A Manual of California Vegetation Online
Vegetation constitutes a large part of our state’s distinctive signature. The dark, fog-shrouded coastal redwood forests, the sunny open foothill woodlands of blue oak, and the aromatic sage scrubs of the central and southern coasts are each as representative of California as are its scenic coastline, Mediterranean climate, and Yosemite Valley.
Vegetation, the particular plant species in a region and their arrangement, is a part of the natural world that humans learn to recognize at an early age. It is our nature to differentiate between grasslands and shrublands or meadows and adjacent forests. They have diverse resource values that we enjoy and exploit. No wonder we distinguish economically important Douglas fir forests from nearby fields of manzanita. We are all vegetation classifiers to some degree.
Unlike an artist’s contrasting brush strokes, patterns of vegetation may be anything but precise. Although we can agree that a dense lodgepole pine forest is qualitatively different from an adjacent wet mountain meadow, the closer we look, the harder it is to discern where the meadow begins and the forest ends. We see individual sedges, asters, and buttercups trailing off in decreasing numbers into the forest as the canopy closes overhead, and we see young pines and scattered mature trees stationed well out into the meadow. The fuzzy boundaries are characteristic of vegetation.
Also unlike the permanence of a master’s brush stroke, vegetation patterns change. Because plants are alive, they create a transitory scene. At any location, trees, shrubs, and herbs form a distinct pattern that changes subtly from season to season and year to year through the growth, death, and reproduction of individual plants. Patterns can also change profoundly over time following a disturbance or long-term climatic shift. However, any given landscape tends to have a limited set of vegetation types. Vegetation following fire, seasonal or periodic drought, avalanche, or disease often approaches conditions similar to those before the disturbance because of the physiological adaptations plants have.
Some changes, however, are not within the adaptive repertoire of the plants, forcing permanent shifts in the scene. New stands of vegetation may arise within a geologic blink of an eye. For example, a wet meadow may change to a dry one after a single flood event. Rapid down cutting of a stream can drop the water table below the reach of the roots of meadow plants.
Ecologists have long thought that environmental variation drives pattern in vegetation to a great degree. Differences in such important environmental factors as moisture, soil texture and chemistry, temperature, and the time since the disturbance exert powerful influences. For example, the moisture content of the soil ranging from a streamside terrace to adjacent upland slopes has a direct influence on the vegetation occupying these areas. We can view these environmental variables as gradients, and the character of the vegetation changes subtly along these gradients. Conversely, the character of vegetation may change dramatically where there is a “break” in environmental gradients, such as between a chemically harsh serpentine soil and a more benign sandstone soil. The study of how this environmental variation influences the character of the vegetation is an important part of vegetation science.
Early in the twentieth century, lines were drawn in an ideological debate over the fundamental character of vegetation. Frederick Clements (1916, 1920) metaphorically equated units of vegetation with organisms. He saw that clusters of species repeatedly associated together. He believed that at least some of these species had obligatory relationships with the community much like those of organs within the living creature. In Clements’s vision of vegetation, units had little overlap because many species were confined to a single type of vegetation. The boundaries among adjacent clusters of plants were narrow, with very little overlap of species’ ranges, except for a few widespread plants.
Henry Gleason (1926, 1939) published his ideas on the continuum view of vegetation soon after those of Clements. Gleason’s contention was that vegetation was the result of two major factors: the fluctuating arrival and departure of plants, and the equally fluctuating environment. In contrast to Clements’s view of plant associations as discrete and interdependent combinations of species, Gleason viewed species as ordered along environmental gradients in ways prescribed by the individual requirements of the species. He saw the overlap in species distribution as largely coincidental and independent. Gleason regarded the clusters of species predicted by the discrete community view of Clements and his followers as merely artifacts of the investigator’s perception or of sampling and analysis methodologies.
Studies by the end of the century supported Gleason’s individualistic view of vegetation. Smith and Huston (1989) also have suggested that competition among species for light and water, as plants grow larger, accounts for the majority of spatial and temporal structure in vegetation. However, others (McCormick and Platt 1980, Parker and Leopold 1983, Rodgers 1980, 1981) have shown that the species composition and structure of the understory is not so much a function of the species composition of the tree canopy, but the presence of shade itself.
Because of these investigations, vegetation ecologists became cautious about acknowledging the existence of plant communities. Wilson (1991) suggested that there is very little evidence to support their existence. Others (Dale 1994, Palmer and White 1994) suggested that we need definitions that are more precise. Palmer and White (1994) suggested that community ecologists should define “community” operationally and thus remove themselves from the ontological dilemma of whether or not communities exist. Keddy (1993) pointed out that a number of defined things do exist that are difficult to prove empirically (e.g., electrons, individual human beings, individual species). Yet these terms are useful.
This operational approach is the one taken today, and we hold less interest in the debate. Yes, trees migrated individualistically as they colonized new ground at the end of the last Ice Age (Davis 1981). However, the recent interest in the nature of biotic interactions indicates that they are important in structuring communities (Begon et al. 2006, Gurevitch et al. 2006). It would seem that both Clements and Gleason were right to some degree.
Ecologists have devised sophisticated techniques to define associations among plant species and their environmental correlates (Bonham 1989, Dufrêne and Legendre 1997, Gauch 1982, Kent and Coker 1992, Jongman et al. 1987, McCune and Grace 2002, Mueller- Dombois and Ellenberg 1974). Although these techniques distinguish groupings of plants following quantitative field sampling and analysis, we recognize that plant communities erected using these techniques are drawn arbitrarily. Community boundaries depend on the investigator, the methods of sampling and analysis, and the sharpness of the environmental gradients in the landscape. For this reason, we use the word vegetation type rather than “plant community” when referring to plant assemblages.
Clements (1916) also popularized the idea that a community grew, matured, died, and reproduced itself though a progression of serial stages, each depending on the previous one. This concept of succession, like his community theory, also has been challenged. Egler (1954), McCormick (1968), and others conducted studies in the eastern United States suggesting that succession depends on initial floristic composition, not serial stages. Progression from dominance by shortlived plants to that of longer-lived plants has elements of chance that can create several possible avenues as plants grow and mature. Clements’s predictable model of succession to a single persistent state is not realistic.
Forest ecologists (e.g., Oliver and Larsen 1990) replaced the term succession with a new one, stand dynamics, when talking about the way forests change with age. Their domain of interest is not the plant community in the abstract, but an existing stand of plants that have the same history. Consider the situation after logging: Seedlings establish in the bare soil after this disturbance. As they grow to be trees, the stand changes structurally and possibly floristically. You can view this change as gradual or as a set of distinct stages, often called seral states. A forested landscape, to the ecologist, becomes a matrix of one to many stands with different histories and ages.
Recent ideas applied to nonforested environments (e.g., Westoby et al. 1989, Huntsinger and Bartolome 1992) are similar. These approaches considered the pattern of vegetation as a set of states and transitions. The states represented visually distinct vegetation types—grassland, grassland with shrub seedlings, and dense cover of shrubs with little grass—all in the same area. The transitions represent the change from one state to another brought about by changes in environmental conditions, disturbance events, or by the inevitable growth of new species overtopping previously taller plants. We see the patterning as an ever-changing mosaic of vegetation types, yet these patterns repeat in the landscape at different intervals and at different locations depending on the plants’ adaptations to environmental variation and through time.
Our philosophy in this book is to treat any vegetation pattern that repeats and persists in the landscape for more than just few years as worthy of describing and defining. This most certainly includes a diverse array of post-fire and other disturbance-related types, as long as they are predictable and include diagnostic species. We find little value in attempting to classify largely ruderal or managed landscapes such as vacant lots, agricultural lands, or managed wetlands. Patterns of plant cover in such areas are usually short, persisting even with a high frequency of human-mediated disturbance such as mowing, plowing, irrigating, burning, or disking. They are altered more readily by these nonnatural processes and have a suite of species, many of which are equally capable of dominating a stand under a shifting palette of natural seasonal and yearly changes, coupled with intensive and often high-frequency unnatural disturbance. We do, however, include some of the more conspicuous types dominated by non-natives in this edition.
We have a practical need to understand how ecological theory applies to vegetation because theory helps us in our conservation and management efforts. However, the accuracy of science will never be a substitute for our aesthetic appreciation of the beauty of vegetation patterns across the landscape. It behooves us to pause and reflect.
A classification is a language created to bring order out of apparent chaos. Because it is a language, the basic goal of a classification is to solve a communication problem. A vegetation classification therefore, develops a single, commonly accepted terminology to discuss various kinds of vegetation.
However, a single classification cannot be all things for all people. Ecologists have developed numerous vegetation systems that reflect a variety of descriptive scales, philosophies, and purposes. They recognize that classifications are artificial: the units are subjective and created to meet a particular need. The user should be aware of these points when using a classification.
Reasons for developing a vegetation classification vary and can include resource inventory, land use planning, land management, conservation, illustration of ecological relationships, or the building of a framework for understanding vegetation dynamics. Likewise, the vegetation units created can vary and include functional resource management criteria, such as timber or range types; descriptions of vegetation associated with landscape units, ecological units, or animal habitats; and classifications with an emphasis on vegetation structure, floristic assemblages, or units recognizable from aerial photographs. All have value.
Any vegetation classification imposes a particular level of detail or scale. For example, maps and classifications for land use and cover published by the USDA or various California county governments list general units, such as grassland, conifer forest, scrub, or marsh. Other classifications place more emphasis on distinguishing among types of agricultural crops rather than among natural vegetation units. Wall maps of the vegetation of California (Küchler 1977, Parker and Matayas 1979, Wieslander and Jensen 1946) show gross patterns of vegetation, but these are not accurate at a fine, local, or regional scale. In contrast, a vegetation classification and map produced for a small nature preserve or a state park exhibit detail at a very fine scale.
The concept of hierarchy has proven valuable in the theoretical organization of vegetation throughout the world (Falk et al. 2006), in much the same way that biologists express relationships among plant or animal taxa. Several well-defined vegetation classification hierarchies exist. For example, the UNESCO classification (Muller-Dombois and Ellenberg 1974), the classification developed by Grossman et al. (1998), and the revised International Vegetation Classification (NatureServe 2007) as applied to the United States incorporate many levels of scale, from the individual stand to world-wide general classes of vegetation.
In traditional approaches to vegetation, the highest classes correlate with basic environmental differences (e.g., aquatic and terrestrial) that define units of classification. Lower levels are arranged around physiognomic features including plant height and life form (e.g., tree, shrub, herb), canopy cover (open to closed), and foliage characteristics (e.g., evergreen, deciduous, broadleaf, needleleaf). It is not until you move well down in the hierarchy that the classifier often uses species dominance: the formation-type level (Barry 1989a, Grossman et al. 1998), the alliance level (Braun-Blanquet 1965), and the series or type level (Daubenmire 1952, 1966, Pfister and Arno 1980). The next lower level, the association, is defined by secondary species.
The broader units of classification are particularly important for developing national, continental, or worldwide relationships. For example, the LANDFIRE program (http://www.landfire.gov/) focuses on aggregations of associations and alliances called “ecological systems” (Comer et al. 2003) to identify current fire conditions across the United States. The revised International Vegetation Classification focuses on aggregations of associations and alliances into groups up through divisions, formations, and classes (Nature- Serve 2007a) to identify ecologically related vegetation based on climatic, edaphic, biogeographic, and disturbance differences.
Our emphasis is at the alliance level. This level is best for considering vegetation at a regional and statewide level because it is based on a tangible number of floristic categories, defined by well-known plant species, some of which are widespread throughout the state. We recognize further patterns at the association level. The association is used at the local scale. State parks or small- to medium-sized natural preserve managers prefer the association level of the classification because it reflects predictable combinations of plant species that typically have more local specificity as it applies in a certain mountain range or an ecological subsection (Miles and Goudey 1997).
Hierarchical approaches to classification of natural environments exist that do not emphasize vegetation. One of the most widely used of these is the wetland and deepwater habitats (Cowardin et al. 1979). Their approach arranges features of the environment hierarchically, with the upper levels related solely to physical characteristics, such as substrate type, water chemistry, and water regime. They consider organisms only at the lowest level. In one application of this classification to central and southern California wetlands, Ferren et al. (1994) recognized hundreds of dominance types. For example, the Palustrine/Woodland and forest (Salix laevigata)/Seasonally and intermittently flooded/Fresh water/Stream banks and floodplain plants type is analogous to our Salix laevigata alliance.
Hierarchical arrangements of ecosystems, for example, the Forest Service Ecoregions project (Bailey 1980, 1995), uses synthetic ecological units of Ecoregions derived from abiotic (primarily soils, climatic zones, and landforms) factors including the Domain, Division, Province levels, the Subregion including the Section, Subsection levels, the Landtype association, and Landtype phase (Bailey 1995, 1998). California involves two different domains, four divisions, eight provinces, and 19 sections (McNab and Avers 1994). We use the section and subsection levels, as they apply to California (Miles and Goudey 1997), to describe the extent and regional variability in our alliance descriptions.