Chapter 4: Historic Conditions

"Evaluating the status of existing ecosystems requires a standard or set of reference points that characterize sustainable ecosystems."

Kaufmann et al. (1994)

Thousands of years of human occupation preceded the first written accounts, paintings, and photographs of Southwestern landscapes. We can only guess what the earliest occupants thought of their landscape. We know that prehistoric peoples altered vegetation by farming and burning; they shaped the land by terracing, field leveling, and stream channeling. Archeologists are just beginning to understand the extent of prehistoric occupation and do not know how prehistoric activities over the centuries have resonated across the landscape. For example, we know little of how burning or nutrient depletion by repeated cultivation affects vegetative patterns hundreds or thousands of years later. We need to view the whole notion of naturalness in the Southwest through the filter of prolonged and sometimes intensive human occupation. In the present and historical past, it is often very difficult to distinguish between natural and human-caused effects. For instance, what did cause widespread arroyo cutting in Southwestern deserts? Was it over-grazing, climate change, or a combination of both (Hastings and Turner 1965, Cooke and Reeves 1976)?

The period used to characterize historic conditions in this chapter is that prior to the Mexican-American War of 1848. This date is selected as a central point between the 1845 annexation of Texas and the 1853 Gadsden Purchase. These events mark the political transition from Hispanic to American sovereignty in the Southwest and its opening to new and increasing waves of settlement. The view through time's fuzzy lens becomes a little clearer as we approach the present. After the Civil War, members of various expeditions and surveys began to photograph the Southwestern landscape. General agreement between 18th century photographs and 16th century written accounts indicate that the landscape had maintained some degree of consistency over those centuries. If this agreement were valid, then photographs taken in the 1880s (and possibly as late as the 1920s for the Colorado Plateau) could represent the general forest character during at least the last few centuries. The consistency of forest attributes in early Colorado Plateau photographs supports this hypothesis (Hastings and Turner 1965, USDA Forest Service photographs 1901-1968³). There is, however, one note of caution in use of old photos to characterize the past. Artistic bias and commercial interests may have favored certain landscapes and views over more common ones; historic photographs may not represent a random sample of early landscapes.

CLIMATE

A description of past forest conditions in the Southwest begins with a review of climate history. Although regional climates persist for centuries, they do change and vegetation responds on a similar scale (Delcourt et al. 1983). The forest ecosystems we see today or in hundred-year-old photographs are the products of species evolution and migration over aeons of time on a constantly shifting landscape driven by changes in climate. From a broader perspective and context, the climate during the few centuries before 1848 was unique.

Climates change at a variety of scales (Delcourt et al. 1983). Long-term, persistent trends in temperature and humidity determine the extent and location of the various life zones, the elevation at which one biotic community replaces another. Shifts from one climatic regime to a new pattern can be abrupt. Evidence from Greenland ice cores suggests the 1300-year cold spell of the Younger Dyras (11,200–10,000 B.P.) ended over a period of only a few years (Betancourt et al. 1993). Short-term fluctuations in the order of years to decades determine drought cycles, fire frequencies, and pulses of tree reproduction. The Southwest is strongly influenced by oscillations in the Pacific ocean-atmosphere system; years of El Nino bring increased annual precipitation (but less rain in summer) and years of La Nina bring the opposite (Betancourt et al. 1993).

Data from geology, paleobotany, and dendrochronology studies at Potato Lake and Chaco Canyon permit a reconstruction of the climate history of the Southwest. Potato Lake (Anderson 1993) is a high elevation site (7,500 feet) in the Arizona–New Mexico Mountains Province. In the mid-Wisconsin Period (35,00–21,000 B.P.), the area was dominated by mixed conifer species; in the late-Wisconsin (21,000–10,400 B.P.) by spruce alone; and for the past 10,400 years by ponderosa pine. The Chaco Canyon and San Juan Basin (Betancourt et al. 1993) is a lower elevation region in the Colorado Plateau Province. Before 8,000 B.P. a relatively cold and moderately wet climate prevailed; the canyons contained a mixed conifer forest and the mesa tops a cold desert steppe. During the Altithermal Period (8,000–4,000 B.P.) pinyon and juniper migrated into the area and replaced the mixed conifer forests; warm desert grasses replaced the sagebrush of the cold desert steppe. The cause of these vegetation changes is thought to have been the arrival under generally warmer conditions of a monsoonal circulation with warm wet summers. A cooler and dryer Neoglacial Period lasted from 5,000 to 2,000 B.P. The climate of the past 2,000 years is considered modern (Cartledge and Propper 1993), but it includes several notable, global events–the Medieval Warm Period from 1000 to 1350 A.D. and the Little Ice Age from 1450 to 1850 A.D. (Nielson 1986). In the Southwest, higher average summer temperature and precipitation persisted from 950–1130 A.D. and prolonged summer droughts occurred from 1130–1180 A.D. (Cartledge and Propper 1993). Cycles are not only in the past. In 1996, Arizona endured one of the worst droughts since 1904; tree damage was detected on 57,000 acres of Federal forests (USDA Forest Service 1996a).

FIRE

Both lightning and human-caused fires, once started, could burn until extinguished by rain, or until they ran out of fuel (typically when they reached an area that had recently burned). Fires could burn for months and cover thousands of acres (Swetnam 1990, Swetnam and Baisan 1996). As a result, most forest stands (except spruce–fir) burned every 2 to 30 years as low-intensity, area-wide fires. Fire reconstructions in the Pinaleno Mountains of southern Arizona demonstrate that pre-settlement mixed conifer forests could have burned as frequently as the ponderosa pine forests (Grissino-Mayer et al. 1995). With greater moisture levels but heavier fuel loads, spruce fir forests burned much less frequently but at high, stand-replacing, intensity (Grissino-Mayer et al. 1995, Veblen et al. 1994). Research by Cable (1975), Dieterich (1980, 1985), Grissino-Mayer et al. (1995), Leopold (1924), and Weaver (1951) establish the range of pre-settlement fire frequencies for Southwestern forest communities (Table 4.1). The role of native peoples in modifying fire regimes of the interior West is examined by Arno 1985, Gruell 1985, Barrett 1988, Savage and Swetnam 1990, and Veblen and Lorenz 1991. Native cultures used fire for a variety of purposes (Pyne 1982, Phillips 1985). In addition to incidental burning, there is ample evidence that fires were intentionally set for increasing desired plant species, improving wildlife habitat, driving game animals, and clearing transportation routes. Pinchot (1947) describes a scene where:

"We looked down and across the plain. And as we looked there rose a line of smokes. An Apache was getting ready to hunt deer. And he was setting the woods on fire because a hunter has a better chance under cover of smoke."

In many areas of the West, American Indians altered succession using fire (Gruell 1985). In the Southwest, however, the historic fire regime does not depend on native burning (Swetnam and Baisan 1996). Lightning is common in many parts of the Southwest during periods of high fire hazard; these rates and patterns would have prevailed in earlier times as well (Schroeder and Buck 1970). Lightning ignition alone is sufficient to produce the fire frequencies revealed in various fire-scar chronologies. The effect of native peoples on fire in the Southwest would probably have been very site- and time-specific (Swetnam and Baisan 1994).

The ecological role of fire in Southwestern forests has always been significant. Where fire was more frequent, forest communities even developed a dependency on fire as a mechanism for ecosystem regulation (Wright and Heinselman 1973). Where fire was infrequent such as in the spruce–fir forest, fire was still a major abiotic factor, but its effects and integration into the ecosystem had a very different character. Like all disturbances (such as grazing, fuelwood collection, mortality from insects and diseases) fire affected species composition, the amount, distribution, and proportion of living and dead biomass, and various ecosystem functions (e.g., nutrient cycling).

DEMOGRAPHICS AND LAND USE

For more than 12,000 years, humans have been an integral component of Southwestern forest ecosystems; but the archaeological record is fragmentary and poorly sampled. Precise estimates of regional populations in prehistoric times do not exist, although information on general population trends are evident (Dean et al. 1994). For at least 10,000 years, a sparse population dependent on wild plants and animals occupied the Southwest. Beginning around 300 A.D., the population in certain areas of the Southwest increased with a shift to greater reliance on domesticated plants and development of more permanent settlements. By 600 A.D., the increase was regionwide. With the growth of larger, more complex communities throughout the region, the population peaked at perhaps 130,000–150,000 in the 11th to 13th centuries (Dean et al. 1994). Beginning about 1300 A.D., many areas of the Southwest were abandoned; and the regional population probably declined, although to what degree is uncertain. Anasazi populations appear to have re-aggregated into large pueblos along the Rio Grande and in the Zuni and Hopi areas. The Hohokam and Salado populations, however, appear to have dispersed. Such shifts in population and settlement had probably been repeated throughout prehistory in various ways and at various scales as societies responded to changes in climate, resource availability, and political, economic, and social pressures (Cordell 1984, Gumerman 1988, Tainter and Tainter 1996). These fluctuations in late prehistoric times, however, were dwarfed by the major reduction of native populations that began with the arrival of Spanish colonists.

Impacts of prehistoric populations are thought to have been minimal or highly localized until the 11th and 12th centuries. At that time, farming, fuelwood cutting, and game hunting greatly increased around new, dense settlements. There is evidence in the Phoenix Basin that main irrigation canals during the Classic Hohokam Period (1200–1400 A.D.), totalled more than 500 kilometers (Spoerl and Ravesloot 1994) and that irrigated fields covered 150 square kilometers (Nicholas and Neitzel 1984). In other areas where dry farming and floodwater farming were practiced, similar intensive use and alteration of the landscape is apparent (Fish and Fish 1992, Lang 1995). Impacts on the environment in late prehistoric times, along with drought, crop failure, and population stress, probably contributed to local and regional cycles of settlement abandonment, and relocation.

In the Southwest, as elsewhere in the New World, the arrival of Europeans had a catastrophic impact on native populations. Epidemics, starvation, hostilities, subjugation, and relocation devastated native peoples. Pueblo populations in New Mexico, for example, were reduced from perhaps as many as 60,000 in the 1500s to fewer than 7,000 by 1706 (Schroeder 1979). Some groups disappeared entirely, including the Piro of central New Mexico and the Sobaipuri of southern Arizona. The degree to which the effects of disease may have preceded the actual arrival of Spanish colonists is unknown.

The Spanish population during the Colonial Period remained relatively small, due to their dependence on irrigation agriculture and hostile relations with the Apaches, Navajos, Utes and Comanches. The Spanish population in New Mexico is estimated to have been only 2,500–3,000 in 1680, and was perhaps no more than 20,000–25,000 by the late 18th century (Simmons 1979). Because the population was small except along certain permanent streams, regional impacts during the Colonial Period were not much different from those of late prehistoric times. The notable changes were the shift to more intensive irrigation and the introduction of new crops and animals. The population grew during the later Mexican Period; sheep grazing increased significantly; and mining expanded in certain areas.

TIMBER AND FUELWOOD RESOURCES

Prehistoric populations used wood for fuel, tools, and construction. In late prehistoric times, pinyon and juniper were locally depleted in some areas (Bahre 1991, Betancourt et al. 1993, Cartledge and Propper 1993, Kohler 1992, Stiger 1979). As many as 200,000 ponderosa, spruce, and fir beams were used for construction of the 10 major puebloes and kivas at Chaco Canyon (Betancourt et al. 1986). At least near large settlements, impacts on adjacent upland forests may have been significant.

In historic times prior to 1848, only small-scale utilization of timber resources was possible because of technological and transportation limitations. Native American and Hispanic populations used the forests for lumber in domestic construction and for firewood. Woodland and riparian forests were affected, especially near areas of population growth. The riparian bosque of the Middle Rio Grande Valley, for instance, had been virtually eliminated by Puebloan and Hispanic farmers before 1848 (Abert 1848a, Wozniak 1987). But impacts on upland forests were probably negligible before the introduction of commercial logging, mining, and railroads in the late 19th century.

FOREST INSECTS AND PATHOGENS

For millennia, trees of Southwestern forests have been host to numerous species of herbivorous insects, pathogenic or saprophytic fungi, and parasitic plants. These species co-evolved with their hosts as members of dynamic, interacting communities. Through their ability to cause widespread tree mortality, defoliation, decay, or deformity, some of these species achieved significant ecological importance as disturbance agents. Along with fire, these agents are among the more important regulators of forest density, composition, and structure. Forest conditions in turn affect the distribution and reproduction of forest insects and pathogens. Directly and indirectly, these species interact with other members of the ecological community influencing various ecosystem processes, providing food and creating habitat for other organisms, affecting nutrient recycling, and influencing fire behavior.

The species of primary interest in the Southwest include bark beetles, several species of defoliating insects, dwarf mistletoes, and root decay fungi. Bark beetles and defoliators are usually present in low populations, but they will periodically increase to outbreak levels. Although populations of dwarf mistletoe and root decay fungi fluctuate, their rates of change are much slower. These species, however, are very persistent and affect forests annually rather than periodically.

Descriptions of previous outbreak patterns for insects or distribution and abundance of pathogens are developed from inference or observation. Details for prior disturbance regimes of insects and pathogens may be inferred if one assumes their fundamental biology (e.g., host preferences) has not changed and if one has data on past climate and vegetation. Observations on the distribution and severity of outbreaks prior to the past few decades are limited to early written reports and photographs and to later reports of forests little affected by harvest, grazing, and fire. Some characteristics of prehistoric outbreaks can also be determined by various reconstruction techniques (e.g., paleobotany and dendrochronology). The host and environmental requirements of native insects, fungi, and parasitic plants has probably changed little over the recent centuries, so it is reasonable to expect that where and when conditions were suitable, these disturbance agents would have been active. Systematic surveys and reporting of insect outbreaks and disease occurrence only began in recent decades (unpublished reports⁴ and USDA Forest Service 1972). As mentioned previously in reference to old photographs, the early reports must be interpreted with caution, not because they are inaccurate, but early entomologists may have viewed forests with different objectives and values ("conceptual filters.")

Bark Beetles

Numerous species of bark beetles attack and kill trees (Furniss and Carolin 1977). Bark beetles generally have a narrow host preference within several related genera or a single genus; within portions of a beetle's range only a single host species may be available (Wood 1982). In the Southwest, the most important bark beetles on ponderosa pine are the roundheaded pine beetle (Dendroctonus adjunctus), western pine beetle (D. brevicomis), mountain pine beetle (D. ponderosae), pine engraver (Ips pini), and the Arizona fivespined ips (I. lecontei). The Douglas-fir beetle (D. pseudotsugae) on Douglas-fir, the fir engraver (Scolytus ventralis) on white fir, the spruce beetle (D. rufipennis) on Engelmann spruce, and the western balsam bark beetle (Dryocoetes confusus) on subalpine fir are also important bark beetles. Successful attack usually leads to rapid tree death, but if attack is restricted to only a portion of the bole, top-killing or strip-killing may occur (Stark 1982). At low population levels, bark beetles are usually restricted to scattered individual trees that have been weakened by disease, old age, or competition and to fresh logs and slash caused by windthrow or snow breakage. In outbreaks, small groups of killed trees eventually merge into large stands of dead trees.

Bark beetles affect and are affected by the forest community in numerous ways. By selectively killing trees of certain sizes and species, bark beetles change tree density, species composition, and size structure of the forest (Schmid and Frye 1977). Extensive and severe outbreaks can increase fire hazard (Martin and Mitchell 1980). The beetles, their associates, and successors provide food for insectivorous birds (especially woodpeckers); the resulting snags provide habitat for cavity-dependent species. Changes in forest conditions brought on by beetle-caused mortality modify the environment for numerous other species. Principal factors that influence bark beetle outbreaks are susceptible host population, weather, and natural enemies. Factors that lower tree resistance, such as poor site, overcrowding, drought, injury, and disease, favor outbreaks. Depletion of suitable hosts, extreme cold temperature, and natural enemies (insect predators and parasites, fungal diseases, and birds) contribute to population declines.

The earliest published reports of bark beetles in the Southwest date from the early 1900s; information prior to 1848 is scarce. In the early 1900s, entomologists from the USDA Bureau of Entomology conducted the first detailed investigations on various species of Dendroctonus beetles in Arizona and New Mexico (Hopkins 1909). From their reports, it appears that large outbreaks occurred in certain forest types and regions but were rare or insignificant in others.

Hopkins (1909) reports an outbreak of spruce beetle on the slopes of Sierra Blanca Peak, south central New Mexico. Baker and Veblen (1990) use historic photos and dendrochronology data to reconstruct disturbance regimes. They determined that the spruce beetle has been a major disturbance agent, comparable to fire, from central New Mexico to Colorado since the 19th century. Because fire suppression and logging have had less effect on spruce–fir forests than other communities, disturbance regimes observed today are probably a good reflection of what they had been in earlier historic times.

Hopkins (1909) notes that levels of pine beetle activity were less in the Southwest than he had observed in other Western regions. Woolsey (1911) concurs that mortality from the mountain pine beetle was less continuous and extensive in Arizona and New Mexico than elsewhere. These are curious reports considering that at the time (early 1900s) the Southwest was experiencing a severe drought which ought to have stressed trees and made them susceptible to attack. On the other hand, even today, forests of large, widely-spaced pines and few understory trees rarely support outbreaks.

The exception to observations of limited pine beetle activity in the Southwest is found on the Kaibab Plateau of northern Arizona. Blackman (1931) provides evidence for early outbreaks of mountain pine beetle in ponderosa pine. This section of the Colorado Plateau Province seems to have had a long history of large outbreaks by the mountain pine beetle. The largest outbreak occurred between 1917 and 1926 and killed about 12 percent of the ponderosa pine growing stock. Five earlier outbreaks dating back to 1837 were identified, but neither their extent nor magnitude are well documented. Why pine bark beetles were so active here and not so elsewhere is an intriguing question.

Defoliating Insects

Western spruce budworm

The western spruce budworm (Choristoneura occidentalis) feeds on foliage of true firs, Douglas-fir, and spruce throughout the western United States. In the Southwest, its principal hosts are white fir and Douglas-fir (Linnane 1986). Larvae feed primarily in buds and on foliage of the current year. Complete defoliation occurs when outbreaks persist for several years. Sustained heavy defoliation results in decreased growth, tree deformity, top-killing, and death. Stand-level effects include changes in stand structure and composition. Tree mortality is generally more prevalent in the smaller, suppressed, understory trees, so outbreaks result in fewer understory trees and increases in the average diameter. Species composition shifts to nonhost or less susceptible species. In mixed-species stands where true fir, spruce, or Douglas-fir are climax species, budworm outbreaks increase the relative abundance of early seral species such as ponderosa pine or southwestern white pine (Wulf and Cates 1987). In stands with only late seral hosts, outbreaks are a natural thinning from below, removing understory trees and stimulating growth of overstory trees.

In addition to the direct effects on forest trees, western spruce budworm affects other members of the forest community. Twenty-six species of birds are known to feed on budworm. Some of these consume large numbers of budworm, and their populations may increase in outbreak areas (Garton 1987). Outbreaks of Douglas-fir beetle sometimes follow after those of the western spruce budworm.

Swetnam and Lynch (1989, 1993) use dendrochronology to reconstruct the long-term, regional outbreak history of western spruce budworm in northern New Mexico (Southern Rocky Mountain Province). Nine regional outbreaks are identified from 1690 to 1989, with a periodicity of 20 to 33 years, and duration within stands of approximately 11 years. Most stands, including one over 700 years old, had endured multiple infestations, suggesting that Douglas-fir and budworm may coexist in the same stands over long periods. Budworm activity tends to coincide with increased spring precipitation. Budworm history and behavior in other Southwestern provinces is expected to be different, but research in these areas is not complete.

Western tent caterpillar

The western tent caterpillar (Malacosoma californicum) is a native insect that feeds on the foliage of aspen and is an important defoliator in the Southwest. Outbreaks occur sporadically and can result in extensive defoliation, growth loss, top kill, or even mortality. Outbreaks typically persist in an area for several years and flare up in one stand and then another. In a few areas, however, there have been repeated sustained outbreaks (Jones et al. 1985). Outbreaks eventually subside from a variety of biotic factors, particularly natural enemies such as viruses, insect parasites and predators, and birds. The most important control is a nucleopolyhedrosis virus which affects larvae (Furniss and Carolin 1977). The authors are unaware of any reports of the incidence or activity of this insect prior to 1848, but if aspen were more extensive then, outbreaks could also have been more common.

Root Decay Fungi

Root diseases are common throughout the Southwest in many stands of mixed conifer and spruce–fir forests and some pinyon or ponderosa pine stands (Wood 1983). The most important root diseases are caused by the decay fungi Heterobasidion annosum (affecting ponderosa pine and white fir), Phaeolus schweinitzii (primarily affecting Douglas-fir), Inonotus tomentosus (affecting spruce) and Armillaria spp. (affecting nearly all species). These fungi injure trees by decaying and killing roots. Spread occurs by wind-disseminated spores which infect through basal wounds, through root contacts between healthy and infected trees, or through rhizomorphs (only for Armillaria). These fungi can persist for decades in the roots of stumps and snags and infect susceptible regeneration through root contact (Shaw and Kile 1991, Otrosina and Cobb 1989, Tkacz and Baker 1991). Root disease infection reduces growth and survival and increases risk of mortality by bark beetles or windthrow.

Although some seedling and sapling trees are killed by root disease fungi, their principal ecological effect is through the death of canopy trees, either as single individuals or groups in slowly expanding patches. Disease centers persist for hundreds of years and appear as openings with progressively more recently-killed trees at the edge. Some regeneration may establish within a disease center, but these trees only escape mortality for a few years. Because there are species differences in susceptibility and tolerance, affected stands may exhibit species conversion (even from trees to brush). Increases in canopy diversity greatly impact habitat quality; whether the change benefits or harms a species depends on its individual requirements.

Although the authors know of no early reports describing the distribution and extent of root disease centers in the Southwest, their longevity and ubiquity suggest that root disease fungi have always been important disturbance agents, especially on more mesic sites (Wood 1983).

Dwarf Mistletoes

The dwarf mistletoes (Arceuthobium) are highly specialized dicotyledonous parasites of conifers (Hawksworth and Wiens 1996). Most conifers species in the Southwest are parasitized by one or another species of Arceuthobium. Because of their abundance and severe damage to infected trees, the most important dwarf mistletoes are the southwestern dwarf mistletoe (A. vaginatum subsp. cryptopodum) on ponderosa pine and the Douglas-fir dwarf mistletoe (A. douglasii) on Douglas-fir. Mistletoes acquire their water, mineral nutrients, and carbohydrates from a living host (Tocher et al. 1984), thereby reducing and re-allocating host growth. Many species of mistletoe, including the southwestern dwarf mistletoe and the Douglas-fir mistletoe, induce proliferation of dormant buds, localized swellings, and retention of infected branches; leading to the formation of distinctive witches brooms. Although intensification within an infected tree is slow (Geils and Mathiasen 1990), survival is greatly reduced (Hawksworth and Geils 1990).

Dwarf mistletoes affect wildlife habitat both directly and indirectly (Hawksworth and Wiens 1996). Mistletoes provide food, foraging sites, and nesting for numerous species; structural changes from brooms, snags, and openings (Parmeter 1978, Tinnin 1984) benefit the abundance and richness of nesting passerine birds (Bennetts et al. 1996).

The spread and intensification of dwarf mistletoe are affected by numerous host, stand, and environmental factors. Site quality, host vigor and age, stand density, composition, and structure are several of the more important factors (Parmeter 1978) in determining the rate of mistletoe increase. In the historic period, fire had been a significant factor in determining mistletoe distribution and persistence (Alexander and Hawksworth 1976). Although severe crown fires can sanitize an infested stand, partial burns leave scattered infected seed trees and insure early re-infection of the regeneration. Increased fine fuels and brooms on infected trees provide a fuel ladder to carry ground fire into the crown, thereby leading to complete, stand-replacing fire. Even in low-intensity fires, mistletoe reduces the survival of infected trees (Harrington and Hawksworth 1990).

Although there is little information on the previous abundance of dwarf mistletoes, they were probably already well distributed throughout the forests of the Southwest by historic times. Paleobotanical evidence supports the hypothesis that these parasites have been in western North America since the Miocene Epoch (Hawksworth and Wiens 1996). Because spread is relatively slow and long-distance dispersal rare, the extent of mistletoe distribution in the historic period is probably well reflected by the current distribution. Based on present understanding of mistletoe ecology (Parameter 1978) and evidence of previous forest conditions and fire frequency, one can infer that mistletoe abundance may have been lower in the historic period.

HISTORIC CONDITION BY FOREST COMMUNITY

Woodlands

Prior to 1848, many of the areas now occupied by dense woodlands were predominantly open, diverse communities of trees, shrubs, and perennial grasses and forbs. However, there were dense woodlands reported early in the 19th century. Abert (1848a, 1848b), for example, describes a trip between Santa Fe and Taos which began in a pinyon woodland with no grass. From the plateau east of the Rio Grande, he entered the canyon at Embudo where he reports little pasturage and that residents raised goats because there was insufficient vegetation for cattle. Leopold (1951) compares 20 pairs of photographs in stands of pinyon–juniper from nine different geographic localities. The earliest of the pairs were taken between 1895 and 1903; the latter were taken between 1937 and 1946. The number of trees increased in 7 pair, remained unchanged in 10 pair, and decreased in 3 pair.

Woodlands consist of dispersed groups of pinyon, juniper, or oak; the areas between tree patches may be mostly bare or covered by sparse litter, shrubs, or grasses. The pattern of tree patches is strongly influenced by ecosystem conditions and processes both below ground and above. Variations in soil depth, nutrients, and microbes interact with seasonal annual drought, plant competition, fire, grazing, and insect–pathogen attack (Gehring and Whitham 1995, Klopatek et al. 1990, Leopold 1924). In the historic period, native use of woodlands for timber and fuelwood had a significant effect (Betancourt et al. 1986).

Ponderosa Pine

Numerous documents (e.g., Biswell et al. 1973, Brown and Davis 1973, Cooper 1960) refer to historic ponderosa pine stands as open, parklike, and with a vigorous and abundant herbaceous understory. Captain Sitgreaves in 1854 describes an apparently typical ponderosa pine scene where "the ground was covered with fresh grass, and well timbered with tall pines". Photographic and written records of historic forest conditions and archaeological reconstructions suggest that the characteristic vegetation was a grass matrix with individuals, clumps, and stringers of large and variously-sized trees of almost exclusively ponderosa pine.

An area now within the Coconino National Forest is described in a U.S. Geological Survey (1904) report as:

"A yellow-pine forest, as nearly pure as the one in this region, nearly always has an open growth, but not necessarily as lightly and insufficiently stocked as in the case in this forest reserve. The open character of the yellow-pine forest is due partly to the fact that the yellow pine flourishes best when a considerable distance separates the different trees or groups of trees. It is very evident that the yellow-pine stands, even where entirely untouched by the ax, do not carry an average crop of more than 40 per cent of the timber they are capable of producing ... The yellow-pine forest in the reserve is, broadly speaking, a forest long since past its prime and now in a state of decadence ... Apparently there has been an almost complete cessation of reproduction over very large areas during the past twenty or twenty-five years (due mostly to sheep use), and there is no evidence that previous to that time, it was at any period, very exuberant."

Although the popular early descriptions of the ponderosa pine forest call attention to the parklike stands, there are also descriptions which refer to dense cover (Woolsey 1911). An accurate picture of the pre-settlement ponderosa pine forest would most likely describe a mosaic not only with an open, grass savanna and clumps of large, yellow-bark ponderosa pine, but also with a few dense patches and stringers of small, blackjack pines (young ponderosa pine). Ponderosa pine naturally regenerates rarely, but then reproduces with an over abundance of seedlings and a high rate of juvenile mortality (Pearson 1931). The large yellow-bark ponderosa pine of late 1800s were probably survivors that emerged from dense patches established during rare episodes of successful reproduction (climatically unusual periods of high moisture and infrequent fires). These patches would have provided needed cover for not only various wildlife species (e.g., wild turkey, Meleagris gallopava) but also the conditions for mistletoes and bark beetles to persist and even locally flourish.

The typical climate over the several centuries prior to 1848 and the development of a fire-dependent vegetation reinforced a fire regime of frequent, low-intensity burns (Covington and Moore 1994b). On an area-wide basis, surface fires burned within the montane zone where ponderosa pine is either climax or seral every 4.8 to 11.9 years (Weaver 1951). Fires of this frequency were sufficient to normally prevent reproduction by ponderosa pine or other species of the mixed conifer community. These fires, however, encouraged development of grassy understories and retention of large, open-grown ponderosa pine.

The typical climate of the ponderosa pine zone includes an adequate, annual amount of moisture for good vegetative growth and conditions favorable for frequent early summer fires (Harrington and Sackett 1992). Winters are relatively mild (average slightly above 30° F) and precipitation as snow saturates the soil (Schubert 1974). Rainfall minimums occur in May and June (some areas receive less than 0.5 inch). The spring dry season is accompanied by increasing air temperatures, low humidity, and persistent winds. The drought is broken in early to mid-July with development of almost daily thunder and lightning storms; July and August are the wettest, warmest months. A second dry season occurs in the fall. This climatic pattern is particularly conducive for development of a pine-grass savanna maintained by frequent surface fires.

Open stands of ponderosa pine under a frequent fire regime are capable of supporting a productive understory and associated grazing populations. Clary (1975) reports that open pine stands can produce at least 200 to 300 pounds per acre of herbaceous material; Cooper (1960) estimates that production could exceed 1,600 pounds per acres in frequently burned stands. These high levels are the result of surface fires which increase nutrient cycling and reduce competition from woody reproduction. Needle cast and litter from the previous year's grassy and herbaceous growth form a highly flammable fuel that is easily dried out in the spring and ignited in the early dry lightning storms (Pyne 1982). Because fires are frequent, large amounts of woody fuel do not accumulate and crown fires are uncommon. These frequent, surface fires kill small trees, but the still dormant grasses and forbs survive, and large trees escape damage because of their high crowns and thick barks (Biswell et al. 1973). These forests (Biswell 1972, Cooper 1960, Hall 1976, Weaver 1947) support elk and deer as the dominant grazers; disease, predators, and other population regulation mechanisms keep vegetation and herbivores in balance.

The more dense and younger stand structures of the historic ponderosa pine forest were the result of special circumstances in the interaction of climate and site. Even though ponderosa pine reproduction was rare, there were occasional wet cycles as long as 15 to 20 years without fires when ponderosa pine could regenerate (Swetnam and Dieterich 1985). The regeneration cycle required seed production, establishment, and survival to an age at which the young tree could successfully compete and endure surface fires. In the historic period, most large trees were killed by lightning (and associated fire), dwarf mistletoe, bark beetles, windthrow, or senescence. When single or small groups of trees died and fell, they were inevitable consumed by surface fires. This more severe, but localized, fire treatment produced mineral soil seedbeds and reduced grass competition, thereby creating a favorable microsite for establishment (Cooper 1960). Within these severely burned microsites with little competition and fuel, seedlings could survive, grow, and develop their competitive ability and resistance to fire. Replication of this pattern within the pine-savanna resulted in an uneven-aged forest composed of small, relatively even-aged groups (Cooper 1960).

Mixed Conifer

Early descriptions of the mixed conifer forest indicate they included a variety of conditions depending on the time since and the severity of the most recent burn. Lang and Stewart⁵ describe the mixed conifer forest on the North Kaibab Plateau (Colorado Plateau Province) in 1909. Although the date is later than 1848, this particular region had been only sparsely settled by that time. They describe most mature Douglas-fir (as well as white fir and blue spruce) as "deteriorating"; they probably mean these trees were decayed, had poor crown form, broken tops, and hollow bases typical of repeatedly fire-damaged trees. Lang and Stewart also note that Douglas-fir regeneration was "healthy and vigorous"; and often dense stands of pole-sized trees covered large areas, especially on more mesic sites and under aspen. The older stands had probably survived numerous, light fires. On xeric or warm–dry sites (white fir and Douglas-fir habitat types, and those with seral ponderosa pine) mixed conifer stands burned about every 5 to 12 years (Weaver 1951). On mesic or cool–moist sites (spruce–fir habitat types) in the White Mountains of Arizona, area-wide fires occurred with an average return interval of 22 years (Dieterich 1983). As the interval between fires increases more fuel accumulates and the likelihood of a stand-destroying crown fire increases. The younger stands described by Lang and Stewart had probably established following a severe fire. Moir and Ludwig (1979) declare that most mixed conifer stands are established in this manner. These severe fires may either directly produced a stand of conifers or a stand which first goes through an aspen stage (Pearson 1931, Pearson and Marsh 1935).

Spruce–Fir

Because many spruce–fir forests in the Southwest had been little affected by logging, grazing, or fire suppression until very recently, the historic conditions and disturbance regimes of this community can be reconstructed with good precision and reliability.

The major disturbances in the spruce–fir forests of the Southwest are fire and bark beetles (Baker and Veblen 1990, Schmid and Frye 1977). Although these agents are capable of reshaping whole landscapes by conflagration or outbreak, these events are relatively infrequent (100+ years between major disturbances, see Veblen et al. 1994). In these wet forests, ignitions are rare, but heavy fuel accumulations and steep slopes result in high-intensity, crown fires lethal to the subalpine vegetation (Grissino-Mayer et al. 1995). The different species of bark beetles are selective for either spruce (usually developing in blowdown) or subalpine fir (often associated with root disease). Snags created by these events can remain for decades (Schmid and Hinds 1974). The species and patterns of regeneration are highly variable and depend in the long term on climate (Anderson 1993) and in the short term by site conditions and immediate disturbance history (Rebertus et al. 1992, Patten and Stromberg 1995). On wetter sites protected from intense radiation, Engelmann spruce and subalpine fir usually take and hold early dominance of the site (Fowells 1965). On drier sites, aspen, Douglas-fir, and southwestern white pine may become established initially and spruce and fir later emerge as dominants (Lebarron and Jemison 1953).

Aspen

Although aspen is usually successional to conifers, the aspen community has always been an important component of Southwestern forests. Precise estimates of forest area occupied by aspen before the mid-19th century are not available, but since that time acreages have declined with cessation in burning (Jones and DeByle 1985). The successional dynamics and ecological role of aspen stands are well known and little changed since the historic period.

Aspen stands are usually very quite wet and do not readily burn; aspen stems, however, have only a thin bark and are easily killed by a light fire (Baker 1925). After a fire, aspen re-sprout or sucker from shallow lateral roots (Gruell and Loope 1974). Although aspen is a climax species on some sites, it is usually seral to conifers (Mueggler 1976). This replacement is gradual and can take from 100 to 200 or more years (Bartos et al 1983). If an aspen stand is within a mixed conifer forest, conifers can become established within a single decade (Jones 1974). Because aspen stands are so different from conifer stands, they are very important for landscape diversity and wildlife habitat. Although aspen stems are short lived and snags do not stand long, the wood is soft, often decayed, and therefore useful to cavity-dependent species. Young sprouts are heavily browsed by elk and deer.

Riparian Wetlands

The Colorado and Rio Grande River systems in Arizona and New Mexico extend from headwater tributaries in high, forested mountains to the lowlands of the subtropical Sonoran and Chihuahuan Deserts (Minckley and Rinne 1985). Continuous corridors of riparian vegetation once covered hundreds of miles along desert rivers in the Southwest. Besides forested riparian communities, there were riparian shrublands, marshlands, and grasslands. These plant communities were found at elevations from high wet meadows and cienegas, to tree-banked streams, to slack water sloughs and marshes–the alpine, montane, and floodplains-plains riparian ecosystems (Dick-Peddie 1993).

Riparian ecosystems served as permanent habitat and seasonal migration routes for many species of birds and mammals. Rivers and spring-fed cienegas supported specialized, endemic fish species. Beaver (Castor canadensis) required water and created ponds to retain it during periods of low flow. Open water for drinking was essential for some animals such as doves and bats (e.g., the spotted bat, Euderma maculatum). Other species like the southwestern willow flycatcher (Empidonax trailii extimus) and the ferruginous pygmy owl (Glaucidium brasillianum) depended on the special plant and animal communities of a riparian wetland.

Alpine plant associations in the Southwest are dominated by graminoids–especially, sedges or Carex, other aquatic plants in the genera Cyperus, Juncus, and Scirpus, and grasses of the genera Deschampsia, Agrostis, and Glyceria. By trapping sediments moving over the channel bottom these native aquatic plants are important for protecting the long-term stability of alpine meadows or cienegas.

A large proportion of the Southwest's "live" streams pass through the upland montane forests of the mixed conifer and ponderosa pine communities. The riparian zones themselves are usually narrow, often following relatively steep stream channels in restricted valleys. The watercourses are flanked by hardwood and coniferous riparian plant communities. Narrowleaf cottonwood, box-elder (Acer negundo), bigtooth maple (Acer grandidentatum), Scouler willow (Salix scouleriana), and arroyo willows (Salix lasiolepis) are typical hardwoods. White fir and blue spruce are the common conifers in and adjacent to riparian ecosystems (Szaro 1989, Minckley and Brown 1994). These streams usually flood from snowmelt in the spring; and many riparian species depend on over-bank flooding for seed transport and burial in fresh, fertile alluvial sediments. Historically, seed shedding and flooding usually coincided.

The vegetation and stream channels between the montane and subtropical floodplains are markedly diverse. In steep terrain, watercourses are often confined to narrow canyon areas where riparian vegetation is restricted. Broader riparian zones occur in wider valley bottoms. This warm, temperate, mid-elevational zone supports mixed broadleaf forests of Arizona sycamore (Platanus wrightii), Arizona walnut (Juglans major), velvet ash (Fraxinus velutina), and Bonpland willow (Salix bonplandiana) (Szaro 1989, Minckley and Brown 1994).

Historically, the low desert, subtropical zone riparian areas supported gallery forests of Fremont cottonwood and Gooding willow (Salix gooddingii). Understory communities include bosque-forming mesquite (Prosopis) on terraces and coyote willow (Salix exigua) in wetter areas. Historical records state that major rivers in the hot desert zone flowed through riparian gallery forests, densely vegetated flood plains, and substantial marshlands (Minckley and Rinne 1985).

WILDLIFE

As forest vegetation and landscapes evolved over millennia, so did their accompanying wildlife. Little is known about the history of many wildlife species. Grass and shrub communities of the intermountain portions of the West evolved largely in the absence of large hoofed herbivores (Mack 1989). Elk (Cervus elphus), mule deer (Odocoileus hemionus), pronghorn (Antilocarpa americana), bighorn sheep (Ovis canadensis), and occasionally bison (Bison bison) were the dominant grazing and browsing animals. The more open canopies in woodland, ponderosa pine, and mixed conifer forests favored wildlife species such as deer, turkey, some songbirds, and small rodents. Some species such as the northern goshawk (Accipiter gentilis atricapillus) preferred forests with more closed canopies; others such as the Mexican spotted owl (Sirix occidentalis lucida) preferred habitats with vertical structure, as provided in steep canyons or tall, diverse forests.

Numerous authors (Kay 1990, 1992, 1994a, 1994b; Koch 1941; Rawley 1985) report that deer, elk, antelope, and bighorn sheep were rare or absent in the Rocky Mountains during the early 1800s. Davis (1982) reviews the status of wildlife and hunting in Arizona from 1824 to 1865. Mule deer and pronghorn were the game animals most often encountered, but hunting success was no better than today. Bighorn sheep remained in rugged mountainous districts that were avoided by early travelers so few encounters were reported. Davis (1982) goes on to state that antelope had been reduced to small scattered bands by World War I and that native elk were gone by 1900. He describes hunting in early Arizona:

"Most of us harbor the notion that game in a virgin wilderness is always abundant, tame, and easy to kill–a hunter's paradise. No such assessment could possibly be made from the writings of the American explorers of Arizona. More often than not they complained that game was scarce and wary, and on a number of occasions several were forced to kill and eat their horses and mules."