Physiological responses of plant populations to herbivory and their consequences for ecosystem nutrient flow.

We explored how responses of two populations variable in grazing tolerance provide feedbacks to nutrient supply by controlling carbon supply to soil heterotrophs. The study focused on differences in production and carbon and nitrogen allocation patterns between the two populations. The grazing-tolerant population, or on-colony population, is found on intensively grazed prairie dog colonies, and a grazing-intolerant population, the off-colony population, is found in uncolonized grasslands. Equations describing the production and allocation responses to defoliation for the two ecotypes described were incorporated into CENTURY, a nutrientcycling simulation model. Simulations showed an increase in plant production that paralleled increases in net nitrogen mineralization. Production was greater with grazing and was maintained at higher grazing intensities for the on-colony than the off-colony population. Differences between the populations provided important controls over nitrogen losses. Feedbacks between plant responses to grazing and nitrogen cycling accounted for increased nitrogen availability with grazing. These feedbacks were more important determinants of ecosystem function than were fertilization effects of urine and feces deposition. The simulation results suggest that ecosystem function may be sensitive to physiological differences in population responses to periodic disturbances like herbivory.

Production and nitrogen responses of the African dwarf shrub Indigofera spinosa to defoliation and water limitation.

The dwarf shrub Indigofera spinosa Forsk. (Papilionacea), a native forage species of arid Northwest Kenya, was propogated from seed, grown in a controlled environment, and subjected to three treatments of defoliation and watering frequencies in a factorial experimental design. Biomass production and nitrogen accumulation in tissue components were measured to determine defoliation responses in a water-limited environment. We hypothesized that plants would maintain biomass and nitrogen flows despite removal of aboveground meristems and tissues by defoliation. Principal experimental results included a slight reduction (11%; P=0.08) of total biomass production by clipping ca. 1/3 or 2/3 of new leaves and stems and all apical meristems every month. Total aboveground production was not affected by clipping, while final root biomass was reduced 17% by the 2/3 clipping. The least water stressed plants were affected most negatively by defoliation, and the unclipped plants responded more negatively to greater water limitation. Plants achieved partial biomass compensation through alterations in shoot activity and continued allocation of photosynthate to roots. A smaller fraction of leaf production was directed to litter in clipped plants although clipping only removed the youngest tissues, suggesting that clipping increased leaf longevity. In turn, each leaf probably contributed a greater total quantity of photosynthate. Photosynthetic rates were also likely to have been increased by clipping water-stressed plants. In contrast to biomass, plants overcompensated for nitrogen lost to defoliation. Total nitrogen uptake by individual plants was stimulated by defoliation, as there was more total nitrogen in leaves and stems. Increased nitrogen uptake was achieved by clipping stimulation of total uptake per unit of root rather than of total root mass.

Patterns in grass silicification: response to grazing history and defoliation.

Morphologically distinct populations of a North American perennial grass, Agropyron smithii, collected from a heavily grazed prairie dog (Cynomys ludovicianus) colony (PDC) and a grazing exclosure (EX), were grown in an environmental chamber to determine whether: (1) leaf silicon (Si) concentrations are greater in plant populations which differentiated under heavy grazing pressure, and (2) leaf silicification is inducible by defoliation. Mean shoot Si concentration of nondefoliated plants was greater in the PDC population (2.2%) than the EX population (1.9%) over the 18 wk experiment, largely as a result of differences in Si concentrations in leaf blades. However, leaf Si concentration was lower in defoliated plants of each population than in nondefoliated plants, indicating that leaf silicification was not an inducible herbivore defense mechanism in A. smithii. The higher leaf Si concentrations from the heavily grazed population may be associated with grazingrelated environmental stresses such as a warmer, drier microclimate or with morphological characteristics related to grazing tolerance or avoidance.

Herbivory tolerance of Agropyron smithii populations with different grazing histories.

The effects of defoliation on growth and nitrogen (N) nutrition were examined in populations of Agropyron smithii (western wheatgrass) collected from a heavily grazed black-tailed prairie dog (Cynomys ludovicianus) colony (ON-colony) and a nearby lightly grazed, uncolonized area (OFF-colony). Defoliated and nondefoliated plants were grown at low soil N availability with similar sized defoliated individuals of A. smithii from a grazing-exclosure population as a common competitor. Sequential harvests were made over 24 days following defoliation. Growth analysis plus biomass and N yield and distribution data were used to identify features which may contribute to plant defoliation tolerance. Defoliation reduced total production 34% across populations. Defoliated plants produced as much new blade tissue, but only 67% as much new root biomass as did nondefoliated controls. Plants from prairie dog colonies accumulated biomass at a faster relative rate than did plants from uncolonized sites, in part, because of a 250% greater mean relative growth rate of blades and more than 200% greater rate of biomass production per unit blade biomass. Total N accumulation was significantly greater in defoliated ON- than OFF-colony individuals. The mean relative accumulation rate of N was increased by defoliation in ON-colony plants, but reduced by defoliation in OFF-colony plants. The mean rate of N accumulation per unit root biomass was more than 300% greater in the ON- than OFF-colony population. Colony plants initially had a greater proportion of biomass and N remaining after defoliation in roots. Initial differences between populations in the distribution of biomass and N were eliminated as colony plants concentrated 24-day accumulation of biomass and N in aboveground structures. The data suggest that the combination of growth, N nutrition, and biomass and N distribution characteristics of the colony population likely confer a high rate of resource capture on heavily grazed prairie dog colonies.

Plant-herbivore interactions in a North American mixed-grass prairie : III. Soil nematode populations and root biomass on Cynomys ludovicianus colonies and adjacent uncolonized areas.

Seasonal dynamics of soil nematodes and root biomass were examined from under western wheatgrass (Agropyron smithii) and little bluestem (Andropogon scoparius) from a heavily grazed prairie dog (Cynomys ludovicianus) colony occupied for 5 to 10 years and an adjacent lightly grazed, uncolonized area in Wind Cave National Park, South Dakota, USA. Nematodes were differentiated into classes of plant-parasitic Tylenchida and Dorylaimida and nonparasitic Dorylamida and Rhabditida. Root-feeding nematodes were generally more numerous from A. smithii than from A. scoparius, while nonparasitic populations were not different in soil from beneath the two plant species. Rhabditida, parasitic Dorylaimida and Tylenchida (from A. scoparius only) were more numerous on the prairie dog colony than from the uncolonized site, but nonparasitic Dorylaimida populations did not differ between the two areas. Mean total (live plus dead) root biomass beneath A. scoparius and A. smithii on the prairie dog colony averaged 71% and 81%, respectively, of values from the uncolonized area. Estimated consumption by root-feeding nematodes averaged 12.6% and 5.8% of annual net root production in the upper 10 cm from the prairie dog colony and uncolonized site, respectively. We conclude that, because of microhabitat modification or reductions in plant resistance to nematodes, heavy grazing by aboveground herbivores apparently facilitates grazing by belowground herbivores. Because heavily grazed plants have less roots than lightly grazed or ungrazed plants, the impact of root-feeding nematodes on primary producers is likely to be greatest in heavily grazed grasslands.

Investigations of trypsin inhibitors in leaves of four North American prairie grasses.

Leaves of four important North American prairie grasses (Agropyron smithii, Andropogon gerardii, A. scoparius, and Bouteloua gracilis) were examined for the presence of trypsin inhibitors which are thought to protect some plant species from herbivory. Leaves of all four species had significant inhibitor activity at levels comparable to those in tomato leaves. Our evidence suggests that part of the inhibitor activity was proteinaceous and part may have been polyphenolic. Young leaves ofA. Smithii had more inhibitory activity than old leaves, but no leaf age differences were observed inB. gracilis. Mechanical wounding of leaves ofA. Smithii caused no consistent increase in inhibitor activity, in contrast to reports for some plant species. Plants ofA. smithii collected from areas heavily grazed by prairie dogs had trypsin inhibitor levels comparable to those in plants collected from a grazing exclosure. Thus, the ecological role of proteinase inhibitors in these Great Plains dominants remains to be demonstrated.

Defoliation responses of western wheatgrass populations with diverse histories of prairie dog grazing.

Photosynthesis and regrowth were compared over a 10-day period following defoliation of about 75% of the tillers of western wheatgrass (Agropyron smithii) plants collected from a black-tailed prairie dog (Cynomys ludovicianus) town and a grazing exclosure at Wind Cave National Park, South Dakota. Prior to defoliation, dog town plants had more tillers, but fewer leaves per tiller, shorter and narrower leaf blades, more horizontal leaves, and higher leaf blade/leaf sheath ratios than plants from the grazing exclosure. Rates of net photosynthesis (PN) did not differ significantly among plants of the two populations, either prior to or following defoliation. From Days 2-10 following defoliation, PN of remaining undamaged leaves averaged 104% of predefoliation rates while PN of similar leaves on non-defoliated plants declined steadily with time. averaging only 79% predefoliation rates during this period. Following defoliation, transpiration rates followed similar trends to CO2 exchange, and rates did not differ between plants of the two populations. Absolute rates of leaf elongation and shoot production were greater in plants from the exclosure. However, defoliation of plants from the exclosure population resulted in a 20% reduction in their cumulative shoot dry weight, while cumulative shoot dry weight of plants from the prairie dog town was not significantly affected by defoliation. This apparent ability of plants from the prairie dog town population to withstand defoliation better than plants from the exclosure was atributed to factors such as the higher leaf blade/leaf sheath ratio and more horizontal leaf angles of plants from the former population.

Plant-herbivore interactions in a North American mixed-grass prairie : I. Effects of black-tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plant species diversity.

Research was conducted to determine the effects of a native, sedentary rodent of North American grasslands, the black-tailed prairie dog (Cynomys ludovicianus), on seasonal aboveground plant biomass and nutrient dynamics and plant species diversity. The study was done on a northern mixed-grass prairie site at wind Cave National Park, South Dakota.Peak live plant biomass was greatest (190 g/m2) on the uncolonized part of the study area and least (95 g/m2) on a part of the prairie dog town colonized for 3 to 8 y. Peak live plant biomass (170 g/m2) of the oldest portion of the prairie dog town (colonized >26 y) was not significantly different from that of uncolonized prairie. However, where-as graminoids composed >85% of the total biomass of the latter area, forbs and dwarf shrubs (Artemisia frigida) were >95% of the total of the former. Both standing-dead plant biomass and litter declined markedly as time since colonization increased. Total plant species diversity (H) was greatest in the young prairie dog town (colonized for 3 to 8 y).Nitrogen concentration of plant shoots varied significantly as a function of time since colonization. Shoot-nitrogen was lowest in plants from the uncolonized site and greatest in plants collected from the longest-colonized areas of the prairie dog town. Shoot-nitrogen declined significantly over the growing season and tended to be higher in C3 graminoids than in C4 graminoids. In vitro digestible dry matter showed similar trends; the differences between C3 and C4 digestibilities were greatest during the last half of the growing season.We suggest that prairie dog-induced changes in plant biomass, plant species diversity, plant nutrient content, and forage digestibility may lead to further alterations of nutrient cycling and trophic dynamics in this mixed-grass prairie ecosystem.

Plant-herbivore interactions in a North American mixed-grass prairie : II. Responses of bison to modification of vegetation by prairie dogs.

Studies were conducted during the 1979 growing season to examine how North American bison (Bison bison) use prairie dog (Cynomys ludovicianus) colonies in Wind Cave National Park, South Dakota. Objectives included (1) determining whether bison selected for prairie dog towns parkwide; (2) characterizing in greater detail bison use patterns of a 36-ha colony in Pringle Valley as a function of time since prairie dog colonization; and (3) relating these bison use patterns to measured changes in structure and nutritional value of vegetation on and off the dog town.During midsummer, prairie dog towns were one of the most frequently used habitats by bison parkwide. Day-long observations at Pringle Valley revealed that bison exerted strong selection (nearly 90% of all habitat use and feeding time) for the dog town, which occupied only 39% of the valley. While there, they partitioned their use of the colony by grazing in moderately affected areas (occupied <8 years by prairie dogs) and by resting in the oldest area (>26 years occupation).Prairie dogs facilitate bison habitat selection for a shortgrass successional stage in this mixed-grass community by causing a broad array of compositional, structural, and nutritional changes in the vegetation.

Effects of carbofuran on growth, biomass allocation, and intraspecific competition in Bouteloua gracilis.

Growth, biomass allocation and competition between blue grama plants were examined with and without application of carbofuran, a pesticide which has been used to study insect and nematode effects on primary production. Carbofuran had no apparent effect on total plant growth, biomass allocation, or competition between neighboring plants.

Examination of North American bison saliva for potential plant growth regulators.

A series of laboratory bioassays were utilized to test for the presence of potential plant growth factors in saliva from a large native ungulate, the North American bison (Bison bison L.). Whole saliva enhancedAvena coleoptile growth at high pH, whether alone or in combination with indoleacetic acid (IAA). However, this enhancement was a result of salts in the saliva (primarily NaHCO3) rather than of other compounds acting hormonally, enhancing IAA activity, or inhibiting IAA oxidase activity as possibly occurs with some insect salivas. Additionally, the absence of detectable cytokinins in the saliva was indicated by its failure to enhance cucumber cotyledon expansion. This suggests that biochemical control of plant growth by salivary compounds following grazing is probably not an important component of this ruminant's interactions with its food plants, as has been suggested for some herbivores.

Relative growth rates and the grazing optimization hypothesis.

A mathematical analysis of the changes in plant relative growth rates necessary to increase aboveground production following grazing was conducted. The equation derived gives an isoline where production of a grazed and ungrazed plant will be the same. The equation has four variables (mean shoot relative growth rate, change in relative growth rate after grazing, grazing intensity, and recovery time) and may be analyzed graphically in a number of ways.Under certain conditions, small increases in shoot relative growth rate following grazing will lead to increased aboveground production. Under other conditions, very large increases in relative growth rate after grazing can occur without production being increased over that of ungrazed plants. Plants growing at nearly their maximum potential relative growth rate have little opportunity to respond positively to grazing and potentially can sustain less grazing than plants with growth rates far below maximum. Plants with high relative growth rates at the time of grazing require large increases in growth rate while slow growing plants require only small increases. High grazing intensities are least likely to increase production and high grazing frequencies require greater responses than infrequent grazing events.

Plant-herbivore interactions: Examination of potential effects of bison saliva on regrowth of Bouteloua gracilis (H.B.K.) lag.

Laboratory experiments were performed to determine whether regrowth of blue grama was affected by potential growth-promoting substances in saliva of North American bison. We observed no statistically significant effects of foliar application of whole bison saliva on net photosynthesis (PN), root respiration (RR), allocation patterns of photosynthetically fixed 14C, or regrowth rates over a 10-day period following clipping to various heights. In a 10-week experiment, there were no significant effects of saliva on leaf, crown or root growth or tiller production in plants clipped to heights of 6, 4 or 2 cm above crowns. Similarly, nitrogen-stressed plants failed to show significant changes in growth rates or tillering in response to saliva over a 3-week period. Clipped blue grama plants did exhibit significant compensatory growth responses, including higher PN rates from 3-10 days following clipping and allocation of a higher proportion of current photosynthate to synthesis of new leaf tissue with increasing severity of defoliation. Nevertheless, unclipped plants invariably outproduced clipped plants following defoliation.

Net photosynthesis, root respiration, and regrowth of Bouteloua gracilis following simulated grazing.

Net photosynthesis (PN), root respiration (RR), and regrowth of Bouteloua gracilis (H.B.K.) Lag. were examined in the laboratory over a 10-day period following clipping to a 4-cm height to simulate grazing by large herbivores. Net photosynthesis rates of tissue remaining immediately following defoliation were only about 40% as great as preclipping rates. Three days after clipping, PN rates of defoliated plants had increased to values about 21% greater (per unit leaf area) than those of unclipped controls and remained at that level through Day 10. No statistically significant changes in RR occurred following defoliation. Biomass of unclipped plants nearly doubled during the 10-day study period, while that of defoliated plants increased 67%. Over half the new growth of defoliated plants was allocated to new leaf blades and only 18% to new roots, while only 33% of the new growth of control plants was allocated to new leaf blades but 29% went to new roots. As a consequence of increased PN rates and increased carbon allocation to synthesis of additional photosynthetic tissue following defoliation, net CO2 uptake per plant increased from 9% to 80% of that of the controls from Day 0 through Day 10.

A simulation model of Bouteloua gracilis biomass dynamics on the North American shortgrass prairie.

A grassland primary producer model for simulating intraseasonal biomass dynamics as a function of temperature, moisture, light, and nitrogen was developed for Bouteloua gracilis (H.B.K.) Lag., the dominant C4 grass of the North American shortgrass prairie. Plant state variables included young and mature leaves, crowns, and roots from three depth categories while simulated processes included spring regrowth, photosynthesis, respiration, photosynthate allocation, death, and litterfall. Sensitivity analyses revealed the model was most sensitive to changes in photosynthesis and photosynthate allocation and least sensitive to changes in initial values of state variables, leaf dark respiration rates, and rate of spring regrowth.An abiotic submodel driven by observed weather data was used in conjunction with the primary producer model to simulate plant biomass dynamics under a variety of conditions including untreated controls (C), nitrogen fertilization (F), irrigation (I), and irrigation plus fertilization (IF). Model predictions of life shoot biomass (B s) and annual aboveground net primary production (NPP A) followed the same trends as field measurements with B sand NPP Aof IF>I>F>C. Failure of the model to accurately predict measured declines in peak B sand NPP Aafter several years of irrigation may have been caused by failure to account for growth lags following water stress, inadequate simulation of interspecific competition, or failure to simulate response to some mineral nutrients which had become limiting after several years of this treatment. A simulated annual carbon budget for plants in the four treatments suggests that from 61% (IF) to 80% (C) of the net carbon fixed above ground is ultimately translocated and utilized below ground.

An empirical model for estimating CO2 exchange of Bouteloua gracilis (H.B.K.) Lag. in the shortgrass prairie.

An empirical model for predicting net photosynthesis (P N ) and dark respiration (R D ) in the field was developed and tested for Bouteloua gracilis (H.B.K.) Lag., the dominant C4 grass of the North American shortgrass prairie. P N is predicted as a function of soil water potential, canopy air temperature, irradiance, and plant age, while R D is expressed as a function of soil water potential and temperature. The model accounted for 85% of the variability in the data base used to estimate parameter values. Results of a validation test showed good agreement between observed and predicted P N rates, suggesting this approach would be useful as a submodel of a grassland ecosystem model.