The climate cycles of the 2 million years of the Quaternary were a major force in the evolution of plant response to change. Quaternary climate has been primarily glacial with interglacials such as the current Holocene a minor component. Plant species responded individually to climate changes and, consequently, species composition has continually changed. The legacy of Quaternary climate change is that plant communities are far less stable than they appear to be from our perspective. They are unique at each location, difficult to define, and communities that are relics from a previous environment can be sensitive to small or transient environmental changes. Plant communities are variable both in space and time. Many ecological principles and concepts, and ecosystem paradigms derived from them, require revision to incorporate this variation. The concepts of habitat type and condition and trend, for example, do not reflect dynamic vegetation response to changes in climate. Our knowledge is presently insufficient to adequately describe interactions between ecosystems changing climate, but the patterns of vegetation response to environmental changes of the past may provide important information on vegetation response to present and future climate change. The concepts of thresholds, multiple steady states, and multiple successional pathways are helpful in understanding the dynamic interrelationships between vegetation and environmental changes.

Quantifying root responses of naturally growing range plants to treatments is difficult. The objective of this study was to develop a containerized technique to study individual plants growing in the field under near natural environmental conditions. Three containers were evaluated: 15 X 100 cm nylon (2,000 pores/cm2) sacks, 15 X 100 cm polyvinyl chloride (PVC) tubes, and 30 X 100 cm PVC tubes. The 15 X 100 cm PVC tubes were easiest to handle, and plants grown in these containers appeared similar in size and growth form to adjacent, undisturbed plants. Survival rate for 165 sand bluestem (Andropogon hallii Hack.) plants grown for 2 years in the 15 X 100 cm PVC tubes was 98.8%. This survival rate was achieved despite clipping treatments during the second year.

This study was conducted in a sagebrush-bitterbrush vegetation type in southcentral Wyoming to compare the effects of mowing and 2,4-D application on shrub cover and bitterbrush (Purshia tridentata Pursh) use, preference values, browse values, and residual twig weight after summer/fall browsing by cattle and winter big game use. Areas 3 to 6 ha in size were mowed, sprayed with 2,4-D, or left untreated on 4 different sites. Total shrub cover was reduced from 38% (53% mountain big sagebrush (Artemisia tridentata spp. vaseyana), 37% bitterbrush) to 23% (38% sagebrush, 52% bitterbrush) by mowing, and to 17% (18% sagebrush, 64% bitterbrush) by herbicide application. Twig weight use was greater on mowed areas, on the more productive sites, and during summer cattle browsing when leaves were present than during winter browsing by big game. Frequency of use was inversely related to bitterbrush availability, although the relationship was confounded by different snow depths on different sites. Utilization values based on basal diameter were more sensitive to site and treatment effects than length and weight utilization values, but were influenced by differences in twig morphology and did not reflect differences in twig use, preference values, browse values, or browsing residual weights. Mowing produced the greatest preference values and browsing residual weights for cattle browsing and the greatest browse values for both cattle and big game. Preference values, browse values, and browsing residual weight appear to be useful indicators of site, treatment, and browsing effects.

Over the past 2 centuries, the island of Kaho'olawe has suffered the ravages of war, slash-and-burn agriculture, and overgrazing. Today, much of the island is barren and severely eroded. A research project initiated in 1988 has sought to identify effective, economical techniques to revegetate portions of the island. Treatments included drill seeding plus several rates of fertilization with monoammonium phosphate (11-52-0). Some treatments also include jute netting for soil moisture conservation and erosion control. The effect of windbreak fencing was evaluated across all treatments. Drill seeding plus broadcast application of at least 62 kg ha-1 N plus 291 kg h-1 P205 was the most cost-effective treatment. Jute netting and windbreak fencing significantly enhanced plant production, but the high cost of materials and maintenance limits their use to critical areas. The planted species with greatest promise for the windy, semiarid conditions on Kaho'olawe were buffelgrass (Cenchrus ciliaris L.), bermudagrass [Cynodon dactylon (L.) Pers.] and weeping lovegrass [Eragrostis curvula (Schrad.) Nees]. Although not included in the seed mixture, Australian saltbush (Atriplex semibaccata R. Br.), a naturalized species, responded favorably to fertilization. A subsequent, larger scale revegetation project using a specially modified chisel plow seeder to scarify, plant, and apply in-furrow fertilization in a single-pass operation reduced the cost and improved the results of the revegetation process.

Revegetation of disturbed rangelands in western Canada is severely impeded by cattle grazing. Fencing to protect emergent vegetation is costly and restricts animal movement. Chemically repelling cattle from emergent vegetation may provide a convenient and economical alternative to fencing. This study determined whether certain repellents could reduce grazing intensity on vegetation to which they were applied compared to untreated vegetation. Canopy measurements were used to compare grazing intensity. Three trials were conducted on reclaimed land within the Aspen Parkland region of central Alberta. Time since reclamation was over 10 years in Trial A, 2 years in Trial B, and 3 weeks in Trial C. Pregnant mares' urine, Hinder® (150 mg ml-1 ammonium soaps), Skoot® (120 mg mi-1 tetramethylthiuram disulfide), and Deer-Away Big Game Repellent® (37% putrescent egg solids) were evaluated. Two concentrations of each repellent were sprayed onto 1 X 3-m treatment plots randomized within blocks replicated 4 to 6 times. Plot canopies were measured either by gently resting a sheet of plastic laminate over the canopy, or by lowering a sliding bar attached perpendicularly to a meter stick until it contacted the uppermost leaves of the canopy, and recording the height of the sheet or bar above the soil surface. Canopies of plots treated with Big Game Repellent® were taller than control plot canopies on each measurement date in all trials, indicating grazing was reduced. Big Game Repellent® plots were also generally taller than plots treated with other repellents. Canopies of plots treated with pregnant mares' urine, Hinder®, and Skoot® generally did not differ from control plots, nor did they differ from each other in any trial. Low repellent concentrations did not reduce grazing in any trial, but high concentrations reduced grazing in all trials. Repellent effectiveness was not permanent since all canopy measurements became shorter with time. Big Game Repellent® was effective in reducing grazing intensity by cattle, but practical use of repellents for grazing management requires further investigation.

We evaluated seasonal changes in browse quality 5-6 years after experimental manipulations to control unwanted woody vegetation using combinations of herbicide and fire on cross timbers rangeland in central Oklahoma. The study area consisted of two 32-ha replications of untreated controls and 4 brush treatments (tebuthiuron and triclopyr used singly or in combination with periodic prescribed burning); herbicides were applied in 1983 and fires initiated in 1985. Nutritional quality of blackberry (Rubus spp.), coralberry, (Symphoricarpos orbiculatus Moench), rough-leaf dogwood (Cornus drummondii Meyer), elm (Ulmus spp.), greenbrier (Smilax spp.), hackberry (Celtis spp.), and smooth sumac (Rhus glabra L.) were assessed by measuring crude protein, in vitro dry matter digestibility, neutral detergent fiber, acid detergent fiber, and moisture content. Crude protein concentrations of browse were 14% higher on herbicide-treated areas compared to untreated controls and 11% higher on triclopyr treatments compared to tebuthiuron treatments. In vitro dry matter digestibility was 9% higher on herbicide-treated areas compared to untreated controls. Fiber constituents and moisture content were not influenced by brush treatments. Prescribed burning combined with herbicide applications did not improve the quality of browse. Our results indicate that browse quality can be improved for white-tailed deer by applications of tebuthiuron or triclopyr and improvements persist for up to 6 years post treatment.

Ponderosa pine (Pinus ponderosa Lawson) needles cause abortions in pregnant cows. We examined pine needle consumption by cattle in 2 trials in eastern Montana. Trial 1 compared pregnant and open cows (n=4) from January to March 1989; trial 2 compared pregnant cattle (n=4) that received either 9 kg alfalfa hay head-1 day-1 or 1.4 kg alfalfa pellets head-1 day-1 from December 1989 to February 1990. Diets were estimated using both bite counts and fecal analysis. During trial 1, bite counts revealed pregnant and open cows consumed 45 and 42% of their grazing diets as pine needles (P>0.1). Fecal analysis showed that pregnant cows consumed more pine needles than did open cows (36% vs. 27%, respectively) (P<0.05). During trial 2, cattle consumed < 1% of their diets as pine needles. In trial 1 cattle consumed less pine litter and consumed more needles from trees as snow depth increased. Consumption of needles from trees increased as ambient temperature declined; no needles were consumed from trees when the minimum daily temperature exceeded -5” C. During both trials, grazing times decreased as temperatures declined, and increased as snow depth and wind speed decreased. We conclude that weather is a major factor influencing needle consumption; other interrelated factors may be forage availability, snow cover, and grazing time. Plne needle consumption, and the risk of abortion, in pregnant cattle appears to be greatly diminished during mild winter weather.

Forage quality and quantity from palatable grasses, like creeping bluestem [Schizachyrium scoparium (Michx.) Nash var. polycladus (Schriber & Ball) Bruner (Syn S. stoloniferum Nash.)], are limited, especially in winter when cows graze Florida range. We anticipated that N fertilizer (0, 40, 60, 120 kg ha-1), P (0, 25 kg ha-1) and K (0, 100 kg ha-1) would increase bluestem yield, tiller density, and forage quality. Within sample dates yield and tiller density increased linearly with N rate. For example 31 days after fertilization, intercepts for equations predicting yield were 319 kg ha-1 and 124 m-2 with coefficients of 1.2 and 0.29, respectively, where the independent variable is N rate. Over sample dates yield responses to N rate were quadratic and tiller densities were cubic. Reproductive tiller density was increased by N fertilization (1989 tiller density, nom-2, = 30 + 0.29N). Neither yield nor tiller density was affected by P fertilizer, but K fertilizer increased reproductive tiller density, hence fall yield. After 3 years of fertilization, N had negative quadratic and negative linear effects on yield and tiller density, respectively. Tissue N concentration in the fall was reduced with N fertilization because of increases in reproductive growth (1988 calendar 145 days postfertilization, g kg-1 = 5.7 - 0.041 N + 0.00031 N2). Fertilization of creeping bluestem is not a recommended practice when bluestem is to be grazed in fall and winter.

In Panicoid grasses, elevation of the crown node above the soil surface caused by excessive subcoleoptile internode elongation is detrimental to seedling establishment. We describe a technique to screen grass seedlings for excessive crown node elevation. Seed of 11 perennial grass cultivars were germinated and grown in a plywood box 1.2 X 1.2 X 1.2 m at an irradiance of 1.5 micromole m-2 sec-1 at 30 degrees C. A subset of 8 grasses were grown similarly but at an irradiance of 0.75 micromole m-2 sec-1. Elevation of crown nodes above the growth media and shoot length (from crown node to leaf tip) were measured 7 days after planting. The crown nodes of 4 Eragrostis species were not elevated above the soil surface, while other species had mean crown node elevations of 1 cm or more. Relative ranking of crown node elevation was similar for a species within each light level, but response to change in light intensity was not consistent across species. Sideoats grama (Bouteloua curtipendula (Michx.) Torr.), blue grama (Bouteloua gracilis (H.B.K. Lag. ex Steud.)), and kleingrass (Panicum coloratum L.) exhibited significant crown node elevation in this system. Estimates of genetic variation based on comparison of apomictic and sexual grasses within this group suggest that substantial genetic variation exists for excessive subcoleoptile internode elongation and that progress by selection for lower crown node placement should be possible in most of the grasses studied. This system allows characterization of grasses for extent of crown node elevation and is being used in a recurrent selection protocol to select for lower crown node placement.