Search

Scaled quail (Callipepla squamata) numbers were evaluated during and after a 2-year drought period using strip census techniques on 2 pastures in late seral rangeland ecological condition and 2 pastures in mid-seral rangeland ecological condtion. This study was conducted on the Chihuahuan Desert Rangeland Research Center (CDRRC) in south-central New Mexico on 4 adjoining pastures that were similar in size and terrain. During part of the study (August 1994 to April 1997) all 4 pastures were destocked due to depletion of perennial grass cover and biomass from a combination of drought and heavy cattle grazing. Scaled quail sightings pooled across sampling periods (9) were different (P = 0.08) on high and low rangeland ecological condition treatments. They averaged 10.72 birds per pasture on late-seral and 4.22 birds per pasture on mid-seral rangeland ecological condition treatments. Autumn perennial grass cover and standing biomass levels was higher (P 0.10) between treatments. Our study indicates that during extended dry periods livestock grazing at moderate intensities may adversely affect scaled quail populations in the Chihuahuan Desert by depleting perennial grass cover. However, in years of above average precipitation there is evidence scaled quail prefer mid-seral pastures over late-seral pastures. Maintaining a mosaic of conservatively (late-seral) and moderately (mid-seral) grazed pastures should best meet the habitat needs of scaled quail in the Chihuahuan Desert. DOI:10.2458/azu_jrm_v56i4_joseph

Raccoons (Procyon lotor) are important furbearers and ecological generalists that exist in a variety of landscapes. Numerous studies have examined raccoon space and macrohabitat use, but information detailing these parameters within pine-dominated landscapes is lacking. Furthermore, no studies using radiotracking have examined raccoon habitat use at multiple spatial scales. We radiomonitored 131 adult raccoons (99 M, 32 F) during 1991-1997 in central Mississippi, USA. We subsequently assessed space use and habitat selection at 3 spatial scales. Males maintained larger home ranges and core areas than females. Size of home ranges differed by season, but raccoons maintained similar-size core areas across seasons. Habitat use differed by gender only at the coarsest spatial scale. Raccoons consistently used mature (>30 years old) pine and hardwood habitats over other available habitats at all spatial scales. Although hardwood-dominated habitats were important to raccoons, our findings suggest that in pine-dominated landscapes, raccoons select multiple seral stages of pine. Raccoons also readily used ≤8-year-old pine habitats, likely because of increased foraging opportunities within these habitats, particularly, during spring and summer. Our findings illustrate the importance of juxtaposition of habitat types for raccoons in pine-dominated landscapes, specifically the availability of mature pine and hardwood habitats, as well as early-successional habitats capable of providing beneficial foraging resources. Additionally, our findings offer evidence that prescribed burning may alter landscape use by raccoons, providing the potential that manipulating burning frequencies may allow managers to manage raccoon habitat-use patterns.

Rangeland ecological condition was monitored over a 48 year period on 41 sites on Bureau of Land Management rangelands scattered across 6 counties in southwestern New Mexico. All sites were grazed by livestock during the study period. Sampling occurred in 1952, 1962, 1982, 1992, 1997, 1998, and 1999. A modified Parker 3 step method in conjunction with Dyksterhuis quantitative climax procedures were used to determine rangeland ecological condition. At the end of the 48 year study period (1952-1999), the average rangeland ecological condition score across study sites was the same (P > 0.05) as the beginning of the study (39% versus 41% remaining climax vegetation, respectively). Major changes (P > 0.05) in rangeland condition occurred within the study period due to annual fluctuations in precipitation. Ecological condition scores increased in the 1980s and early 1990s due to above average precipitation. However, drought in the early to mid 1950's and again in the mid to late 1990's caused rangeland condition scores to decline. At the end of the study (1997-1999), 38% of the sites were in late seral ecological condition, compared to an average of 25% in the 1952 to 1982 period. The amount of rangeland in late seral ecological condition increased while the amount of rangeland in mid seral and early seral condition decreased in the 1990s compared to the 1952-1962 period. The average percent cover of black grama (Bouteloua eriopoda Torr.) and tobosa (Hilaria mutica Buckley), the primary forage grasses in the Chihuahuan Desert, were the same (P > 0.05) in 1952 and 1999. Over the 48 year study period, the average cover of shrubs including honey mesquite (Prosopis glandulosa Torr.) showed no change (P > 0.05). However major increases in honey mesquite basal cover occurred on 1 site and creosote-bush (Larria tridentata [Pursh] Nutt.) increased on another. Grazing intensity was evaluated during the last 3 years of study (1997, 1998, 1999). Overall grazing use of forage across sites and years averaged 34% or conservative. Our research shows controlled livestock grazing is sustainable on Chihuahuan Desert rangelands receiving from 26-35 cm annual precipitation. DOI:10.2458/azu_jrm_v55i4_navarro

A 20-year set of cover data on sagebrush semi-desert plant communities responding to wildfire and livestock grazing near Mills in central Utah provided an opportunity to compare the assumptions and adaptability of classical and state-and-transition models for describing secondary succession. Cover data were organized and analyzed by plant species, growth forms, and other ground cover classes. Graphical analysis, ordination (employing semi-strong hybrid multi-dimensional scaling), regression, and analysis-of-variance were used to determine whether the patterns observed were best described as community change (tightly linked species) or individualistic change (each species acting independently). Distinct differences in total plant cover, growth form, and species composition were found between burned (both grazed and ungrazed) and the unburned and grazed treatments. Conventional graphical and statistical analyses of burned and ungrazed plots showed greater and earlier expansion of perennial grasses and then relatively less cover-weighted compositional change in recent years compared to the other treatments. Vegetation on none of the treatments appears to have stabilized toward either the pre-burn sagebrush semi-desert, a new state or the potential natural community for the site involved. Pathways of change reflected in the ordinations have been complex in all treatments. The only obvious trends in responses of individual species were to fire and the inverse relationship of cheatgrass to total perennial vegetational cover. All this evidence points to few tight linkages between species or growth form groups and thus favors viewing these patterns individualistically. While the state-and-transition model allows greater flexibility than the classical model in the depiction of plant community/individual species changes consequent to any management action, it doesn't apply readily everywhere, as exemplified by this case study. DOI:10.2458/azu_jrm_v55i2_west

We evaluated the visual obstruction method as a non-destructive means of estimating herbage standing crop in tallgrass prairie. Prediction models were developed for both plot-level and pasture-level estimates by regressing standing crop from clipped plots on visual obstruction measurements (VOM) from 48, 20-sample trials. Trials were conducted year-round on burned and non-burned sites in different seral stages and with various levels of productivity and grazing pressure. Separate models were required for burned and non-burned pastures, but both applied across all other variables and were unaffected by community heterogeneity. Coefficients of determination were 0.95 and 0.90 for burned and non-burned pastures, respectively. Use of a more precise measurement scale for visual obstruction did not improve the prediction models. Models for standing crop based on individual quadrats explained less variation than models based on transect averages. The highest correlations with visual obstruction were obtained with 20 x 50 cm quadrats placed adjacent to the measurement pole and oriented toward the observer. The visual obstruction method required little training and mean deviations of student readings from those of the trainer were less than 1 cm. Sampling efficiency is improved with the visual obstruction method because it is reasonably accurate and 6 times faster than clipping. Standing crop estimates can be calculated immediately and less field equipment is needed. DOI:10.2458/azu_jrm_v54i1_vermeire

Twenty-six, 0.04 ha macroplots were sampled on 9 range sites in southwestern Montana to compare successional scores and condition classifications of range condition analysis and United States Forest Service (USFS) Ecodata and Ecopac (Strata) analysis methods. Range condition scores (0-100%) and range condition classes (poor, fair, good, excellent) were derived from the traditional Soil Conservation Service range condition analysis method, with the exception that only major decreaser and increaser graminoids and shrubs were individually clipped and bagged. Ecological status scores (1-100%) and ecological condition classes (low, mid, high, very high) were determined with United States Forest Service Ecodata methods. Range condition score means were greater (p < 0.02) than ecological status score means (48% vs 41%). Standing crop biomass affected differences (p < 0.001) between range condition scores and ecological status scores. Lower producing sites had greater range condition scores than ecological status scores and higher producing sites had greater ecological status scores than range condition scores. Range condition classes and ecological condition classes were not independent (p < 0.02). Differences between the 2 methods were attributable to the use of species composition by weight for the range condition analysis and the use of percent canopy cover by Ecodata methods. Rangeland managers trying to determine successional status should realize that range condition analysis and Ecodata methods produce similar condition classes but different condition scores. DOI:10.2458/azu_jrm_v53i2_winslow

We studied the distribution of American marten (Martes americana) and short- tailed weasels (Mustela erminea anguinae) on northern Vancouver Island, British Columbia, over 2 winters from 1996 to 1998. We tested a combination of track plates and remote cameras to detect marten and weasels in 4 broadly different habitats: early seral cut-blocks, regenerating cut-blocks, old forests, and grassy edge habitats. We measured vegetation structure at many sample sites in order to describe habitats where marten and weasels were detected. We detected similar numbers of marten in all of the low elevation forest stands sampled; fewer marten were detected in grassy edge habitats than in forested stands. Weasels were detected at sites with open forest and little structure compared to forested stands. Weasels were detected primarily in edge habitats associated with forest openings or riparian areas dominated by grass ground cover. We used rotten bait and lure, instead of fresh bait, to make sites most attractive to marten and weasels. We used 7-day checks which allowed us to trap more sites than shorter check in- tervals. Track plate boxes worked well for detecting marten and weasels though we suggest extending the roof and sides at both ends of the box 20 cm beyond the floor to provide greater protection from rain. Cameras were expensive, heavy, and failed to detect marten more often than track plates. We suggest that cameras be used only to identify unknown tracks or early in a study to determine appropriate methods.

The Chihuahuan Desert of the southwestern United States and northern Mexico is the primary habitat of the scaled quail (Callipepla squamata), but populations have declined over much of this area since 1960. Because changes in ecological condition of Chihuahuan Desert rangelands may explain the decline in scaled quail populations over the last 37 years, we conducted scaled quail counts seasonally (1988-89) on sites in late- and mid-seral ecological condition in southcentral New Mexico. On the basis of percent cover, grasses were the most important plant group on the late-seral site (65-70% remaining climax vegetation), while shrubs dominated the mid-seral site. The main effect of ecological condition (35-40% remaining climax vegetation; P = 0.48) and the interaction of ecological condition with season (P = 0.69) were not significant for scaled quail observations, but the main effect of season was significant (P < 0.05). Scaled quail observations in grassland, shrub-grass, and grassland communities differed from expected values. When data were pooled across seasons, more scaled quail were observed in shrub-grass mosaic communities than shrubland and grassland communities (P = 0.07). Fewer sightings of scaled quail were made in grassland communities near climax than in shrubland communities in an early seral stage. Availability of food sources and protective cover were apparently primary determinants of habitat use by scaled quail. Data from our study indicate that areas with a mixture of late-and mid-seral plant communities will provide the best available habitat for scaled quail in the Chihuahuan Desert.

Wildfire in the growing season is relatively frequent and interest is increasing in using growing-season fire in management of tallgrass prairie. However, the influence of fire in the growing season on forage production and species composition, especially in mid-successional tallgrass prairie, is largely unknown. Our objective was to compare vegetation composition and production on Loamy Prairie and Very Shallow ecological sites in mid-successional stages in response to late growing-season fire at different frequencies. We applied 4 burning treatments (no burn, or 1, 2, or 3 burns in 5 years) in the late growing season in southern Oklahoma during a series of years of above-average precipitation. The sites were dominated at the beginning of the study by early- and mid-successional species including prairie threeawn (Aristida oligantha (Michx)), a species indicating a disturbance history. After the initial burns in 1990, tallgrasses, little bluestem, and perennial grasses were reduced by burning on the Loamy site. Forbs were more productive on burned plots (1,980 kg ha-1) than on plots that were not burned (1,290 kg ha-1) averaged across sites. Total production was not reduced by burning in 1990. Growing-season burns in 2 consecutive years had little influence on species composition or production as compared to a single burn in 2 years. Warm-season perennial grasses other than tallgrasses and little bluestem increased on the Loamy site, but decreased on the Shallow site. Production of cool-season perennial grasses increased to almost 40% of total production on twice-burned plots averaged across sites. Other than the effect on cool-season perennial grasses, 2 burns over a two-year period had little effect beyond the first growing season after the second burn. Twice-burned plots and plots burned 3 times produced more forbs than either plots that were burned once or not burned. Production of perennial grasses was opposite that of forb production. Total production was not reduced on either site regardless of fire frequency. Results indicate managers may expect a short-term reduction in production of forage grasses and an increase in forbs following late growing-season fire in mid-successional tallgrass prairies.

A field experiment compared the effects of herbicide treatments (picloram, clopyralid and clopyralid + 2,4-D, each at two timings, at the recommended rates for control of the exotic forb Centaurea maculosa) on the structure and species diversity of native plant communities during an 8-year period in western Montana, USA. Floristic composition of replicated treatment plots was sampled before spraying and for 3 years after the initial herbicide applications at two grassland and two early seral forest sites. Following the third year post-spray measurements, half the treatment plots were randomly selected to be resprayed and community sampling was continued for two more years. Diversity was quantified as species richness and Shannon Diversity Index. Standing crops by growth form were estimated by double sampling at the conclusion of the experiment. Herbicide treatments had high efficacy on the target weed, shifting the plant communities back to a grass-dominated structure. Depressions in plant community diversity were small and transitory. In the third year after the initial applications, there were no significant differences among treatments and some herbicide-treated plots had begun to surpass the untreated plots in community diversity measures. With most treatments, respraying 3-4 years after the initial applications did not reduce plant diversity compared to untreated levels. Late season applications, made after most herbicide-susceptible forbs had entered summer drought-induced dormancy, minimized impacts on plant community diversity. The behaviour of herbicide residues in the soil is described in relation to community-level effects. Implications of the results for exotic weed management in conservation settings are discussed.

In this study the impact of a rotation grazing system on the soil and vegetation of a Stipa-Bouteloua-Agropyron community in the mixed prairie ecoregion was compared with the ungrazed treatment in exclosures. At a low stocking rate, grazing had no effect on the vegetation but did alter soil quality. Grazing pressure was so light in the rotational grazing treatment that recovery of productivity, as measured by standing crop and litter, was not significantly different from the ungrazed treatment. Conversely, the species distribution was unchanged but was indicative of a lower seral state for this mixed prairie. The effect of grazing on this community was indirect, possibly by altering the microenvironment. The relationships observed among forage production, soil chemistry, and species composition raise questions on the importance of any one variable expressing range condition on the mixed prairie.

Seasonal wildlife observations were made along transects on 2 pastures in late seral and 2 pastures in mid seral condition in southcentral New Mexico in non-drought and drought years (1993, 1994). Remaining climax vegetation was about 64% and 57% on late seral pastures. About 37% and 32% of the climax vegetation remained on mid seral pastures. Total wildlife and total bird sightings/km2 during the study period were higher (P 0.10) between late and mid seral condition rangelands. Black-tailed jackrabbit (Lepus californicus J.A. Allen) numbers were higher (P

We studied cattle and deer diet quality within replicated grazing treatments of continuous and short-duration grazing at heavy and moderate stocking rates. The study was conducted at the Welder Wildlife Refuge, Sinton, Tex. from October 1987 to July 1989. We obtained cattle diet samples from esophageally fistulated steers. Deer diets were reconstructed using data obtained through the bite-count technique. Digestibility (IVDOM) and crude protein (CP) of cattle diets were similar between grazing systems and stocking rates. Digestibility of deer diets was affected by both grazing systems and stocking rate. Dietary CP and IVDOM of deer and cattle diets both differed among seasons. Dietary CP levels met maintenance requirements for deer throughout the study. Also, CP levels were high enough to meet low- to mid-gestation requirements. Deer dietary protein requirements for growth and lactation were never met regardless of grazing strategy. Although protein content of cattle diets was relatively low, these values satisfied cattle maintenance needs. Nursing cows, however, would not have met their requirement in any season sampled regardless of grazing system or stocking rates. Continuous grazing and moderate stocking rates may provide white-tailed deer the opportunity for selecting diets containing more desirable forbs and greater nutrient concentration. Less intensive rotational grazing at moderate rates may be preferred to maintain a relatively high seral stage.

Management of northern spotted owls (Strix occidentalis caurina) in northwestern California is aided by knowledge of its prey base, the dusky-footed woodrat (Neotoma fuscipes). From late April through September 1990, we radiotracked 25 dusky-footed woodrats representing both sexes and a cross section of age classes (ad, subad, and juv). Animals were captured in sapling/poletimber shrubfields resulting from past clearcut timber harvest. In northwestern California woodrats reach their highest densities in this seral stage. During evening radiotracking sessions, we detected intermittent, short distance movements of juvenile and subadult males and females, and adult females from shrubfields into adjacent old-growth forests occupied by spotted owls. None of the radiotagged woodrats permanently emigrated to older-aged forests during our study period. Between-year emigration from shrubfield to old forest, however, was documented for marked individuals. Our telemetry results indicate that woodrats are not averse to crossing sharp ecotonal boundaries from shrubfields into adjacent old forest, into canopy openings in these forests, or into natural openings within dense shrubfields. A substantial number of our radiotagged woodrats were killed by predators, with carcasses most often (5 of 9 depredated woodrats) found in adjacent old forest. These observations suggest that during these between-habitat forays, woodrats may be more vulnerable to avian and mammal predators. Therefore, to some degree, the existence of brushy shrubfields adjacent to older forest may increase the availability of woodrats to predators, such as spotted owls, that exploit prey from a variety of habitats but spend the majority of their time hunting in late seral stage forests.

AssTRkcT Knowledge of lynx (Lynx canadensis) use of burned areas is desirable to anticipate the effects of fire management in the taiga. During 1991 to 1994 we tested whether L. canadensis and snowshoe hares (Lepus americanus) selectively used post-fire seres during winter in the Alaskan taiga and whether selection could be explained by population trends and the concept of refugia for L. americanus. Forest seral stages included early-successional tall shrub-sapling (1985 burn), mid-successional dense tree (1966 burn), and mature coniferous (100 to 115 yr old). Track counts, pelt sealing records, and anecdotal observations indicated that numbers of L. canadensis and L. americanus declined during 1991 to 1994. Tracks of both species were most abundant in the 1966 burn, which was characterized by dense coniferous and deciduous regeneration. Lepus americanus and L. canadensis aIso had the highest preference for the 1966 burn, which is consistent with the prediction that refugia for L. americanus would also contain most of the L. canadensis during the decline phase of the hare cycle. Fire is the predominant disturbance in upland taiga, and mid-successional forest is clearly important to L. canadensis and L. americanus. However, stands of mature forest may be important to L. canadensis for denning and as a source of alternative prey during declines of L. americanus. Thus, conservative fire management to benefit L. canadensis in interior Alaska should ensure some stands of mature forest to fulfill their potential habitat needs.

The traditional milpa agriculture system (slash-burn) of the Lacandon Maya people in eastern Chiapas, Mexico has created and uses a variety of habitat patches including the whole range of seral stages during forest development. This study examines seed bank attributes in habitats associated with slash-burn agriculture and with the more recent expansion of grasslands. We estimated the abundance and floristic composition of seeds in the soil from induced grasslands, milpa fields neighboring early and mature forests, and from secondary forest and mature tropical rain forest (15 samples by site, 2-3 replicated sites per successional stage). A total of 130 species germinated from the soil samples. The seeds of late successional tree species were scarce in all the seed soil samples (

During surveys of upland forests in the Tsuga heterophylla Zone in Washington and Oregon, 1986 to 1996, we caught 192 S. gracilis. In the Oregon Coast Range, more (85%) S. gracilis were caught in oldeilar to Coast Range old growth (P = 0.60) with no differences between seral stages (P = 0.74). No S. gracilis were caught in upland forests > 500 m in Washington or at low elevations in the Puget Trough. Spilogale gracilis was widely distributed in upland forests, but may be tied to riparian areas at elevational extremes. Old growth provides good habitat in some dry areas (Coast Range), perhaps through better developed forest floors (coarse woody debris, litter, humus) than are found in young stands. Absence from the Puget Trough study areas may reflect dryness, disturbance history, and forest fragmentation by agriculture and urban development. Future research on S. gracilis may lead to a better understanding of forest-floor processes and differences in ecosytem processes that occur elevationally within relatively small geographic areas.

Botanical composition of diets and dietary overlap were investigated among free-ranging cattle (Bos taurus), Rocky Mountain elk (Cervus elaphus nelsoni), and white-tailed deer (Odocoileus virginianus) in coniferous forests of northern Idaho. The study was conducted within the Abies grandis/Clintonia uniflora (grand fir/queencup beadlily) and Thuja plicata/Clintonia uniflora (western redcedar/queencup beadlily) habitat types. Botanical composition of ungulate diets was determined via microhistological analysis of fresh fecal samples collected in early summer, mid-summer, and early fall of 1987 and 1988. Dietary overlap was examined using Kulcyznski's similarity index. This formula also was used to compare the botanical composition of ungulate diets vs. plant community composition at 5 different seral stages: herb-shrub, sapling, pole, mature, and potential natural community. Cattle consumed graminoid-dominated diets from within early successional communities. Elk also foraged predominantly on graminoids, but elk foraging habits were more diverse and more variable amongst seasons than cattle. White-tailed deer diets were dominated by forbs and shrubs from within late successional communities. Competition for forage between cattle and elk was more likely in the grand fir habitat type, while forage competition between elk and white-tailed deer was more likely in the redcedar habitat type. There was little evidence of potential forage competition between cattle and white-tailed deer in either habitat type.

The germinable seed bank in a grassland affects the succession of degraded range and the recolonization of disturbed sites, and must be understood to predict potential responses to management. The germinable seed bank on the fescue prairie was characterized and its relationship to grazing, season, and depth of burial determined. The study was conducted in the fescue prairie of southwestern Alberta in livestock exclosures and on paddocks that, since 1949, have been stocked at fixed rates to achieve light, moderate, or heavy grazing pressures. Surface debris was sampled in fall and spring, and soil was sampled to a depth of 6 cm in spring. The samples were spread on vermiculite in trays and the seeds allowed to germinate over a 90-day period. In fall, total surface seed numbers m(-2) increased from 1,785 to 7,783 from the ungrazed to heavily grazed site, and most of the differences were accounted for by whitlow-grass (Draba spp.) and Kentucky bluegrass (Poa pratensis L.). These species also contributed most to differences between fall and spring on the grazed sites. Total seed numbers were similar (1,790 vs 1,803) in spring and fall on ungrazed sites. The species composition of the seed bank did not change with depth. In the soil, the annual forb pygmyflower (Androsace septentrionalis L.) was the most common seed but was not detected in a vegetation survey. Soil disturbance in the fescue prairie is more likely to lead to a seral community dominated by annual forbs, than a rough fescue (Festuca campestris Rydb.) dominated grassland.

Although the use of herbicides and prescribed fire have been shown to increase density of cottontail rabbit (Sylvilagus floridanus) populations, the impact of such brush management practices on their condition has not been explored. We used discriminant analysis to investigate responses of overall physical condition of cottontail rabbits (n = 422 adults) to brush management and succession on replicated disturbed and undisturbed upland hardwood forest-tallgrass prairie over a 6-year period. Five different disturbed habitat types were experimentally created using herbicides (tebuthiuron or triclopyr), fire, or a combination of both. Parameters that were important discriminators of rabbit physical condition among habitat types and post-disturbance successional changes included indices of kidney fat and parasitism, and relative masses of spleen, liver, and dried stomach digesta. Brush management practices using herbicides influenced overall condition of rabbits, but the type of habitat disturbance was not important. Effects on overall body condition of cottontail rabbits from burning disturbed habitats were not apparent until later seral stages when production of herbaceous dicots declined and vegetative composition more closely resembled that of undisturbed areas.

A heath community found on Pulau Sibu, an island lying off the east coast of Peninsular Malaysia, was found to consist of a low scrub, generally under 2 m in height, containing 24 species of vascular plants growing on hills overlooking the sea. It grew on a very shallow and acidic (pH 4.3) soil that was low in total nitrogen concentration (0.2%). Leaves of the seven most common shrub species, Ixonanthes reticulata, Melastoma malabathricum, Rhodamnia cinerea, Rhodomyrtus tomentosa, Schima wallichii, Styphelia malayana and Tristaniopsis obovata, were collected for morphological, anatomical, and chemical analysis. The community was characterized by very small leaves (a dominance of microphylls), with very low nitrogen, phosphorus and total chlorophyll concentrations, and chlorophyll a/b ratio. Lamina thickness, specific leaf area, crude fiber content and potassium concentrations were less extreme. A number of the species in the community are generally considered to be early successional, but no strong evidence could be found for any recent disturbance event that would have initiated the production of a seral community.

According to theory, maximum populations of a species occur within the stage of ecological succession to which the species is adapted. The maximizing seral stage may differ among areas, which illustrates a need for understanding factors that regulate the relationship. Thus, we tested the hypothesis that optimum seral stage for northern bobwhites (Colinus virginianus) on southern rangeland varies inversely with site productivity. Data were collected on 6 sites (2 each in early, mid-, and late seral status) in each of 3 regions of different productivity (potential annual herbaceous biomass production) in the Rio Grande Plains and Gulf Coast Prairies of Texas during 1989-90. Results supported a predicted interaction; on more productive sites, bobwhite density was highest with early seral status, whereas on less productive sites, density was unaffected by seral status

Because dusky-footed woodrats (Neotoma fuscipes) are important prey species of northern spotted owls (Strix occidentalis caurina), we estimated their abundance in Douglas-fir (Pseudotsuga menziesii)/tanoak (Lithocarpus densiflora) forests of different ages in northwestern California. We surveyed and woodrats in 6-8 stands in each of 5 seral stages during the summers of 1988-90. We used belt transects to compute nest density, and livetrapped at nest sites to estimate woodrat numbers per nest. Combining these 2 sources of information provided an estimate of woodrat densities, which were highest in sapling/brushy poletimber stands, followed by seedling/shrub and large old-growth stands. Woodrats were not found in small sawtimber stands and rarely occurred in large sawtimber stands

Metrosideros polymorpha Gaud. (MYRTACEAE), 'ohi'a, dominates pioneer dry forest communities in Hawai'i, while other tree taxa dominate later seral stages. Tree density is greater in the later seral communities, and the dominant species there exhibit significantly lower mid-day plant water potentials than Metrosideros from nearby pioneer communities. This pattern suggests that the greater water stress experienced by the later seral trees may be due to competition for limited water, and this may restrict Metrosideros from these sites. The replacement of Metrosideros in later seral stages of Hawaiian dry forests contrasts with its maintenance as the dominant canopy component in both early and late seral rain forest communities.

SUMMARY (1) The growth patterns of tree height and stem diameter in crowded Abies veitchii, A. mariesii and Betula ermanii populations were investigated based on the diffusion model for growth dynamics in size-structured populations. (2) Size-dependent growth was found in each species, but the pattern was different between the species. Small, suppressed individuals of Abies spp. showed little growth in tree height but grew in stem diameter. In small, suppressed individuals of B. ermanii, the converse was found. (3) The diffusion model revealed that stem diameter growth of Abies spp. and tree height growth of B. ermanii are affected only a little by competition between individuals even under crowded conditions, and thus that growth in these dimensions is determined in a genetically strict manner for each species. These growth patterns correspond to the life-history traits and the successional status of Abies spp. (shade-tolerant late-successional species) and B. ermanii (a shade-intolerant early-successional species). (4) This study indicates the quantitative differences of assimilate partitioning between tree height growth and stem diameter growth in size-structured plant populations in relation to the characteristics and successional status of species.

(1) A community of saxicolous lichens and the seral communities of primary succession on subalpine moraines were compared with respect to (i) the treatment of 'no-contact' samples, (ii) the hypothesis of proportionality (that the probability of a species being a contact neighbour is proportional to its probability of being an initial sample), and (iii) the procedure for examining all pairwise associations. The effect of spatial autocorrelation in the lichen data was evaluated, and the analysis was modified to accommodate the possibility of self-contact in the seral communities. (2) The data show that in the lichen community, treating uncolonized rock as if it were a species can be justified, and that while there is significant spatial autocorrelation, it does not have a large effect on the investigation of species pairs. (3) In the moraine communities, it was found that the number of species with significantly many no-contact samples declined from the youngest surface and that the interspecific associations among the major plant categories were remarkably consistent through the series of moraines. (4) In both kinds of vegetation, the proportionality hypothesis was rejected. The analysis of data from these very different communities shows how the treatment of contact data can be modified to accommodate and evaluate the differences.

Range condition is usually defined by similarity of current to climax or potential vegetation. It is often assumed that rangelands in low condition are biologically less productive than those in higher condition. The objective of this study was to determine if range condition (ecological status) is related to total productivity or to forage production for livestock. Adjacent areas along fencelines representing differences in range condition were sampled in 58 locations. These comparisons represented 31 different range sites across southern Arizona. Weight by species of above-ground peak standing crop current year's growth of vegetation was estimated using the dry-weight-rank/comparative yield methods. Range condition was rated with Soil Conservation Service range site descriptions. Species were classified as forage or non-forage to estimate forage available for cattle. In 75-85% of comparisons of good condition sites to fair condition, good to poor, and fair to poor, total current year's standing crop did not differ significantly. Where differences were significant, productivity was not consistently more on the high condition class. Forage production, however, was more from the stand in the higher condition class in about 2/3 of the comparisons. We concluded that in southern Arizona rangelands in higher condition (higher seral) classes usually produce more forage for cattle than lower condition classes on the same range site. Nevertheless, it is not usually true that total biomass productivity on low condition range is less than the same range site in higher condition.

FACELLI, J. M. AND W. P. CARSON (Department of Biological Sciences, Rutgers University, P.O. Box 1059, Piscataway, NJ 08855). Heterogeneity of plant litter accumulation in successional communities. Bull. Torrey Bot. Club 118: 62-66. 1991.-We sampled the litter accumulated in five seral sites (aged 5, 15, 23, 29, and 47 yr) and in an oldgrowth forest in early spring and fall. We collected and weighed all the litter in twelve 0.24 m2 quadrats per field, arranged in three parallel transects. In the four younger sites, the transects were 12, 24, and 36 m distant from a forest edge. Total litter was lower in the four younger fields than in either the older site or in the forest. Litter composition in the four younger fields was rather similar, and mostly composed of herb litter. In the 47 yr site and in the forest, litter of tree leaves and wood were the main components. Spatial heterogeneity in the litter layer was high, specially in the 23 and 29 yr oldfields. In the four younger fields, the composition of the litter was significantly affected both by the age of the field and by the distance from the forest edge.

(1) A randomized block design experiment on succession at Aston Rowant NNR, Oxfordshire, has been recorded for 14 years from 1969 to 1982. The treatments were annual grazing, and no grazing, alone, or after disturbance by burning or rotavating in 1969. (2) Changes in cover of thirty-two common flowering plants in the field layer over 14 years are tabulated. In relation to succession in the ungrazed plots, species were grouped into four categories, those which (i) gained in cover (especially tall coarse grasses), (ii) increased initially, then declined (more characteristic of disturbed treatments), (iii) decreased (mainly smaller and shorter-lived species), and (iv) showed no consistent gains or losses. (3) In the disturbed plots, especially with rotavation, many seed-bank species appeared, several subsequently persisting for many years. (4) Corylus avellana and Tamus communes colonized a few plots after several years, indicating the importance of dispersal rather than the stage of succession reached. Seed dispersal from plants in the different treatments suggested that long-term succession experiments should include barrier areas. (5) The original flora has not been entirely eliminated during 14 years of secondary succession and could probably be reclaimed. This highlights the value of information about expected seral changes as well as descriptive accounts of the present vegetation for managers of nature reserves.

Multivariate analyses were used to describe graphically the post-fire patterns of change in the structure of forest and heathland small-mammal communities over successional time in Australia. Data were collected from 16 forest sites representing a temporal sequence after fire, and from one heath (shrubland) site which has been followed continuously over a 15 yr period, with regeneration from two successive fire cycles. These small-mammal communities follow a successional trajectory that may be described using factor analysis to represent seven species as factors in two or three dimensions. The trajectory illustrates the shift in community structure from one dominated by species of Pseudomys (native mice), through a period of domination by dasyurid marsupials into the longest period where species of Rattus (native rats) dominate, until they too are reduced in numbers. The strongest factor reflects the differences observed between communities from unburnt or mature seral stages and those in the early stages of regeneration after fire. This form of multivariate analysis is senstitive to the scale of the temporal data analysed and successional patterns are likely to emerge only when data from annual or greater temporal scales are used. The trajectories tend to follow a cycle as they move back towards their unburnt or undisturbed mature state. The order in which species enter the succession appears linked to the functional group to which the species belongs, which is consistent with assembly rules for these functional groups proposed previously for these habitats.

Two changes in traditional range condition analyses are recommended: (1) to replace the terms excellent, good, fair, and poor with ecological equivalents of climax, late seral, mid-seral, and early seral in cases where this is practical; and (2) to develop relationships between products (e.g., livestock, wood products, water) or conditions (e.g., infiltration, site stability, erosion) and successional stage or state. Such information will allow the land manager to evaluate possible tradeoffs between managing for a particular successional stage or state and particular goods or

The natural landslide disturbance regime and the major trends in post-landslide succession were studied within a 44-km 2 belt between 530 and 850 m a.s.l. (subtropical lower montane wet forest, sensu Holdridge) in the Luquillo Mountains of Puerto Rico. Examination of aerial photographs from 1936 to 1988, and field surveys, revealed forty-six landslides. Forty per cent of the landslides had map areas (the horizontal projection of their actual areas) between 200 and 400 m 2 . Landslides>1800 m 2 comprised less than 10% of the sample, but accounted for about 40% of the total map area disturbed (...)

Summary. For the simplest geometry of a treefall gap, a circle, the area open to colonization is Ai = x (rg - di), where rg is the radius of the gap and di is the distance into the gap from the gap edge that is normally required for seedlings of species i to avoid suppression by overhanging limbs. The proportion of a gap available to a given species increases with gap size according to a convex function for species tolerant of partial shade near gap edges (low di), but approaches linearity for less shade-tolerant pioneers (high di). Border effects of other gap shapes are easily modeled. Tropical ecologists recognize that pioneer tree species regenerate in larger rather than smaller gaps (Hartshorn 1978), and that seedlings and saplings of pioneer species both require and can take advantage of enhanced light under a broken canopy better than those of primary forest species (Bazzaz and Pickett 1980, Bazzaz 1984). Generally unrecognized are the simple geometrical relationships between habitat available to pioneers and gap size. Geometry allows rough predictions of the relative abundance of pioneer versus primary forest occupants of maturing gaps. This note explores the implications of gap geometry for gap colonization by tree species; empirical tests will be reported as they become available.

Cluster analysis followed by stepwise discriminant analysis was used to delineate ecological stages on a mixed grass prairie in western South Dakota. Forty-seven variables were analyzed for 48 sites ranging from potential vegetation to early seral stages. A cover-frequency index for western wheatgrass (Agropyron smithii), blue grama (Bouteloua gracilis), and buffalograss (Buchloe dactyloides) was the most valuable in identifying 4 different (P

We measured seasonal habitat use by 30 radio-collared Sitka black-tailed deer (Odocoileus hemionus sitkensis) on northern Admiralty Island in southeastern Alaska from February 1979 to July 1982. Habitat use varied seasonally as deer moved from low-elevation ( 600 m) summer ranges in open canopy subalpine and alpine habitats. Deer used oldgrowth forest almost exclusively during winter and spring, and high-volume old growth (>74 mbf/ha) was used in much greater proportion than its abundance. To minimize the impacts of timber harvesting on deer populations, emphasis should be placed on maintaining stands of high-volume old growth on low-elevation deer winter ranges. J. WILDL. MANAGE. 54(3):371-378 The Sitka black-tailed deer in southeastern Alaska occurs at the northern limit of its natural range (Wallmo 1981). This subspecies inhabits a densely forested coastal environment where deep winter snows are the major factor limiting population size (Klein and Olson 1960). In high snowfall regions of the Pacific Northwest and in southeastern Alaska, low-elevation old-growth forest has been identified as essential deer winter range (Bloom 1978, Wallmo and Schoen 1980, Bunnell and Jones 1984). The 6.8 million-ha Tongass National Forest encompasses over 80% of the land area of southeastern Alaska. Approximately 7,000 ha of old growth on National Forest lands, as well as additional thousands on state and private lands, are scheduled for harvest annually (U.S. For. Serv. [USFS] 1988). Converting old-growth forests to younger seral forests in Alaska will reduce forage availability and consequently deer carrying capacity (Wallmo and Schoen 1980, Alaback 1982). In southeastern Alaska, the old-growth forest is variable in structure and composition on a fine-grained (

(...) Relationships of leaf area to sapwood area, stem respiration to sapwood volume and vigour to growth efficiency provide equations for tree growth, leaf-area dynamics and canopy structure. The model simulates mixed-age, mixed-species populations of trees on a 1000-m 2 patch. The test site, Fiby urskog (central Sweden), has been unmanaged since a catastrophic storm in 1795. (...)

We investigated the range and extent to which different microdisturbances could have an impact on woody seedling establishment in seral oldfields and mature forest on the New Jersey Piedmont, USA.

Forests dominated by Picea sitchensis (Bong) and tree growth in gaps to determine if P. sitchensis can Carr. and Tsuga heterophylla (Raf.) Sarg. occupy a narrow persist under a small gap disturbance regime. Tree replace- coastal strip along the Pacific Northwest coast of North ment models derived from tree replacement patterns in America. P. sitchensis is considered seral to T. heterophylla gaps were used to predict abundances of P. sitchensis and but descriptions of groups of P. sitchensis saplings and small T. heterophylla in the next generation. Tree replacement trees in light gaps created by small canopy disturbances models did not predict a decline in P. sitchensis abundance (

Vegetation composition and structure were quantified in 21 pine stands converted from second-growth hardwood-pine forest at two state forests in Virginia. These stands included three replications each of seven developmental stages ranging in age from 1-22 years. Species richness and vegetation cover in the ground stratum showed the same trend: high values in stands 1-5 years of age followed by a decline from 5-15 years, at which point canopy closure was complete, and these variables were relatively stable for the next 7 years. A peak in transgressive stratum richness in the 9th year was followed by a similar decline and leveling off. Evenness (J') values over time were nearly constant for these understory strata, except in 3-year-old stands where a decline in ground stratum evenness occurred due to dominance by Andropogon virginicus. Increasing richness and evenness in the overstory stratum was due to growth of hardwoods. Trends in richness and evenness differed from the old-field successional model due to a high diversity of seeds and sprouts at the start of succession. Despite such differences, the seral stages typical of old-field succession were discernible in converted stand succession. However, they were somewhat abbreviated and masked by the abundance and variety of reproductive units and the rapid closure of the canopy by codeveloping pines and sprout hardwoods. INTRODUCTION The central Piedmont of Virginia is part of the Oak-Pine Forest Region delineated by Braun (1950). Young forests are characteristically dominated by Pinus virginiana and P. echinata and are gradually replaced during succession by an oak-hickory complex (Braun, 1950). From the 1700s to the early 1800s Piedmont lands were cleared, intensively farmed and abandoned (Gottmann, 1969). Abusive crop farming depleted soils and led to severe and extensive topsoil erosion (Culbertson, 1948; Gottmann, 1969). Much abandoned cropland reverted to native forest through succession. The end of the 19th century marked the beginning of a new period of deforestation when most of the commercially valuable forest of the Piedmont was harvested (Gottmann, 1969). Pinus taeda (loblolly pine) has been the principal species planted in the central Piedmont since the 1930s, even though it is incapable of natural regeneration throughout much of this region (S. F. Warner, pers. comm.). Up to the mid-1950s, most planting of P. taeda was in old fields. Owing to a decline in agricultural land abandonment, hardwood and pine-hardwood forests have been directly converted to P. taeda plantations at an accelerated rate. Whereas much is known about the vegetation dynamics of old-field succession in the Piedmont region of the Southeast (Crafton and Wells, 1934; Billings, 1938; McQuilkin, 1940; Oosting, 1942; Keever, 1950; Nicholson and Monk, 1974, 1975), little is known about such changes associated with the establishment and development of pine plantations on sites converted from second-growth forests. Thus, the objective of this study was to examine temporal changes in the composition and structure of vegetation in loblolly pine stands converted from such forests. 1 Present address: 435 Jackson Street, Fall River, Massachusetts 02721. 2 Present address: Department of Biology. 3Department of Biology.

Changes in pioneer community structure in an 0.8-ha central New York field were monitored during a year of cultivation and the 1st 2 years of abandonment. Species composition was unlike that of old fields in more southerly location, but similar to southeastern Michigan old fields. High diversity (Shannon-Wiener Index) and high proportion of perennials in the 1st year of abandonment and the establishment of a perennial grass stage during the 2nd year were attributed to persistence of propagules in the soil after 1 year of cultivation or possibly to rapid restocking due to the small size of the field. Maximum productivity occurred during June of the 2nd year of abandonment when dominance was established by Phleum pratense; however, total production over the growing season was lower than in the 1st year when dominance was shared by several species early in the growing season. Eighty-four percent of the species identified during the 1st year of abandonment were exotics, primarily European in origin. This percentage was much higher than reported for 1st-year fields in southeastern and mid-Atlantic regions. INTRODUCTION Secondary succession has been intensively studied in the eastern deciduous forest region of the United States. Much of the work has been done in the southeastern and mid-Atlantic Coastal Plain (McCormick and Buell, 1957; Odum, 1960; Woodwell and Oosting, 1965; Golley, 1965; Golley and Gentry, 1966; Levin, 1966), Piedmont (Oosting, 1942; Keever, 1950; Bard, 1952; Daniel and Platt, 1968; Hanks, 1971; Buell et at., 1971), and in the Midwest (Evans and Cain, 1957; Beckwith, 1954; Wiegert and Evans, 1964; Bazzaz, 1968). Very little has been published about secondary succession in the northeastern Appalachian Plateau. Swan (1970) studied the response of various seral stages to burning in S-central New York. Hurd et al. (1971) studied two abandoned hayfields in central New York in an attempt to ascertain the effect of nutrient enrichment on stability and diversity on the old-field ecosystem. This paper reports on the first phase of a study designed to compare succession on a central New York old field with old-field succession in southeastern, mid-Atlantic, and Midwestern regions. MATERIALS AND METHODS The study is being carried out at the Suny College at Oneonta Biological Field Station near Cooperstown. The elevation is 512 m (1680 ft) above sea level. The soil type is Volusia channery silt loam, an imperfectly drained, acid soil of glacial origin. Soils of this type typically have a fragipan at a depth of 30 cm (1 ft). These soils are used principally as pasture land, or for the production of oats or

Algal samples from a hot spring in Yellowstone National Park, Wyoming, were taken at 10 temperature intervals, 10 m apart, from 57.2 C downstream to 45.7 C, and then the entire visible community was removed. Sampling was continued at the same intervals during the 1st, 3rd, 5th 7th-week successional stages. No successional stage had more than 17 species but four species were noticeably the dominants in the descending order of: Synechococcus lividus, Mastigocladus laminosus, Phormidium laminosum and P. angustissimum. These dominants comprised 94.56%o of the frequency and 70.69% of the volume in the climax community, increasing to 97.37% of the frequency and 91.77% of the volume in the 1st week seral stage. The four dominants alone were sampled at the highest temperature interval in the climax community but only two were present in the lst-week seral stage at the same temperature interval. Other species entered the community as the water cooled in all seral stages. The minor species reappeared in the lst-week seral stage at a mean temperature 2.23 C lower than in the climax community. The uniform conditions in the hot spring environment, and the uniformity of the community (blue green algae) created a situation paradoxical with the competitive exclusion principle by which hierarchal order and diversity were maintained during the seral stages. Succession progressed without the community changing the environment, and without being affected by changing external factors. The combined actions of individual populations, each in exponential growth, may have been responsible for the community overshooting the carrying capacity of the environment for diversity before returning to the carrying capacity of the climax community. INTRODUCTION Community succession is one of the oldest of ecological concepts and has numerous interpretations resulting from varied field studies and, more recently, ecological models. An interpretation by Clements (1905, 1916) considered the community as a single entity that develops, as would an organism, to maturity. This organismic interpretation of succession has been embraced by numerous ecologists for its apparen.t logic, ease of comprehension and the status of its advocate. Clements believed that succession was an orderly process in which a community progressed to a more complex community and that all successional stages in a region would culminate in one monoclimax. From Clements' aspect "community and succession concepts were lineal descendants of a long tradition of natural philosophy which held that there were design, purpose and unity in nature and that these required a holistic approach to their understanding" (McIntosh, 1980). Gleason (1926), as perhaps the most prominent early ecologist to reject Clements' interpretation of succession, considered succession to be an ecological process of the mass effect of individual plants acting within the community the core of the individualistic concept of succession. The long span of time involved in successional studies has been a disadvantage in studying the process on one site (McIntosh, 1980). Henry and Swan (1974) believed that one generation of white pine can dominate a forest for 200-250 years before other species begin replacing them. Gleason (1927) wrote that, "Succession in general is very slow, and the vegetational history of any particular habitat is more apt to be measured in thousands of years, or even in tens of thousands, than in centuries or decades." Because of this slow rate of terrestrial succession and because meteorological records at the time (1927) extended less than a century, Gleason further stated that a studied community might actually be in retrogression. Hot spring communities offer an excellent 1 This study was supported in part by the Grants and Research Funding Committee, Southeast Missouri State University.

took about eight years. By then the composition of the community was similar to that recorded forty years earlier. The successional changes in number of species, individuals and biomass are illustrated for the total fauna as well as for dominating groups. The sequential changes of some numerically dominant populations showed a bell-shaped curve pattern. During the first years after pollution abatement, when polychaetes dominated, these population changes were drastic but evened out in later seral stages. The role of larval recruitment in succession is discussed. Three diversity indices were used to assess the community structure: Shannon's formula, its measurement of evenness, and Sanders' rarefaction technique. As tools for assessing pollution or recovery, the two former had to be used with care, as the highest values were recorded at the beginning of the recovery process when the individuals found were few but evenly distributed among the few species present. The rarefaction technique and the measure of species richness were more satisfactory for this kind of assessment.

We tested the effects of predation, above-ground diffuse competition and belowground diffuse competition on the survival and growth of Carya tomentosa (Juglandaceae) seedlings in a five year oldfield bordering a mature forest. We planted eighty seedlings under cages, in pots or with surrounding vegetation tied back (in all combinations) to realize the treatments. We found that seedlings unprotected from herbivores suffered great mortality (i.e., survival < 10%). Seedlings planted away from the forest edge suffered significantly less predation than those closer to the forest. Seedlings often resprouted after browsing. Because only a few seedlings survived outside the cages, we focused on the caged seedlings to examine competition effects on seedling growth and survival. No significant differences were found in leaf, stem, root, and total biomass of surviving seedlings under any of the competition treatments. Insect herbivores affected 15% of the seedlings by removing 10% of their leaf area during the first year. In the second year, 30% had 1-10% damage to leaf tissue. No seedlings died due to invertebrate herbivores. Browsing by mammalian herbivores appears to be the primary factor determining woody seedling establishment patterns in seral oldfields.

A sagebrush-grass range in northern Utah, a critical winter area for deer and elk, was grazed by domestic horses in order to evaluate their potential to reduce selectively herbaceous vegetation which competes with bitterbrush (Purshia tridentata), an important big game winter forage. The diets of horses were recorded during spring and summer grazing under two stocking levels on small pastures. Horses consumed largely grass species throughout the study period in this high seral community. Forbs were important forage only under heavy stocking. No use of bitterbrush was recorded. Rate of forage disappearance, expressed on a peranimal-weight basis, was found to be lower on heavily stocked pastures. All pastures grazed by horses responded with increased seasonal twig production of bitterbrush over the non-grazed state. Seasonal twig production was greatest in response to heavy stocking during early July. Herds of migratory mule deer (Odocoileus hemionus hemionus) and elk (Cervus elaphus nelsoni) in the Rocky Mountains and Great Basin often rely on foothill rangelands dominated by big sagebrush (Artemisia tridentata) and other more palatable shrubs for winter browse. Winter deer mortality in Utah appears to be inversely related to the amount of palatable browse species available (Robinette et al 1952). Bitterbrush supplies important browse on many big game winter ranges in this area. Much of present bitterbrush range in our western states is considered to be seral communities created and maintained by perturbations of some sort (Ellison 1960). Extensive livestock grazing on palatable grasses and forbs in the late 1800’s and early 1900’s led, in part, to widespread successional changes favoring bitterbrush, which was frequently subordinate in the pristine vegetation (Hull and Hull 1974). In response to an increase in palatable shrubs and intensive management, elk and deer prospered by mid-century. In recent years, increasing urbanization near many foothill winter range areas has promoted a trend towards protection of big game winter range from livestock use because of water quality and soil erosion concerns. On many protected rangelands, succession is now favoring grasses since they receive little use during the growing season to balance heavy winter use of shrubs by big game (Smith 1949, Urness 1976). As palatable shrubs disappear from the community, productivity of these lands as deer winter habitat decreases. Declines in the Intermountain Region of deer herd productivity in the mid-1970’s (Connolly 1981) may have been partially due to these plant community changes. It is well recognized that controlled livestock grazing can be used in a manipulative sense to direct plant community succession (Anderson and Scherzinger 1975, Lewis 1969). Research with catAt the time of the research authors were graduate research assistant and associate professor, Range Science Dept., UMC-52, Logan, Utah 84322. Reinerisnow research assistant Range and Wildlife Manage. Dep., Texas Tech Univ. Lubbock, 79409. The authors wish to acknowledge Charles H. Jensen for his advice on the study design and layout. Statistical analysis was assisted by Dr. Ronald V. Sisson and Mr. David L. Turner. Thanks are also due to John Moody and Deborah Bowdenfortheir help in collecting the field data, and to the Starwood Ranch, Aspen, Cola., for generously loaning the study animals. The study was supported in part by the Utah Division of Wildlife Resources, Project W-105-R. Manuscript received August 20, I98 I. JOURNAL OF RANGE MANAGEMENT 35(5). September 1982 tle and sheep has shown that the productivity of bitterbrush can be increased by springtime grazing of competing herbaceous vegetation (Smith and Doe11 1968, Jensen et al. 1972, Smith et al. 1979). One problem encountered when using sheep or cattle for this purpose is that both animals consume bitterbrush. Smith and Doe11 (1968) found that on moderately stocked pastures bitterbrush utilization by cattle increased dramatically in July, in some cases exceeding utilization of grasses. Jensen et al. (I 972), in developing guidelines for sheep grazing on big game winter range, also reported heavy use of bitterbrush beginning in July. Both investigations suggested that stocking and season ofhvestock use must be precisely controlled to avoid excessive utilization of bitterbrush. The use of horse grazing as a “biological tool” to enhance bitterbrush on big game winter range was investigated in this study because horses were presumed to select for grass species (Stoddart et al. 1975, p. 257, Wagner 1978). The objectives of this study were twofold: (1) to investigate the use of spring-summer horse grazing as a means of increasing the seasonal production of bitterbrush, and (2) to determine the effect of grazing season and stocking intensity on the diets of horses grazing sagebrush-grass rangeland.

During secondary succession in tropical America, reproductive characteristics of the constituent plants change as species constitution changes in the direction of 'climax' or equilibrium. Self-compatibility is more prevalent in early stages, while outcrossing is more the rule in later stages as dicliny, dioecism, and self-incompatibility beome more frequent. Observations indicate that plants of increasingly large stature and seral position are more widely spaced, and tend to have larger brightly colored flowers and larger pollinators. Associated understory plants come to have small unicolorous flowers and small pollinators as light availability decreases. Plants of early stages have primarily small, many-seeded dehiscent fruits, while plants with large, few-seeded fleshy fruits come to dominate equilibrium forests, particularly in wet forest. Canopy emergent trees, epiphytes, and lianas of mature forest are exceptional in employing wind-dispersal. Many of these changes correlate well with the thesis that pioneer plants are more highly r-selected, while plants of more mature forest tend to be K-selected. ALTHOUGH THE STUDY OF SECONDARY SUCCESSION in temperate latitudes signalled the onset of ecological endeavor (Cowles 1899, Clements 1916), only in recent years has the study of succession in the tropics made significant advances. The study of reproductive characteristics of plants in seral sequences has lagged even further behind. Our studies and those of others (Bawa 1974, Vasquez-Yanes 1974a, McKey 1975, Ruiz Zapata and Kalin Arroyo 1978) allow some preliminary conclusions and hypotheses on the nature of reproductive characteristics of plants in different seral stages. In this paper we consider the various portions of reproductive biology as a sequence of topics, i.e. floral biology, seed biology, and breeding systems. We then contrast findings from lowland wet and dry forest habitats, and finally integrate our findings and hypotheses with

We determined associations of breeding birds with seral stages following clearcutting in a northern hardwood forest. We surveyed birds and measured habitat variables on 90 study plots. Principle components (PC) and cluster analysis were used to group 24 species of birds into 4 seral associations. The 4 avian assemblages were associated with young clearcuts; pole and mature stands with dense understories; mature deciduous stands; and mature, mixed deciduous-coniferous stands. Habitat attributes responsible for the classification were related to canopy height and volume, understory density, tree diameter, number of snags, and basal area of deciduous and coniferous trees. Managers can encourage the presence of all species assemblages by maintaining a mixture of forest age classes. J. WILDL. MANAGE. 52(4):771-777 Habitat use by birds is influenced by vegetation structure (Noon et al. 1980, Rice et al. 1984, Cody 1985). Silvicultural practices can modify structural characteristics of forests and affect distributions of breeding birds (Webb et al. 1977, Titterington et al. 1979, Crawford et al. 1981). Detailed habitat models for individual bird species (Straw et al. 1986) are desirable for featured-species management, but information on habitat associations of groups of birds may be more useful for general resource planning and environmental monitoring (Landres 1983, Verner 1983). Several authors have identified groups of birds based on similarities in habitat use. Neimi and Pfanmuller (1979) identified groups (species associations) that were distributed in a similar manner across a range of communities. Titterington (1977) and Crawford et al. (1981) also identified groups of birds based on habitat affinities; Titterington (1977) referred to these groups as seral associations. Similar assemblages have also been called management guilds, defined by Verner (1983) as a "group of species that respond in similar ways to changes in their environment." Verner's (1983) definition is more narrow than Titterington's (1977) however, because guilds represent only a subset of bird species found in a given habitat (Verner 1984). Published information on habitat use by breeding birds in managed hardwood forests of New England is limited. Titterington et al. (1979) described species assemblages found in early seral stages of spruce-fir forests in Maine, and more recently Chadwick et al. (1986) described the effects of fuelwood cutting on bird communities in southern New England. DeGraaf and Rudis (1986) assigned bird species in New England to forest cover types and age classes, but these classifications were based mostly on general habitat descriptions found in the literature. Few of these habitat associations have been verified by field data. We describe avian assemblages, based on use of habitats created by silvicultural practices. Our study plots represented an array of seral stages common in hardwood forests of New England. We efine species assemblages through quantitative descripitions of habitat characteristics and refer to these groups of birds as seral associations. Seral associations may also be viewed as management indicators because most of the study plots were affected by recent management activities. B. Roberts assisted with fieldwork. Our study was supported by the Windham Foundation, Grafton, Vermont and Cooperative Forest Research Funds administered through the School of Natural Resources at the University of Ver