IS THERE STILL NATIVE DIVERSITY IN CALIFORNIA GRASSLANDS? by Megan E. Lulow and Truman P. Young

ur understanding of the A LOST LEGACY We have been involved in sur- native flora of Cali- veying grasslands in two regions of fornia’s Central Valley The Central Valley and adjacent California, both of which Keeler- is undergoing a sea hills of California have been so Wolf et al. (2007) refer to as the Ochange. Although this community heavily invaded by non-native spe- “Valley and south coastal” type in has traditionally been described as a cies, particularly annual grasses (e.g., their classification of the state’s ma- grassland once dominated by native bromes and wild oats) and herbs jor grasslands. Large-scale grassland perennial bunchgrasses (Burcham (e.g., mustards and starthistle), that restorations were being planned and 1957, Heady 1988, Schiffman 2007, few clues remain of the original veg- implemented throughout the two Stromberg et al. 2007), there have etation. Hiking through most former regions we surveyed: the foothills been many over the decades who rangelands in the Central Valley or of the Santa Ana Mountains within have suggested that the emphasis within the California Coast Range, the Irvine Ranch National Natural on grasses was misplaced, and that one finds only a scattering of native Landmark in Orange County, and native wildflowers (non-grass herbs) plant species, or perhaps rare pock- the Coast Range foothills on the were a significant component as well ets where natives truly dominate western edge of the Sacramento Val- (John Muir 1894; Hamilton 1997). (Noss et al. 1995). The majority of ley in Yolo County. For both re- Recently these voices have risen areas, particularly those recently re- gions we sought to document the again, spearheaded by Glen Holstein leased from grazing, have accumu- occurrence and general abundance of the CNPS and Richard Minnich in lated high levels of thatch and deca- of upland native “grassland” spe- his book, California’s Fading Wild- dent growth of non-native species. cies, and to generate a reference list flowers. Our own survey work has There is very little record of the to facilitate the enhancement of spe- revealed a wonderfully rich flora of composition of California grassland cies diversity in grassland restora- flora prior to the invasions of non- tion projects. Both regions consisted native species, cattle ranching, and of sites within a larger, often de- large-scale cultivation (but see Muir graded landscape of grassland, as sidebar, page 9). The invasion by well as smaller grasslands that were exotic plants has been so extensive part of an oak woodland or scrub that formal classifications of Cali- mosaic. fornia flora recognize the grassland dominated by non-native annuals as its own distinctive subtype, or IRVINE RANCH NATURAL alliance (Sawyer et al. 2009). In fact, LANDMARK there are few plant communities on the planet for which we have so A map of native grasslands and little confidence about their original the extent of their degradation was A post-fire wildflower display representing species frequently found across areas with composition. developed for management planning native remnants in grassland and grassland We have observed, as have oth- within the Lomas Ridge and Lime- scrub. Photograph by Jutta Burger. ers (e.g., Ayzik Solomeshch, Glen stone Canyon regions of the Irvine Holstein, Peter Hopkinson, pers. Ranch National Natural Landmark. wildflowers in remnant grasslands comm.; see also Tables 14.3 and 14.4 National Natural Landmarks are des- throughout the state, and the pages in Bartolome et al. 2007) that in ignated by the U.S. Secretary of In- of Fremontia are often filled with both “remnant” sites and sites con- terior for natural areas in both pub- fields of magnificent wildflowers. In verted to dominance by exotic an- lic and private ownership. They are addition to their aesthetic appeal, nual grasses, many native wildflower recognized as outstanding examples wildflowers contribute to ecosystem species persist. It is clear that calling of the natural heritage of the coun- function, including nitrogen fixation, these communities “grasslands” may try, alongside national parks, recre- pollinator resources, habitat struc- be misleading (although we will use ation areas, and monuments. ture, and forage quality. the term here, for form’s sake). After a fire in October 2007, our

6 FREMONTIA VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 A mosaic of grassland (background, left), scrub (foreground), and ecotone (center), plant communities common to the Irvine Ranch Natural Landmark survey region. Photograph by Jutta Burger. team of scientists visited all but the species occurring at more than half and corms). However, the woody least accessible grasslands in these of all sites, and only 15 occurring at and geophyte species were more hills. Grasslands were defined as more than one-fifth of all sites. The common and consistently found having less than 20% cover of woody number of native wildflower spe- across multiple grasslands. species, including both coast live cies ranged from 5–61 across sites, There were substantially more oak (Quercus agrifolia) and coastal and always far exceeded the num- native species in ecotone grass- sage scrub species. At each of 475 ber of native grass species (Figure lands—a transition area between sites across the hills, we developed a 1), and also often exceeded the na- two plant community types, such list of all native species, and esti- tive grasses in cover. as native grassland and woodland, mated the percent cover of each life Some of the most frequent wild- or native grassland and scrub. How- form (grass, wildflower, shrub, and flowers included: fascicled tarweed ever, at Irvine Ranch Natural Land- tree). Based on floristic composi- (Deinandra fasciculata), blue dicks mark, ecotones contained half the tion, these sites were divided into (Dichelostemma capitatum), minia- cover of the most common native six community types, or alliances ture lupine (Lupinus bicolor), coastal grass, purple needlegrass (Stipa (Figure 1). goldenbush (Isocoma menziesii), and pulchra), than the more open grass- Across all surveyed grasslands common goldenstars (Bloomeria land. While these ecotone areas are in the Irvine Ranch Natural Land- crocea). Across all surveyed grass- species rich, they likely serve a dif- mark, there were a total of 140 lands, the overall abundance and ferent function to the ecosystem native species: 99 wildflowers, 12 species richness of native wildflow- because the relative abundance (or grasses, 1 rush, 26 shrubs, and 2 ers was greater than that of woody cover) of species is so different from trees. Most species were observed species or geophytes (wildflowers that in the core area of each plant at low frequencies, with only two growing from underground bulbs community type.

VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 FREMONTIA 7 WESTERN SACRAMENTO dominated by blue oak (Quercus cies were found in the remnant VALLEY FOOTHILLS douglasii) in this region. At these grassland flora (10 grasses, 67 wild- Northern California sites, only pres- flowers, 1 shrub, and 1 tree). The We had also carried out a survey ence or absence of native species number of native wildflower spe- of grassland remnants in the west- was recorded. Four of the seven sites cies in each area ranged from 20 to ern Sacramento Valley foothills, lim- had cattle or sheep grazing occur- 36 (Figure 2). Many of these spe- ited to grasslands known to have ring at low levels of intensity. In cies were widespread; 25 species either a native wildflower or grass general, ecotones, rocky ridge tops, were observed in more than half the component to the flora, even if com- and north facing slopes were better sites surveyed. Geophytes accounted prising only 10–15% of the total places to find the remnant patches for 20% of wildflower species that plant cover at some sites (Lulow than flatter, open grassland. None- occurred in at least one-half of all and Young 2009). Seven areas were theless, all sites had at least 50% of sites surveyed. Some of the more chosen, with a total of 58 sites sur- their cover represented by exotic frequent wildflowers included Cali- veyed during the spring and sum- annual plant species. fornia yarrow (Achillea millefolium), mer of 2002–2004 (Figure 2). Grass- Again, wildflowers dominated miniature lupine (Lupinus bicolor), lands were defined as having less the native species lists. Across the brodiaea (Brodiaea spp.), valley tas- than 20% cover of woody species, seven sites, at least 79 native spe- sels (Castilleja attenuata), tree clo-

Degraded grassland dominated by non-native annual grasses at Limestone Meadow in the Irvine Ranch Natural Landmark. Photograph by David Olson.

8 FREMONTIA VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 TABLE 1. COMMON SPECIES SHARED BETWEEN SURVEY REGIONS. Common Name Species Fascicled tarweed, hayfield tarweed Deinandra fasciculata or Hemizonia congesta luzulifolia Miniature lupine Lupinus bicolor Blue dicks Dichelostemma capitatum Arroyo lupine Lupinus succulentus Mariposa lily: splendid, Catalina, or yellow Calochortus splendens, C. catalinae, or C. luteus Gum plant Grindelia camporum Short-podded lotus, strigose lotus Acmispon (formerly Lotus) brachycarpus or A. strigosus California plantain Plantago erecta

Note: Tarweeds, mariposa lilies, and annual lotuses did not share the exact same species between regions, but had similar taxa. Source: Lulow and Young, 2009.

ver (Trifolium ciliolatum), and hay- surveys they were among the more ern foothills of the Sacramento Val- field tarweed (Hemizonia congesta). frequent species across sites. This ley, whereas more shrubs were found group of species are often not in- in the grasslands on the Irvine Ranch cluded in grassland restoration spe- Natural Landmark. Greater total spe- NORTHERN VS. SOUTHERN cies lists and can be cryptic in rem- cies richness was recorded on the SURVEYS nant native communities, appearing Irvine Ranch Natural Landmark, but aboveground primarily only after this may be because a far greater While the survey within the fires or wet years. These results sug- number of sites were surveyed there. Irvine Ranch Natural Landmark gest this group may be an over- found that most species occur at looked, yet significant component low frequencies across a variety of of native grassland composition GRASSLAND? FORBLAND? sites, the two survey areas shared 8 across the state. PRAIRIE! of their 25 most common native There were conspicuous differ- wildflowers (see Table 1). While ences between the two areas. In par- In summary, throughout the sur- geophytic species represent a small ticular, native clover species were veyed remnant sites, the number of fraction of the total richness, in both found more frequently in the west- native wildflower species was far

AN EARLY ACCOUNT OF THE CENTRAL VALLEY PRAIRIE

here are several accounts of the flower-rich flora of the Central Valley in the era before large-scale T agriculture, but perhaps none as eloquent as this, from John Muir: The Great Central Plain of California, during the months of March, April, and May, was one smooth, continuous bed of honey-bloom, so marvelously rich that, in walking from one end of it to the other, a distance of more than 400 miles, your foot would press about a hundred flowers at every step. Mints, gilias, nemophilas, castilleias, and innumerable compositæ were so crowded together that, had ninety-nine per cent of them been taken away, the plain would still have seemed to any but Californians extravagantly flowery. The radiant, honey-ful corollas, touching and overlapping, and rising above one another, glowed in the living light like a sunset sky—one sheet of purple and gold, with the bright Sacramento pouring through the midst of it from the north, the San Joaquin from the south, and their many tributaries sweeping in at right angles from the mountains, dividing the plain into sections fringed with trees. — from The Mountains of California by John Muir, 1894

VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 FREMONTIA 9 FIGURE 1. SPECIES RICHNESS OF NATIVE PLANTS AT applying herbicide—can stimulate IRVINE RANCH NATIONAL NATURAL LANDMARK. native wildflower species to germi- nate that were not evident prior to the clearing (personal observation, M. Lulow). When putting together a species list for restoration, identifying those species that are able to coexist with non-native annual species can be as important as determining which spe- cies are most appropriate for a given site. This is particularly true for Cali- fornia grasslands, where it is virtu- ally impossible to completely eradi- cate non-native annuals due to the extent of their invasion throughout the state and their prolific seed pro- duction. Collaborative research between Note: Based on floristic composition cluster and similarity analyses, six plant communities researchers and practitioners can were identified. All sites had a minimum of 50% relative cover of non-native species. greatly inform which methods and species are most effective for restor- Source: Lulow and Young, 2009. ing components of California’s na- tive wildflowers to our grasslands. FIGURE 2. SPECIES RICHNESS OF WILDFLOWERS IN While there will be some species or WESTERN SACRAMENTO VALLEY FOOTHILLS. species groups shared among re- gions, given the diversity of species and physical characteristics of the state, restoration approaches and species used should be honed for a given region and environment. The large number of wildflower species included in these rather coarse sur- veys, their functional diversity, and their aesthetic variety suggest a rich palette for ecological restoration. These surveys also remind us of an essential question that warrants Note: Each site had a maximum of three native grass species, and had a minimum of 50% relative cover of non-native species. our revisiting, namely what should we call these plant communities that Source: Lulow and Young, 2009. are (were!) a mixture of native grasses together with herbaceous non-grasses greater than the number of native more highly invaded sites to the that vastly outnumber the former in grass species, particularly in grass- wildflower components of grassland species richness, and often dominate lands with less non-native cover or communities that are more intact, it in cover. “Grasslands” seems clearly ecotone grasslands. Individual native may be possible generally to recon- misleading. “Forblands” is dismiss- grass species often occurred at higher struct their lost components, includ- ive of the grasses. While some feel abundance (cover) than wildflowers, ing perennial grasses. that “prairie” evokes images on the but as a class were rarely dominant. Disturbed areas may also have tall-grass and short-grass regions of The persistence of native wild- surprisingly resilient seed banks or the Midwest, we believe that it is the flowers may offer a tool for recon- corm banks. In grasslands disturbed most neutral and therefore most de- structing the original composition for over a century and with virtually scriptive term, and we are increas- of California’s grasslands—an essen- no native cover, any of the follow- ingly using it to describe the original tial step in restoration. By compar- ing processes—clearing built up vegetation of open habitats in the ing remnant wildflower species in thatch, a fire, or after responsibly Central Valley.

10 FREMONTIA VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 Catalina mariposa lily (Calochortus catalinae), a native grassland geophyte, found in fine soils. Geophytes were found frequently with remnant native species in both survey regions across grassland sites. Both photographs by Fred Roberts.

REFERENCES ing wildflowers: Lost legacy and bio- logical invasions. University of Cali- Burcham, L.T. 1957. California range fornia Press, Berkeley, CA. land: An historico-ecological study Muir, J. 1894. The Mountains of Cali- of the range resource of California. fornia. The Century Company, New Division of Forestry, Department of York, NY. Natural Resources, Sacramento, CA. Noss, R.F., E.T. LaRoe III, and J.M. Hamilton, J.G. 1997. Changing percep- Scott. 1995. Endangered ecosystems tions of pre-European grasslands in of the United States: A preliminary California. Madroño 44:311–333. assessment of loss and degradation. Heady, H.F. 1988. Valley grassland. In National Biological Service, Wash- Terrestrial Vegetation of California, ed. ington, DC. M.G. Barbour and J. Major. Califor- Sawyer, John O., T. Keeler-Wolf, and J. nia Native Plant Society, Sacramento, Evens. 2009. A Manual of California CA. Vegetation, 2d ed. California Native Keeler-Wolf, T. et al. 2007. Commu- Plant Society Press, Sacramento, CA. nity classification and nomenclature. Schiffman, P.A. 2007. Species compo- In California Grasslands: Ecology and sition at the time of first European Management, ed. M. R. Stromberg, settlement. In California Grasslands: J.D. Corbin, and C.M. D’Antonio. Ecology and Management, ed. M.R. University of California Press, Ber- Stromberg, J.D. Corbin, and C.M. keley, CA. D’Antonio. University of California Lulow, M.E., and T.P. Young. 2009. High Press, Berkeley, CA. Megan Lulow, Irvine Ranch Conservancy, native wildflower richness in Central Stromberg, M.R., J.D. Corbin, and C.M. 4727 Portola Parkway, Irvine, CA 92620, Valley “grassland” sites in the west- D’Antonio, eds. 2007. California [email protected]; Truman ern Sacramento Valley and adjacent Grasslands: Ecology and Manage- Young, Department of Plant Sciences, foothills. Grasslands 14(3):7-11. ment. University of California Press, University of California, Davis, CA 95616, Minnich, R.A. 2008. California’s fad- Berkeley, CA. [email protected]

VOLUME 39:2/39:3, MAY/SEPTEMBER 2011 FREMONTIA 11