Serpentine Plant Assemblages: A Global Overview

Serpentine substrates are found in many parts of the world, but there is considerable variation in the structure, composition, and diversity of the flora they support. To place western North America in a worldwide context, this chapter provides a brief sketch of global patterns in serpentine plant life, drawing on the reviews by Brooks (1987), Baker et al. (1992), and Roberts and Proctor (1992), as well as other sources. Following this is an overview of some of the main physical factors known to cause variation in the vegetation on serpentine both at the regional and local levels. Finally, we discuss what is known about the roles of competition, fire, herbivory, and other ecological processes in shaping plant assemblages on serpentine. The availability of botanical information varies considerably around the world. In most countries where serpentine occurs, it is possible to name at least some of the plant species and vegetation types found on it. But in countries where surveys are incomplete, or where information has not been synthesized at a national or larger level, it is generally not possible to estimate the number of serpentine-endemic taxa or to describe patterns of variation within the serpentine vegetation. Indonesia, Malaysia, the Phillippines, and Brazil are particularly notable as countries with serpentine floras that are potentially rich but in need of more study. With this caveat, however, some of the major global trends can be described based on available knowledge. Flora and vegetation of selected parts of the world are summarized in table 10-1, and global contrasts between the vegetation of serpentine and other soils are summarized in table 10-2. New Caledonia and Cuba lead the world in known serpentine endemic diversity with 900+ species each, >90% of which are also endemics to these islands. Depending on elevation, rainfall, and fire history, the serpentine vegetation on both islands varies from sclerophyllous scrubland that contrasts visibly with the neighboring vegetation, to medium-stature rainforest that is not strikingly different in appearance from the vegetation growing in other soils.

Nature of Soils and Soil Development

We walk on soils frequently, but we seldom observe them. Soils are massive, even though they are porous. Soil 1m (40 inches) deep over an area of 1 hectare (2.5 acres) might weigh 10,000–15,000 metric tons. It is teeming with life. There are trillions, or quadrillions, of living organisms (mostly microorganisms), representing thousands of species, in each square meter of soil (Metting 1993). In fact, species diversity, or number of species, may be greater below ground than above ground. We seldom see these organisms because we seldom look below ground or dig into it. The many worms and insects one finds digging in a garden are a small fraction of the species in soils because the greatest diversity of soil-dwelling species exists among microscopic insects, mites, roundworms (or nematodes), and fungi. Even though individual organisms in soils are mostly very small or microscopic, the total mass of living organisms in a hectare of soil, excluding plant roots, may be 1–5 or 10 metric tons. More than one-half of that biomass is bacteria and fungi. Living microorganism biomass generally accounts for about 1%–5% of the organic carbon and about 2%–6% of the nitrogen in soils (Lavelle and Spain 2001). The upper limit of soil is the ground surface of the earth. The lower limit is bedrock for engineers, or the depth of root penetration for edaphologists. Unconsolidated material that engineers call soil can be called “regolith” (Merrill 1897, Jackson 1997) to distinguish it from the soil of pedologists and edaphologists. Regolith may consist of disintegrated bedrock, gravel, sand, clay, or other materials that have not been consolidated to form rock. Pedologists investigate the upper part of regolith, where changes are effected by exchanges of gases between soil and aboveground atmosphere and by biological activity. This soil of pedologists may coincide with that of edaphologists or include more regolith. In fact, the lower limit of soil that pedologists investigate is arbitrary, unless this limit is a contact with bedrock that is practically impenetrable with pick and shovel.

Introduction

Ultramafic, or colloquially “serpentine,” rocks and soils have dramatic effects on the vegetation that grows on them. Many plants cannot grow in serpentine soils, leaving distinctive suites of plants to occupy serpentine habitats. Plants that do grow on serpentine soils may be stunted, and plant distributions are commonly sparse relative to other soils in an area. Plant communities on serpentine soils are usually distinctive, even if one does not recognize the plant species. Because of these distinctive features, ultramafic rocks and serpentine soils are of special interest to all observers of landscapes. Geology underlies both conceptually and literally the distinctive vegetation on serpentine soils. The occurrence of special floras on particular substrates within particular regions makes rocks and soils of key significance to plant evolution and biogeography. Sophisticated interpretations of these interrelationships require a combined knowledge of geology, soils, and botany that few people possess. Even highly specialized professionals generally lack the requisite expertise in all three disciplines. The science of ecology, which in principle concerns interactions among all aspects of the environment, seldom incorporates a deep understanding of rocks and soils. Some scientists have attempted to bridge this gap through creating a discipline known as geoecology (Troll 1971, Huggett 1995), which forms the basis for our interdisciplinary exploration of serpentine rocks and soils in western North America. The term “serpentine” is applied in a general sense to all ultramafic rocks, soils developed from them, and plants growing on them. Ultramafic rocks are those with very high magnesium and iron concentrations. The word serpentine is derived from the Latin word serpentinus, meaning “resembling a serpent, or a serpent’s skin,” because many serpentine rocks have smooth surfaces mottled in shades of green to black. The distinctive chemistry of ultramafic rocks and serpentine soils restricts the growth of many plants and makes them refuges for plants that thrive in serpentine habitats, including serpentine endemics (species that are restricted to these soils) and other species that have evolved means of tolerating these habitats. Often the means of tolerance include visible adaptations such as slow growth and relatively thick, spiny foliage.

Denali-Yukon, Domain 9

The Denali-Yukon domain occupies a broad arc that, in general, follows the path of the Denali Fault along the Alaska Range and southwestward into the Yukon Territory. An ophiolite in the northwestern corner of British Columbia that is northeast of the projected Denali fault is included in this locality. A projection of the Denali fault system southwestward from the Alaska Range passes through the southwestern part of the Ahklun Mountains physiographic province, as the province was defined by Wahrhaftig (1965), to Kuskokwim Bay between the mouth of the Kuskowim River and Cape Newenham. Three mafic–ultramafic complexes on the southwestern edge of the Ahklun Mountains province are included in this domain. Glaciers covered this entire domain during the Pleistocene, and mountain glaciers and ice caps are still present at the higher elevations. Permafrost is currently discontinuous. The highest mountain in North America (Mt. McKinley, 6194 m) is in the Alaska Range, but the ultramafic rocks are all at much lower elevations. The climate is very cold throughout the domain, with severe winters and short summers. The mean annual precipitation ranges from 45 to150 cm in the Ahklun Mountains, from 30 to 60 cm in the Alaska Range, and from 30 to 75 cm, or more, in the Atlin area of northwestern British Columbia, which is in the rain shadow of the Coast Mountains. The greatest precipitation is during summers, from June or July to September or October. The frostfree period is on the order of 60–90 days, or shorter, but it may be longer in some of the Atlin area of British Columbia. Localities 9-1 through 9-3 are from Cape Newenham northeastward in the Ahklun Mountains. The ultramafic rocks in the Cape Newenham area were accreted to North America by north directed thrust faults during the Late Triassic and Middle Jurassic time. Localities 9-4 through 9-7 are in the Alaska Range. Locality 9-8 is along a projection of the Denali fault to the eastern edge of the Coast Ranges in British Columbia.

Sierra Motherlode, Domain 2

The Sierra Motherlode domain is in a series of allochthonous terranes, sometimes called the “Foothill Belt,” along the western edge of the north-northwest–south-southeast trending Sierra Nevada, adjacent to the Great Valley of California. It is a discontinuous belt from the southern Sierra Nevada, in Tulare and Fresno counties, to Butte County in the northern Sierra Nevada , but a branch within the belt is practically continuous from El Dorado County about 140 km north to Plumas County at the north end of the range. Cenozoic block faulting has lifted the Sierra Nevada and tilted the mountain range toward the west; therefore the highest elevations are on the east side of the range. Uplift is more pronounced in the southern than in the northern Sierra Nevada. Altitudes range from <200 m adjacent to the Great Valley to more than 4000 m along the crest of the central to southern part of the mountain range. The highest altitudes in the Sierra Motherlode domain are 1939 m (6360 feet) on Red Mountain and 1935 m (6335 feet) on Red Hill in Plumas County, and even higher on some of the granitic plutons that are within the outer limits of the serpentine domain. These plutons were intruded into the allochthonous terranes after the terranes had been accreted onto the continent. Much of the western slope of the northern Sierra Nevada is an undulating to rolling plateau. This plateau is a remnant from the early Tertiary when its surface was deeply weathered to produce lateritic serpentine soils with silica deposited in the subsoils and in fractures in the bedrock (Rice and Cleveland 1955, Rice 1957). The ancient plateau was capped by volcanic flows that produced a practically continuous cover in the northern Sierra Nevada (Durrell 1966). Uplift along the eastern side of the northern part of the Sierra Nevada to initiate its current relief commenced 4 or 5 Ma ago (Wakabayashi and Sawyer 2001). Since the range began to rise a few million years ago, the larger streams flowing across it have cut deep canyons up to about 600 m below the plateau.

Serpentine Soil Distributions and Environmental Influences

Serpentine soils occur in all but one of the twelve orders (Alexander 2004b), which is the highest level in Soil Taxonomy (Soil Survey Staff 1999), the primary system of soil classification utilized in this book (appendix C). They occur in practically every environment from cold arctic to hot tropical and from arid to perhumid (always wet). Thus the variety of serpentine soils is very great even though they occupy only a small fraction of the earth. Serpentine soils have been found in all states and provinces that are adjacent to the Pacific Ocean from Baja California to Alaska. They are most concentrated in the California Region, where they have been mapped in 34 counties in California and in 5 counties in southwestern Oregon. Serpentine lateritic (or “nickel laterite”) soils, which have not been mapped separately from other soils, are economically significant in California and southwest Oregon, even though they are not widely distributed in western North America. A representative serpentine soil is shown in figure 6-1. Serpentine soils, or soils in magnesic (serpentine) families, are represented in 11 of the 12 soil orders. Spodosols and Histosols in magnesic families occur only where there is a thin cover of nonserpentine materials over the serpentine materials, and there are no serpentine Andisols. Andisols contain amorphous and poorly ordered aluminum-silicate minerals, which are responsible for andic soil properties of these soils. Serpentine soil parent materials do not contain enough aluminum for the development of andic soil properties that are definitive of Andisols. Alfisols are soils with argillic (or natric) horizons having more than 35% exchangeable bases (Ca2+, Mg2+, Na+, and K+) on the cation exchange complex. Al3+ and H+ are the common nonbasic (acidic) cations on the exchange complex. The Mg2+ that serpentine soil parent materials release upon weathering keeps the basic cation status of soils high, unless they are leached intensively. Some of the soil horizon sequences are A-Bt, A-Btn, and A-Bt-Btk in Alfisols. Soils of Dubakella Series and other moderately deep Mollic Haploxeralfs with a mesic soil temperature regime are the most extensively mapped serpentine Alfisols in California and southwestern Oregon. Figure 6-1 is representative of the Mollic Haploxeralfs.

Mineralogy and Petrology of Serpentine

“Serpentine” is used both as the name of a rock and the name of a mineral. Mineralogists use “serpentine” as a group name for serpentine minerals. Petrologists refer to rocks composed mostly of serpentine minerals and minor amounts of talc, chlorite, magnetite, and brucite as serpentinites. The addition of “-ite” to mineral names is common practice in petrologic nomenclature. For instance, quartzite is a name for a rock made up mostly of quartz. Serpentinites are rocks that form as a result of metamorphism or metasomatism of primary magnesium–iron silicate minerals. This entails the replacement of the primary silicate minerals by magnesium silicate serpentine minerals and the concentration of excess iron in magnetite. “Mafic” is a euphonious term derived from magnesium and ferric that is used for dark colored rocks rich in ferromagnesian silicate minerals. “Ultramafic” is used when the magnesium–ferrous silicate minerals compose >90% of the total rock. Olivine, clinopyroxene, and orthopyroxene are the minerals in primary ultramafic rocks, with minor amounts of plagioclase, amphibole, and chromite. Ultrabasic has been used by some geologists in referring to ultramafic rocks. The most common ultramafic rocks are harzburgite, containing <75% olivine and 25% orthopyroxene; dunite, with 100% olivine; and lherzolite, which has 75% olivine, 15% orthopyroxene, and >10% clinopyroxene, with or without plagioclase. Very small amounts of chromite are present in all of the mantle ultramafic rocks (Coleman 1971). The alteration of primary ultramafic rocks to serpentine mineral assemblages is incremental due to episodic invasion of water into the ultramafic rock. It is difficult to distinguish and map the gradations from primary ultramafic rock to serpentinite. Because of this difficulty in distinction, we prefer to use the term ultramafic or serpentinized peridotite for all gradations to serpentinite. Pedologists and botanists commonly group serpentinites with primary ultramafic rocks and refer to these substrates as serpentine because all of them have similar chemical compositions. As will become apparent later, there is great variability in the mineralogical compositions of these rocks and the soils derived from them.

Northern Alaska–Kuskokwim Mountains, Domain 10

The ultramafic rocks in this domain are in the western part of the Brooks Range, the interior Alaska lowlands of the Koyukuk–Yukon Basin, the interior Alaska highlands of the Tanana–Yukon Upland, and the Kuskokwim Mountains. This domain extends east to the Seventymile River, a tributary of the Yukon River that is near the Canadian border, and presumably to the Clinton Creek area in the Yukon Territory. Although the highest elevations in the Brook Range are near 2700 m, those in the western mountains of the range are mostly <1400 m. Flatlands, hills, and low mountains dominate the Koyukuk–Yukon Basin and Tanana–Yukon Uplands, Elevations in the Kuskokwim Mountains are mostly <1000 m. Some of the mountains in uplands of the Tanana–Yukon Uplands are higher than 1600 m. Although the Brooks Range was glaciated during the Pleistocene, there was no glaciation in the Koyukuk–Yukon Basin, and only the higher elevations in the Tanana-Yukon Upland were glaciated during the Quaternary. Today, permafrost prevails throughout the Brooks Range, but it is discontinuous in the Koyukuk–Yukon Basin and Tanana–Yukon Upland and in the Kuskokwim Mountains (Ferrians 1965). Loess is extensive in the basins of interior Alaska and at lower elevations in the Kuskokwim Mountains, with some deposits >60 m thick (Péwé 1975). The climate is very cold throughout the domain, with severe winters and relatively short summers, although mean maximum summer (July) temperatures are >20°C (up to 24°C or 25°C) in the interior basins. With latitudes from 61°N to 68°N, days are very long during summers and very short during winters. The mean annual precipitation is 15–45 cm, with the greatest precipitation during summers. Even though the precipitation is low, the climate is not arid because evapotranspiration is limited by short and relatively cool summers. The freeze-free period is on the order of 60–90 days. The northern and interior Alaska ultramafics (serpentine) consist of Paleozoic and Mesozoic thrust slices emplaced onto Precambrian and Paleozoic marine sediments. They all belong to well-defined belts and are related to low-angle thrust faults or to later high strike–slip faults.

Gulf of Alaska, Domain 8

The Gulf of Alaska domain extends eastward from Kodiak Island across the Kenai Peninsula, around the Chugach Mountains, and beyond Tonsina, curves southward across the glacier-covered St. Elias Mountains. The southeastern segment of the domain is west of the Canadian Coastal Mountains and includes the coastline and islands in southeastern Alaska. Ultramafic rocks occur sporadically in this region from British Columbia northwestward through southeastern Alaska and around the Gulf of Alaska to the Kenai Peninsula. Most of this domain is mountainous, especially in areas where there are ultramafic rocks. Elevations range from sea level to >5000 m, although ultramafic rocks are not found at the highest elevations. This area was glaciated during the Pleistocene. Many glaciers persist at the higher elevations, and some descend to sea level. The present climate ranges from cold to very cold and from humid to very humid, or perhumid. Mean annual precipitation ranges from 30 or 40 cm on the north side and west end of the Chugach Mountains, to >400 cm along the coast around Prince William Sound and in southeastern Alaska. Southeastern Alaska is a humid-to-perhumid area with dense forests at lower elevations that grade upward into alpine areas. The north side of the Chugach Mountains is drier, but still humid. Precipitation exceeds evapotranspiration in all months of every year. The frost-free period ranges from no more than a few days or weeks at the higher elevations to about 220 days in sheltered areas near sea level in southeasern Alaska. Some of the highest snowfall in North America is in this domain, and ice caps persist in some of the higher mountains with many glaciers flowing into the sea. Serpentine rocks of the Gulf of Alaska domain occur in both ophiolites and in concentric bodies, and some are from the roots of volcanic arcs complexes. West of the Chatham Straight fault, a discontinuous belt of ultramafic bodies extends for >1600 km from Kodiak Island across the Kenai Peninsula, and around the Chugach Range, arching southward as far as Baranof Island (Burns 1985).

Northern Cascade-Fraser River, Domain 7

The Northern Cascade–Fraser River domain conforms to the Northern Cascade Mountains physiographic province in northwestern Washington and southern British Columbia, the San Juan Islands between the southern tip of Vancouver Island and the Northern Cascade Mountains, and much of the Interior Plateau province of British Columbia. The thread that connects these areas is the north–south Straight Creek–Fraser River fault system that runs through the Northern Cascade Mountains and northward along the Fraser River. The localities of domain 7 are along faults that branch off from this major fault system. The Northern Cascade Mountains are indeed mountainous, and the Interior Plateau of British Columbia is an area of dissected plateaus and scattered mountains. The Fraser River flows northwest in the Rocky Mountain Trench, which separates the North American craton on the northeast from accreted terranes on the southwest; then it turns around the northwest end of the Cariboo Mountains to the Interior Plateau. In the Interior Plateau, the Fraser River flows from Prince George south about 500 km to the Northern Cascade Mountains before turning westward toward the Pacific Coast. The northern part of domain 7 is in that part of the Fraser River basin, including tributaries northwest of Prince George, which is in the Interior Plateau province. Low, hilly terrain dominates the San Juan Islands. All of these areas in domain 7, except the Ingalls complex on southeast margin of the Northern Cascade Mountains, were covered by the Cordilleran ice sheet during the last stage of the Pleistocene glaciation, leaving <15 ka years for soil development on the current ground surfaces. Although alpine glaciers formed in the southeastern margin of the Northern Cascade Mountains, they did not cover all of the soils, allowing some of them longer time for development. Elevations in domain 7 range from sea level on San Juan Islands to mostly in the 600–1500 m range on the Interior Plateau of British Columbia, and up to 4392 m on Mt. Rainier in the Northern Cascade Mountains.