidn.clTuhdiensge tRahxaaminnaccleudaeemaanndyRcohsaap-- endar calibration B.C. 661 ± 150, Stuvier and Rei- aarnroatlhussp,ecRihesamannuds,prAodbeanbloystroempar,eseCnetrctohceagrepnuesr,aPCreu-- amter111098c6m) (dFeipgt.h3)i.sTahsesifginrsetdoaccduartreenocfe1o8f3E0ro±di4u0m. The data and error estimate are approximated from nus, and Heteromeles. Artemisia is primarily Arte- the Santa Barbara Basin data (Mensing and Byrne misia californica. Other Asteraceae are difficult to binadraieetnnietvndieoftiyntaibFxdieeagnlutofirwrfeioem2fd.amtPbihoelealycoBewlraeaevtsehslaeincdafanacPdmeoialleheyamvaolreeneviebalpe.creeenSsaeeevncteaorlmasi-lon iioandfnecpn1ert9ei5sofs3if)e,daCteabdtnyhrdeuCts4ha7eppoocslnpmlieecgdnireeopasttdh(3i15si9sp7e6cbr)ams.saeldTdaahtboeeinsliftatiyro.s13t7T1Ch9asp6ep4pedeaaaartsk-e- suming a 15 year maturation period following the small quantities early in the record; however, alien taxa became important in California in the early first planting in 1949 (Peterson 1980). The disparity in sedimentation rate between the upper core (110 r1e9stehntcsenitnutrryodauncdedthspeecpioelsl.enErioncdrieuamsehparsobbeaeblnyidreenp-- cm in 160 years) and the lower core (165 cm in tified to the species Erodium cicutarium, a Medi- 1500 years) suggests that the radiocarbon age may terranean annual (Mensing and Byrne in press). be artificially old. Even assuming no changes in sedimentation rate, the lowerhalfofthe core would Pollen analysis. Below 1760, Quercus shows span at least two centuries ofvegetation history pri- few changes; however, after 1760 the record be- or to Spanish settlement. . 1998] MENSING: 560 YEARS OF VEGETATION CHANGE 0a3 <U wo "C fi V -- u _ii 1 X TCa3 J*<-Su> i B ^ % i <CoU3 AO a 1 1 II! ilii; ii i i U 2 'vTr \ X ft. $5 B 6 3 t £ c -3a u MADRONO 6 [Vol. 45 Taxonomy. Fifty-four pollen and spore types an interesting short term increase at the 60 and 65 were identified (Mensing 1993), of which ten taxa cm levels (ca. 1920 to 1930), jumping up to 14%. are graphed in Figure 3. Species of oaks in the re- Zea mays (corn) pollen was also found in the 65 gion that most likely contribute pollen to the site cm level, suggesting a period of local cultivation. % include Q. agrifolia, Q. lobata, Q. durata, Q. chry- Brassicaceae is present at about 1 in the core sec- solepis, Q. xvislezenni, and Q. douglasii. Of these tion below 65 cm (ca. 1920). From 65 cm up to the species Q. agrifolia and Q. lobata dominate the re- surface level, Brassicaceae steadily increases from gion today. Pinus includes P. ponderosa, P. sabi- 1% up to 6%. Erodium first appears in the core at niana, andP. coulteri as well asP. attenuata, plant- 110 cm depth and is present in ten levels. Eucalyp- ed around the turn ofthe century. Rhamnaceae/Ro- tus is present in three levels with approximate dates saceae includes taxa similar to those described for of 1920, 1930, and 1960. Cedrus first appears in the Santa Barbara Basin as well as Cercocarpus 1964 and increases in abundance in the surface betuloides, Prunus ilicifolia, Heteromeles arbuti- samples. folia. Asteraceae around Zaca Lake primarily rep- resent herbaceous taxa. Brassicaceae probably in- Discussion clude introduced mustards in the post-European pe- riod. Erodium was identified as Erodium cicutar- Oak woodlands. The Santa Barbarapollenrecord ium. Eucalyptus is no longer present at the site and shows no significant vegetation changes during the the type is unknown. Cedrus is from Cedrus deo- pre-European period. The evidence suggests that dora planted near the lake. oak woodland populations remained stable for up to four centuries. Beginning around 1870 and con- Pollen analysis. Twenty-six levels were analyzed tinuing until 1985, percent Quercus pollen steadily at approximately 10 cm intervals. Six levels were increases to its highest level in 560 years. Principal excluded from the analysis because of extremely components analysis of pollen accumulation rates high percentages of Asteraceae pollen presumably indicates that the abundance of Quercus pollen has associated with erosion events. Dense clay layers indeed increased over the last century (Mensing are present in the core at 130-150 cm and 200- 1993). The twofold increase ofQuercus pollen dur- 240 cm depth. These lenses are associated with ing the last century strongly suggests an increase above average magnetic susceptibility and gener- in oak woodlands in the Santa Barbara region. This ally low organic content (Mensing 1993). High increase may be from an increase in woodland den- magnetic susceptibility readings generally result sity, expansion of oak woodland habitat, or a com- from deposition ofiron bearing sediments. Peterson bination of the two. (1980) hypothesized that these layers may be as- The Zaca Lake record is less clear concerning sociated with periods of higher than average ero- oak woodlands. For most of the record, Quercus is sion. Asteraceae pollen is particularly resistant to the dominant pollen type with maxima averaging biodegradation and during periods of high runoff, 50 percent. Periodic declines in Quercus consis- pollen accumulated on the soil surface may have tently correspond with increases in Asteraceae. An been washed into the lake biasing the sample. increase in individuals of the Asteraceae, locally For most of the record, Quercus pollen remains composed primarily of herbaceous annuals, would between 40-50%. Low percentages are seen at the not displace oak woodlands. Since decreases in 50-60 cm 120 cm and 170-180 cm depths. The Quercus are not associated with increases in woody declines are mirrored by increases in Asteraceae. taxa, I suggest that oak populations in this area re- These strataare notclay layers; however, they show mained stable prior to the mid-1900's. Since 1950 above average magnetic susceptibility suggesting another woody taxon, Pinus, has increased substan- that they may also be associated with erosion tially. The increase in the importance of Pinus re- events. Organic rich lake sediments (gytja) com- corded in the pollen record is confirmed by repeat monly had 50% Quercus pollen. In the upper part photography (Fig. 4). Scattered pine groves, visible ofthe record, from about 1970 to thepresent, Quer- on the distant slopes in the 1895 photograph, now cus averages 35%. Unlike previous decreases in appear as dense forest. Today, the understory sur- Quercus, Asteraceae also declines during this time. rounding the lake is thick with young pines and Pinus values remain stable at about 10% through oaks, but pines over-top oaks in most places. Zaca most of the core then begin to increase rapidly to Lake is located at the transition between oakwood- 29% in the mid 1900's. The Rhamnaceae/Rosaceae land and coniferous forest. Although oaks may be curve shows little variation averaging about 12%. increasing to some extent at this site, the primary Similarly, Artemisia varies little, reaching as high signal in the pollen record, supported by evidence as 7% between 40-65 cm, but remaining at less from repeat photography, is an increase in the im- than 4% for most of the record. portance of pines. Asteraceae is the most important herbaceous pol- This study presents the first high-resolution pae- len type. The record is highly variable, ranging loecologic records to document changes in Califor- from 1% to 34%. Poaceae, which remains fairly nia oak woodlands from the pre-European period constant at about 2-3% for most ofthe core shows to the modern period. Of significance here is that 1998] MENSING: 560 YEARS OF VEGETATION CHANGE 7 Fig. 4. Repeat photography ofZaca Lake taken from the north shore looking east. The upper photograph was taken by a local Santa Barbara photographer ca., 1895 (courtesy Santa Barbara Historical Society Museum). The lower photograph is by the author, 1992. Q. agrifolia populations in the Santa Barbara area European period. A resampling of permanent plots have increased in the recent century, after a long in northern California found that Q. douglasii have period with no apparent changes. Other studies, pri- increased over the last 60 years, with live oaks be- marily stand age analyses ofQ. douglasii, have also ginning to emerge as co-dominants (Holzman documented increases in woodland density (White 1993). 1966; Vankat and Major 1978; Mensing 1992); In general, there is concern that California oak however, these studies do not extend to the pre- woodlands are in decline. A recent assessment of MADRONO 8 [Vol. 45 Q. douglasii found that 87% ofthe study locations 1993). The Santa Barbara Basin pollen record sug- were experiencing a net loss in both tree density gests that the last 100 years have produced such an and canopy cover (Sweicki et al. 1993). Studies increase in oaks in the Santa Barbara region. have documented negative impacts of human- Reduction in fire frequency may also be respon- caused environmental change to oak woodlands sible for the recent increase in woodland and shrub throughout the state including direct loss of wood- cover at Zaca Lake. Fires have been systematically lands through clearing (Bolsinger 1988; Rossi recorded in the Los Padres National Forest since 1980) and poorregeneration as a result oflivestock 1911. Three fires have burned on the chaparral grazing, invasion of annual grasses, and other slopes to the northwest of the lake; however, no changes (Griffin 1971; Bartolome et al. 1987; fires larger than a few acres have burned the wood- Borchert et al. 1989; Harvey 1989; Gordon and ed slopes in the upper Zaca Lake watershed (Los Rice 1993; Muick 1995). Padres National Forest Fire Statistical Database). To understand the full extent ofthese changes on Here, absence offire appears to have favored pines California oak woodlands, it is valuable to have overoaks. ZacaLake is located atthe pine/oakeco- data on how current populations compare with tone. The tendency for pine to invade oak wood- those from the pre-European period. In this regard, land following fire suppression has been clearly paleoecologic studies provide important informa- demonstrated in Yosemite Valley where open oak tion to understand the long term implications ofhu- meadows were converted to closed coniferous for- man-caused environmental change. This study sug- est after fire suppression (Reynolds 1959; Gibbens gests that Q. agrifolia populations in the SantaBar- and Heady 1964; Anderson and Carpenter 1991). bara region have increased during the modern pe- The Zaca Lake pollen record suggests that at upper riod, a time of significant human-caused elevation sites where coast live oak grows with environmental change. pines, coniferous forest will replace oak woodland The increase in Q. agrifolia is most apparent in in the absence of frequent fire. the 1900's. I believe that the environmental change most likely to have resulted in an increase of oak Chaparral, coastal-sage scrub, and herbaceous woodland is a change in fire regime. In the absence vegetation. There is some debate concerning the of fire, Q. agrifolia tends to increase. Density and impact of European settlement on chaparral, coast- canopy cover for Q. agrifolia at Burton Mesa in al-sage scrub, and herbaceous vegetation. Dodge Santa Barbara County was found to be highest on (1975) argued that grassland was much more ex- sites without recent fires (Davis et al. 1988). tensive during the pre-European periodbecause fre- McBride (1974) examined plant succession in the quent low-intensity fires cleared out young shrub Berkeley hills and suggested that in the absence of seedlings. He concluded that heavy grazing and fire recurrent fires, Q. agrifolia and Umbellularia cali- suppression have reduced low-intensity fires and fornica would succeed Baccharis pilularis. In a permitted shrub invasion of vast areas formerly comparison of vegetation dynamics on burned and dominated by grasses. Timbrook et al. (1982) ech- unburned plots at Gaviota State Park west of Santa oed this sentiment and concluded that chaparral has Barbara, Callaway and Davis (1993) found that increased in density and extent over the last 200 chaparral was being converted to oak woodland at years because of suppression of grassland burning. arate of0.12% peryearin the absence offire. They Furthermore, they suggested that a grassland which predicted that with the absence of fire and grazing, dominated the SantaBarbaracoastal plain andfoot- oak woodland would dominate a larger proportion hills has been largely replaced by coastal-sage of the landscape. scrub as a result of fire suppression. The Chumash regularly set fires along the coastal The pollen record does not support the idea that plain, and this practice continued even after estab- chaparral has expanded overthe last 200 years. The lishment of the missions (Timbrook et al. 1982). Rhamnaceae/Rosaceae curve from each site show Many of these grass fires probably burned through virtually no consistent trends (Figs. 2, 3). If any- the understory of adjacent oak woodlands, killing thing, the Santa Barbara Basin diagram shows a oak seedlings and saplings. Trees of less than 7.5 modest decline in chaparral taxa in the recent cen- cm diameter breast height have bark approximately tury. The pollen record suggests that chaparral has 0.6 cm thick and may be killed by low intensity not expanded its range in response to European im- fires (Plumb and Gomez 1983). This process would pacts. have maintained open oak woodlands similar to the There is some evidence to suggest a modest in- oak parks typically described by early Spanish ex- crease in the importance ofcoastal-sage scrub over plorers. Since the turn of the century, urban and the last 200 years. Artemisia averages 7% of the agricultural development has concentrated in areas pollen sum from the period between 1425 and 1820 dominated by grassland and oak woodland. Al- (Fig. 2). However, beginning in 1820, it increases though urban and agricultural development have to 12%, and averages 10% between 1820 and 1985. been responsible for clearing oaks, fire protection Pollen percentage remains at the higher levels ex- in developed areas favors oaks in nearby wildland cept for two brief declines centered on 1920 and settings (Davis et al. 1988; Callaway and Davis 1980. At Zaca Lake Artemisia averages 2.7% prior 1998] MENSING: 560 YEARS OF VEGETATION CHANGE 9 to about 1800, then increases to an average of5.0% The pollen record shows that prior to European in the upper core (Fig. 3). Although this may rep- settlement, oak populations had been stable for at resent a true increase, it is difficult to interpret too least three centuries. In the past two centuries, oak much from such a small change. populations have changed in response to European Comparison of burned and unburned plots in impacts, including the introduction ofgrazing, sup- Santa Barbara County found that coastal-sage scrub pression of fire, and a shift in understory compo- invaded grassland in the absence of fire, but fre- sition. In some cases, these changes appear to have quent fire favored grassland (Callaway and Davis favored oaks, creating woodlands more dense than 1993). Westman (1976) also found thatcoastal-sage during the pre-European period. scrub replaced undisturbed grassland when fire was Chaparral does not appear to have expanded sig- removed. In northern California, Baccharispilular- nificantly in response to European land use is was found to invade grassland during periods of changes. Coastal-sage scrub may have expanded low fire frequency (McBride and Heady 1968). Re- some; however, the evidence for this change is duced fire frequency along the coastal plain may weak. Invasion and transformation ofgrassland ap- have favored a slight expansion of coastal-sage pears to have begun particularly early with the first scrub; however, there is no evidence that this im- alien taxa reaching the area even before the first pact affected the distribution or abundance ofchap- Spanish settlement of California. arral. Pollen evidence of herbaceous taxa shows that Acknowledgments the invasion of alien species into grasslands began very early. Erodium first appears in the pollen rec- theIftihealdnkanRdoglearborBaytronrey aanndd fEorricreEvdileuwnidngfoeraralisesrisvtearnscieonisn ord in 1760, nearly a decade prior to the first Span- of this paper and three anonymous reviewers for their ish settlement in San Diego and more than 20 years valuable comments. I am particularly indebted to Arndt before the founding of the Mission Santa Barbara Schimmelmann and Carina Lange for samples and devel- (Fig. 2). The pollen has been identified as Erodium opment ofthe Santa Barbara Basin core chronology. This cicutarium (Mensing and Byrne in press), a Medi- work was supported by the National Science Foundation, terranean native, and provides evidence that the in- The California Department of Forestry Integrated Hard- vasion and transformation ofherbaceous vegetation wood Range Management Program, and Sigma Xi. began prior to European settlement. Polemoniaceae averages nearly 1% and is consistently present Literature Cited through the 1700's. Asteraceae, the dominant her- baceous pollen type, averages 18% in the pre-Eu- Anderson, R. S. and S. L. Carpenter. 1991. Vegetation change in Yosemite Valley, Yosemite National Park, ropean period. Both taxa decline markedly in the California, during the protohistoric period. Madrono modern period. The decline becomes particularly 38:1-13. pronounced after the arrival of alien Brassicaceae Anderson, M. V. and R. L. Pasquinelli. 1984. Ecology which became widespread along the coastal plain and management ofthe northern oak woodland com- in the early 1800's (Cleland 1951). munity, Sonoma County, California. M.A. thesis. Sonoma State University, Rohnert Park, CA. Conclusions Bartolome, J. W., P. C. Muick, and M. P. McClaran. 1987. Natural regeneration of California hardwoods. The evidence from this study suggests that oak Pcopo.rd2s6.-),31PrionceT.edRi.ngPsloufmbtheanSdymNp.oHs.iuPmillosnbuMruylt(itpelceh-. woodlands in the Santa Barbara region have in- Use Management of California's Hardwood Re- creased during the last 100 years. The nature ofthis sources. U.S. Department ofAgriculture, Forest Ser- increase varies between sites. In the Santa Barbara vice, Pacific Southwest Forest and Range Experiment area, Q. agrifolia appears to have increased begin- Station, Berkeley, CA. ning in the late nineteenth century. Fire suppression Bolsinger, C. L. 1988. The hardwoods of California's on the coastal plain has probably been the main timberlands, woodlands and savannas. U.S. Depart- factor contributing to this increase. The Chumash ment of Agriculture, Forest Service, Pacific North- wariethreapnortiendcrteoasheavienpesreitotdliecmaelnlty asentdfidreesv;elhoopwmeevnetr,, BorcwOheyeslrtetr,R.eMs.1e9a8Ir.9,c.hF.ISnWtta.etriDaocant,viioPsno,srtJol.afMndif,acchtOaoRer.slsaefnf,ectainndg Ls.eeDd.- burning has been suppressed. In the absence offire, ling recruitment of blue oak {Quercus douglasii) in A Q. agrifolia has increased in density. policy of California. Ecology 70:389-404. fire suppression appears to favor Q. agrifolia, and Brown, R. W. and F. W. Davis. 1991. Historicalmortality where fire return intervals are long, oaks would be of valley oak {Quercus lobata, Nee) in the Santa expected to continue to increase in density. Ynez Valley, Santa Barbara County, 1938-1989. Pp. At higher elevations where Q. agrifolia grow 202-207 in Standiford, R. B. (tech. coord.), Proceed- alongside pine, such as at Zaca Lake, fire suppres- iwnogosdofrathnegeslyanmdpomsainuamgeomnenota.k wUo.oSd.laDnedpsaratnmdenhtardo-f sion appears to have favored pine over oak. Here, Agriculture, Forest Service, Pacific Southwest Forest coniferous forest is expanding into oak woodland. and Range Experiment Station, General Technical The taller pines may eventually shade out the Report PSW-126. slower growing oaks ifthe present trend continues. Callaway, R. M. and F. W. Davis. 1993. Vegetation dy- MADRONO 10 [Vol. 45 namics, fire, and the physical environment in Central 1973. Geochronological studies in Santa BarbaraBa- California. Ecology 74:1567-1578. sin. Limnology and Oceanography 18:763-770. Caponigro, M. A. 1976. The diatom stratigraphy and pa- Mayfield, D. W. 1981. Ecology ofa discovered land. Pa- leolimnology of Zaca Lake, California. M.A. thesis. cific Discovery. Pp. 12-20. University ofCalifornia, Santa Barbara. McBride, J. R. 1974. Plant succession in the Berkeley Cleland, R. G. 1951. The cattle upon a thousand hills. Hills, California. Madrono 22:317-329. Huntington Library Publications, San Marino, CA. McBride, J. R. andH. F. Heady. 1968. Invasion ofgrass- Cole, K. L. and G. Liu. 1994. Holocene paleoecology of land by Baccharis pilularis. Journal of Range Man- an estuary on Santa Rosa Island, California. Quater- agement 21:106-108. nary Research 41:326-335. McClaran, M. P. 1986. Age structure of Quercus doug- Davis, F. W., D. E. Hickson, and C. O. Dennis. 1988. lasii in relation to livestock grazing and fire. Ph.D. Composition ofmaritime chaparral related to firehis- dissertation. University ofCalifornia, Berkeley. tory and soil, Burton Mesa, Santa Barbara County, McClaran, M. P. and J. W. Bartolome. 1989. Fire-re- California. Madrono 35:169-195. lated recruitment in stagnant Quercus douglasii pop- Dodge, J. M. 1975. Vegetational changes associated with ulations. Canadian Journal ofForestry 19:580-585. land use and fire history in San Diego County. Ph.D. Mensing, S. A. 1992. The impact ofEuropean settlement dissertation. University ofCalifornia, Riverside. on blue oak (Quercus douglasii) regeneration and re- Faegri, K. and J. Iversen. 1975. Textbook of Pollen cruitment in the Tehachapi Mountains, California. Analysis, 3rd ed. Hafner Press, New York. Madrono 19:36-46. Ferren, W. R. 1985. Carpinteria Salt Marsh: Environ- Mensing, S. A. 1993. The impact ofEuropean settlement ment, History, and Botanical ResourcesofaSouthern on oak woodlands and fire: Pollen and charcoal evi- California Estuary. The Herbarium, Department of dence from the Transverse Ranges, California. Ph.D. Biological Sciences, University of California, Santa dissertation. University ofCalifornia, Berkeley. Barbara. Mensing, S. and Byrne, R. inpress. Pre-missioninvasion Gibbens, R. P. and H. A. Heady. 1964. The influence of ofErodium cicutarium in California. Journal ofBio- modern man on the vegetation of Yosemite Valley. geography. University ofCalifornia, DivisionofAgriculturalSci- Muick, P. C. 1995. Effects ofshade and simulatedherbiv- ences Manual 36. ory on seedling emergence, survival and growth of Gordoofn,grDa.ssRl.anadndanKn.uaJl.sRiocne.s1o9il93.waCtoemrpeatnidtivbelueeffeocatks ftowloiaCNaeleif)orannidabolaukes:oaCkoa(sQtuelrivceusoadkoug(lQauseiricuHsooakgri&- (Quercus douglasii) seedlings. Ecology 74:68-82. Arn). Ph.D. dissertation. University of California, GordRoinc,e.D.198R9.., JC.omRp.etWietlikoenr,forJ.soWi.l wMaetnekre,betawnedenK.anJ-. MuicBke,rkPe.lCe.y.and J. R. Bartolome. 1987. An assessment nual plants and blue oak (Quercus douglasii) seed- of natural regeneration of oaks in California. Final lings. Oecologia 79:533-541. Report, California Department of Forestry: The For- Griffin, J. R. and W. B. Critchfield. 1976. The distri- est and Rangeland Assessment Program. bution of forest trees in California. U.S. Department Peterson, S. 1980. Late Holocene sedimentation at Zaca ofAgriculture, Forest Service, Pacific SouthwestFor- Lake, Santa Barbara County, California. B.A. senior est and Range Experiment Station, Research Paper thesis, University ofCalifornia, Berkeley. PSW-82. Plumb, T. R. and A. P. Gomez. 1983. Five southern Cal- Griffin, J. R. 1971. Oak Regeneration in the Upper Car- ifornia oaks: identification and postfire management. mel Valley, California. Ecology 52:862-868. U.S. Department of Agriculture, Forest Service, Pa- Hall, C. A. 1981. Map of geology along the Little Pine cific SouthwestForest andRangeExperimentStation, General Technical Report PSW-71. Fault, parts of the Sisquoc, Foxen Canyon, Zaca Lake, Bald Mountain, Los Olivos, and Figueroa Reynolds, R. D. 1959. Effect ofnatural fires and aborig- Mountain quadrangles, Santa Barbara County, Cali- ivnaadla.buMr.nAi.ngthuepsoisn,tUhneivfeorressittsyoofftChaelicfeonrtnriaal,SBieerrkraelNeey-. fornia. U.S. Geologic Survey Miscellaneous Field Studies. Map MF-1285. Rossi, R. S. 1980. History of cultural influences on the distribution and reproduction of oaks in California. Harvey, L. E. 1989. Spatial and temporal dynamics of a Pp. 7-18 in T. R. Plumb and N. H. Pillsbury (tech. blue oak woodland. Ph.D. dissertation, University of coords.), Proceedings ofthe Symposium on the Ecol- California, Santa Barbara. ogy, Management, and Utilization ofCaliforniaOaks. Heussifeorr,niaL:. AE. 1129,7080.0PyoelalrenreicnorSda.ntGaeoBlaorgbiacraalBSaosciin,etCyalo-f cUi.fSi.c SDoeuptahrwtemsetntForoefstAgarnidcuRltaunrgee, EFxorpeesrtimSeenrtviSctea,tiPoan-, America Bulletin 89:673-678. General Technical Report PSW-44. Hickman,J.C. (ed.) 1993.TheJepsonManual:HigherPlants SCHIMMELMAN, A., C. B. LANGE, ANDW. H. BERGER. 1990. ofCalifornia. University ofCalifornia Press, Berkeley. Climatically controlled marker layers in Santa Bar- Holzman, B. A. 1993. Vegetation change in California's bara Basin sediments, and fine-scale core-to-corecor- blue oak (Quercusdouglasii) woodlands, 1932-1992. relation. Limnology and Oceanography 35:165-173. Ph.D. dissertation. University ofCalifornia, Berkeley. Schimmelman, A., C. B. Lange, W. H. Berger, A. Simon, Hulsemann, J. and K. O. Emery. 1961. Stratification in S. K. Burke, and R. B. Dunbar. 1992. Extreme cli- recent sediments ofSantaBarbaraBasin ascontrolled matic conditions recorded in Santa Barbara Basin by organisms andwatercharacter.Journal ofGeology laminated sediments: the 1835-1840 Macoma Event. 69:229-290. Marine Geology 106:279-299. Koede, M. A., A. Soutar, and E. D. Goldberg. 1972. Soutar A. and P. A. Crill. 1977. Sedimentation and cli- Marine geochronology with Pb-210. Earth and Plan- matic patterns in the Santa Barbara Basin during the etary Science Letters 14:442-446. 19th and 20th centuries. Geological Society ofAmer- Krishnaswami, S., D. Lal, B. S. Amin, and A. Soutar. ica Bulletin 88:1161-1172.