DOI: 10.1007/s002670010064 Ecological Restoration of Coastal Sage Scrub and Its Potential Role in Habitat Conservation Plans PETER A. BOWLER bird species, accelerate “spread” or expansion of CSS, and Department of Ecology and Evolutionary Biology can also introduce many epiphytic taxa that otherwise would University of California, Irvine be slow or unable to occupy developing CSS creations. Rep- Irvine, California 92697-2525, USA and tile, amphibian, butterfly, and rodent diversity in a salvaged White Mountain Research Station canopy restoration case study at the University of California, 3000 E. Line Street Irvine, showed CSS species foraging and inhabiting trans- Bishop, California 93514, USA planted canopy patches. ABSTRACT / Extensive acreage loss of coastal sage scrub Using restoration techniques to expand existing CSS (CSS), isolation of surviving stands, and the federal listing of stands has more promise than creating isolated patches, several animal species with obligate relationships to this plant and the creation of canopies resembling CSS mid-fire cycle community, particularly the threatened California gnatcatcher stands is now common. Gnatcatchers and other birds use (Polioptila californica), have led to attempts to create CSS to restorations for foraging and occasional nesting, and in mitigate habitat lost to urban development and other causes. some cases created stands along “biological corridors” ap- Many of these creations lie within habitat conservation plan pear to be useful to bird movement. Patches of trans- (HCP) sites, and they could play a more prominent role by be- planted sage scrub shrubs along habitat edges appear to ing repositories for plants taken from a single site having site- break up linear edge effects. There are no data on which specific genetics. long-term survival, succession, or postfire behavior can be predicted for CSS restoration sites, and postfire community Among others, one technique that increases initial resem- changes are not part of either mitigation or restoration plan- blance to natural stands uses digitized, to-scale photography, ning at present. Long-term planning including burning is which has been ground-truthed to verify vascular plant associ- needed so that a fire-adapted habitat will develop. Restora- ations, which appear as mosaics on a landscape. A combina- tion is important in retaining genetic resources, for amelio- tion of placing patches of salvaged, mature canopy plants rating edge effects, as habitat extenders in buffer zones within larger matrices of imprinted or container plant plots ap- around HCP sites, and by providing areas into which natu- pears to significantly enhance immediate use by CSS obligate ral stands can expand. Coastal sage scrub (CSS) restoration faces regional sage scrub (Diegan, Riversidian, Venturan, and Dia- landscape problems, including the deposition of air- blan) and two in the closely related coastal succulent borne nitrogen from anthropogenic sources (Allen scrub in northern Baja California, Mexico (Westman 1997), foliar damage from air pollution (Westman 1983; see DeSimone and Burk 1992 for a review of CSS 1985), lack of species richness (Davis 1999), increasing associations), nearly a dozen subassociations have been weediness of natural stands (Westman 1987, Bowler recognized (Kirkpatrick and Hutchinson 1977, 1980, 1990a), large seed banks of exotics, a lack of mycorrhi- Jones and Stokes 1992); however, few attempts have zal fungi at restoration sites, and the inherent difficulty been made to restore these finer-grained groupings. of establishing many species outside natural stands. The finer-grained aspects of this complex community, Lack of understory annuals as compared with near including its subassociation and exposure differences, neighbor stands (Bowler 1999b) and a lack of demon- should be recognized in conservation (DeSimone and strated fire cycle behavior (Bowler 1999a; see Figs. 1–3) Burk 1992) and restoration planning (Read 1994). are problems linked with mitigation and restoration. Habitat conservation plan settings provide excellent The term “coastal sage scrub” is a generic one, like opportunities for this. “montane forest.” Since Westman (1981a, 1981b) de- fined four major vascular plant associations in coastal Coastal Sage Scrub KEY WORDS: Habitat conservation plan; Ecological restoration; Coastal Coastal sage scrub once covered perhaps 2.5% of the sage scrub; Mitigation; Plant salvage coastal lowlands in California and extended along Pa- Environmental Management Vol. 26, Supplement 1, pp. S85–S96 © 2000 Springer-Verlag New York Inc. S86 P. A. Bowler Figure 1. An idealized representation of postfire succession in coastal sage scrub, illustrating interburn cycles of 20ϩ years. Following fires of this frequency, there is a burst of species richness as fire-following taxa appear immediately after a burn, then decline at around 7 years, with shrubs recovering rapidly and an accretion of non-fire-following annuals present until around 17–20 years. After burning, the cycle begins again. cific coastal Baja California, but today is highly frag- Leymus condensatus and Malacothamnus. Both annuals mented in California, with most fragments having a and understory perennials persist perhaps 7 years fol- dominance of edge, frequently in direct contact with lowing a fire. Some of the species, particularly fire- urban development or roads (see Saunders and others followers, are present in the seed bank, whereas others 1991 for a review of the consequences of habitat frag- invade from adjacent unburned stands and from the mentation). In a sense this is also a “lack of edge,” regional seed rain. The shrubs regenerate through root because natural “edges” abut oak woodland, grassland, sprouting, and a gradual accretion of nonfire-following or chaparral. Prior to the 1970s CSS was a poorly un- herbaceous taxa occurs until about 17–20 years, after derstood community, in part because some of the dom- which understory species richness declines. At about 40 inant canopy plants can invade burned stands of chap- years after a fire, there are few remaining. arral for a few years, thus it was viewed as successional Fire suppression has generated large fuel loads pro- (see DeSimone and Burk 1992 for an excellent litera- ducing vast burns. Prior to this, canopies had many age ture review). In the early 1980s Westman (1981a–c, classes for each species (Fig. 1). This cycle of postfire 1883, 1985) published a series of studies demonstrating diversity (Fig. 2) and its emulation has not been incor- that CSS is a distinct community, and there has been porated into restoration plans, which instead attempt extensive research on it since then. to mimic mid– or late–fire cycle conditions. Example of Sage scrub exhibits a postfire cycle of species appear- CSS restoration strategies that do emulate the fire cycle ance and replacement, beginning with characteristi- are presented in Fig. 3. cally fire-following taxa, such as Lupinus, Phacelia, Most stands today are surrounded by grassland dom- Escholtzia and Amsinkia spp., and perennials, such as inated by annual European grasses or the exotic black Ecological Restoration of Coastal Sage Scrub S87 Figure 2. An idealized representation of postfire succession in an unburned condition beyond 40 years. In this situation, after fire a similar postfire succession begins, with understory taxa declining after about 17–20 years, eventually being nearly absent by 40 years after the previous fire. mustard (Brassica nigra). Native grassland is nearly com- agreements, if additional species qualify for ESA pro- pletely extirpated and southern oak woodland is a re- tection, the areas to be developed would remain ex- stricted, endangered habitat. Altered fire cycles, lack of empt (“no surprises”; see Smallwood and others 1999). adjacent stands of different age classes, and isolation Preserving designated blocks of habitat does not neces- from “regional seed rain” has led to decline in diversity sarily benefit all the sensitive species within them, as throughout the postfire cycle. Late successional epi- many species have widely ranging, often coastal distri- phytic species, such as lichens, have suffered extensive butions so that single blocks reflect little of the genet- losses with many regional extinctions (Bowler and Rief- ics, remaining populations, or actual distribution of ner 1999). As a result, over 60 vascular plant and over such taxa (Westman 1987). Conservation plans, such as 30 animal taxa in CSS are viewed as rare, threatened, or the NCCP and habitat conservation plans (HCPs), have endangered. In the face of the magnitude of habitat been widely criticized in both the environmental activ- loss and extent of decline of plant and animal species ist (Luoma 1998, for example) and academic commu- with obligate relationships to CSS, regional preserve nities (Shilling 1997). systems (the Natural Communities Conservation Plan, or NCCP, for example) have been designed in an at- Mitigation tempt to protect surviving large blocks of habitat while resolving political conflicts. The trade-off of preserving Although some papers published as early as 1990 some areas and exempting rare, threatened, and en- (Bowler 1990b) recommended no further sacrifice of dangered species from legal protection was adopted by CSS, compensatory mitigation was not required until state and federal agencies, as well as the developers who the California gnatcatcher was federally listed as a were largely its promoters and architects. Under these threatened species in 1993. Around