Plant nitrogen and phosphorus limitation in 98 North American grassland soils

The availability of nutrients is a critical determinant of ecological dynamics in grasslands, but the relationships between soil resource availability and nutrient limitation across ecosystems are not clear. To better understand how soil nutrient availability determines nutrient limitation in vegetation, we grew the same species of grass (Schizachyrium scoparium) in 98 North American grassland soils and fertilized them factorially with nitrogen (N) and phosphorus (P). On average adding N, P, and the two nutrients together increased biomass relative to unfertilized plants by 81%, 22%, and 131%, respectively. Plants grown on low-P soils were not primarily limited by P. Instead, these plants were colimited by N and P, while plants grown on high-P soils were primarily limited by N and only secondarily limited by P. Limitation was not predicted by total soil N. The preponderance of colimitation between N and P on low-P soils suggests that low P availability alters the N cycle to constrain supplies to plants such that N and P are made available in proportion to their demand by plants.     

Shading reduces exploitation of soil nitrate and phosphate by Agropyron desertorum and Artemisia tridentata from soils with patchy and uniform nutrient distributions

 Shading may both lessen the demand for soil nutrients and also the energy supply for nutrient acquisition. Since root foraging for nutrients in patchy environments can be energy-costly, especially for an immobile nutrient such as phosphate (P), the effects of shading may be most expected in heterogeneous soils. Plant acquisition of nitrate (N) and phosphate from soils with patchy and uniform nutrient distributions was determined in a field study under open sunlight and with shading for two common perennial Great Basin shrub steppe species, Agropyron desertorum and Artemisia tridentata. Partial shading in a pattern which can occur in shrub steppe vegetation significantly decreased plant N and P acquisition from soils both in the patchy and the uniform nutrient treatments. Artemisia was more affected by the shading than was Agropyron. Exploitation of the rather immobile P ion by both species was reduced to a much greater degree by the shading in the patchy distribution treatment than in the uniform nutrient treatment. As expected, plant acquisition of the more mobile N varied little with nutrient distribution treatment for both species and the depression of N acquisition by shading was the same in both nutrient distributions. The effects of shading appeared to have had its primary influence on different components of root foraging in the two species, especially in the nutrient-rich patches. For Agropyron shading primarily affected root proliferation, as indicated by reduced root density in patches. For Artemisia, shading most influenced root physiological uptake capacity and this was most pronounced in the nutrient-rich patches. While aboveground competition for light may generally reduce nutrient acquisition, the effects appear to be most pronounced if root systems of these steppe species are foraging for nutrients such as P in spatially heterogeneous soils. Received: 29 February 1996 / Accepted: 16 July 1996     

Combined effects of nitrogen enrichment, sulphur pollution and climate change on fen meadow vegetation N:P stoichiometry and biomass

Nitrogen (N) and sulphur (S) deposition, as well as altered soil moisture dynamics due to climate change can have large effects on fen meadow biogeochemistry and vegetation. Their combined effects may differ strongly from their separate effects, since each process affects different nutrients through different mechanisms. However, the impacts of these environmental problems are rarely studied in combination. We therefore investigated the separate and interactive effects of current levels of N- and S-deposition and changes in soil moisture dynamics on fen meadow vegetation. We focused on vegetation biomass and N:P stoichiometry, including access to soil P through root surface phosphatase activity, in a 3-year factorial addition experiment in an N-limited rich fen meadow in the Biebrza valley in Poland. We applied 29.5 kg N ha^?1 year^?1 and 32.1 kg S ha^?1 year^?1, which correspond to current deposition levels in Western Europe. Changes in soil moisture dynamics due to climate change were mimicked by amplified drying of the soil in summer. This level of N-deposition had limited effects on plant biomass production in this rich fen, despite low foliar N:P ratios that suggest N limitation. This level of S-deposition, however, resulted in decreased vegetation P-uptake and biomass. We also showed that increased summer drought resulted in considerable increases in vegetation biomass. We found no interactive effects on vegetation biomass or N:P stoichiometry, possibly as a result of the limited main effects of the separate processes.     

Fire, native species, and soil resource interactions influence the spatio-temporal invasion pattern of Bromus tectorum

Bromus tectorum (cheatgrass) is an invasive annual that occupies perennial grass and shrub communities throughout the western United States. Bronus tectorum exhibits an intriguing spatio‐temporal pattern of invasion in low elevation ponderosa pine Pinus ponderosa/bunchgrass communities in western Montana where it forms dense rings beneath solitary pines following fire. This pattern provides a unique opportunity to investigate several indirect effects of native vegetation that influence the invasion pattern of B. tectorum, and specifically how native species, disturbance, and soil resources interact to influence the spatio‐temporal pattern of invasion. We established four replicate field sites, each containing burned‐tree, burned‐grass, unburned‐tree, and unburned‐grass sampling locations, and initiated a series of field sampling and greenhouse experiments utilizing these locations. The objective of our first greenhouse experiment was to identify whether belowground factors contributed to the pattern of B. tectorum biomass observed in these field locations. This experiment generated a B. tectorum biomass response that was nearly identical to the invasion pattern observed in the field, suggesting further investigation of belowground factors was necessary. We measured resin‐sorbed NH₄⁺ and NO₃⁻ during one generation of B. tectorum, and measured a suite of P fractions through a sequential extraction procedure from these soils. These data revealed that a resource island of high N and P exists beneath pine trees. Through a second greenhouse experiment, we determined that N limited B. tectorum biomass in tree soil, whereas P limited biomass in bunchgrass soil. Finally, through a germination experiment we determined that pine litter strongly inhibited B. tectorum germination. These data suggest B. tectorum is regulated by P in bunchgrass soil, and by N and inhibition by pine litter beneath trees, effects that are likely alleviated by fire. These data demonstrate the combined role of direct and indirect interactions between native and invasive species in regulating biological invasions.     

Fire effects on ecosystem nitrogen cycling in a Californian bishop pine forest

Fire can cause severe nitrogen (N) losses from grassland, chaparral, and temperate and boreal forest ecosystems. Paradoxically, soil ammonium levels are markedly increased by fire, resulting in high rates of primary production in re-establishing plant communities. In a manipulative experiment, we examined the influence of wild-fire ash residues on soil, microbial and plant N pools in a recently burned Californian bishop pine (Pinus muricata D. Don) forest. Ash stimulated post-fire primary production and ecosystem N retention through direct N inputs from ash to soils, as well as indirect ash effects on soil N availability to plants. These results suggest that redistribution of surface ash after fire by wind or water may cause substantial heterogeneity in soil N availability to plants, and could be an important mechanism contributing to vegetation patchiness in fire-prone ecosystems. In addition, we investigated the impact of fire on ecosystem N cycling by comparing ¹⁵N natural abundance values from recently burned and nearby unburned P. muricata forest communities. At the burned site, ¹⁵N natural abundance in recolonising species was similar to that in bulk soil organic matter. By contrast, there was a marked ¹⁵N depletion in the same species relative to the total soil N pool at the unburned site. These results suggest that plant uptake of nitrate (which tends to be strongly depleted in ¹⁵N because of fractionation during nitrification) is low in recently burned forest communities but could be an important component of eco- system N cycling in mature conifer stands. Received: 29 June 1999 / Accepted: 24 October 1999     

Interactive Effects Between Reindeer and Habitat Fertility Drive Soil Nutrient Availabilities in Arctic Tundra

Herbivores impact nutrient availability and cycling, and the net effect of herbivory on soil nutrients is generally assumed to be positive in nutrient-rich environments and negative in nutrient-poor ones. This is, however, far from a uniform pattern, and there is a recognized need to investigate any interactive effects of herbivory and habitat fertility (i.e., plant C/N ratios) on soil nutrient availabilities. We determined long-term effects of reindeer on soil extractable nitrogen (N) and phosphorus (P) and their net mineralization rates along a fertility gradient of plant carbon (C) to N and P ratios in arctic tundra. Our results showed that reindeer had a positive effect on soil N in the more nutrient-poor sites and a negative effect on soil P in the more nutrient-rich sites, which contrasts from the general consensus. The increase in N availability was linked to a decrease in plant and litter C/N ratios, suggesting that a shift in vegetation composition toward more graminoids favors higher N cycling. Soil P availability was not as closely linked to the vegetation and is likely regulated more by herbivore-induced changes in soil physical and chemical properties. The changes in soil extractable N and P resulted in higher soil N/P ratios, suggesting that reindeer could drive the vegetation toward P-limitation. This research highlights the importance of including both the elements N and P and conducting studies along environmental gradients in order to better understand the interactive effects of herbivory and habitat fertility on nutrient cycling and primary production.     

Smoke interacts with fire history to stimulate soil seed bank germination in Mediterranean woodlands

Aims Fire has important consequences on vegetation dynamics. In fire-prone areas, natural selection favors plant species, characterized by a large soil seed bank, and that their germination is stimulated by fire. Although seed germination stimulated by fire heat is common in the eastern Mediterranean Basin, only little is known about germination stimulation by smoke. We examined the interactive effect of aerosol smoke and fire history on the germinable soil seed bank (GSSB) community in eastern Mediterranean woodlands.     

Four-year growth dynamics of beech-spruce model ecosystems under CO2 enrichment on two different forest soils

To elucidate how atmospheric CO₂ enrichment, enhanced nutrient supply and soil quality interact to affect regrowth of temperate forests, young Fagus sylvatica and Picea abies trees were grown together in large model ecosystems. Identical communities were established on a nutrient-poor acidic and on a more fertile calcareous soil and tree growth, leaf area index, fine root density and soil respiration monitored over four complete growing seasons. Biomass responses to CO₂ enrichment and enhanced N supply at the end of the experiment reflected compound interest effects of growth stimulation during the first two to three seasons rather than persistent stimulation over the whole duration of the experiment. Whereas biomass of Picea was enhanced in elevated CO₂ on both soils, Fagus responded negatively to CO₂ on acidic but positively on calcareous soil. Biomass of both species profited from enhanced N supply on the poor acidic soil only. Leaf area index on both soils was greater in high N supply as a consequence of a stimulation early in the experiment, but was unaffected by CO₂ enrichment. Fine root density on acidic soil was increased in high N supply, but this did not stimulate soil respiration rate. In contrast, elevated CO₂ stimulated both fine root density and soil CO₂ efflux on calcareous soil, especially towards the end of the experiment. Our experiment suggests that future species dominance in beech-spruce forests is likely to change in response to CO₂ enrichment, but this response is subject to complex interactions with environmental factors other than CO₂, particularly soil type.     

Positive feedbacks between decomposition and soil nitrogen availability along fertility gradients

Background and aims

We determined the relationship between site N supply and decomposition rates with respect to controls exerted by environment, litter chemistry, and fungal colonization.

Methods

Two reciprocal transplant decomposition experiments were established, one in each of two long-term experiments in oak woodlands in Minnesota, USA: a fire frequency/vegetation gradient, along which soil N availability varies markedly, and a long-term N fertilization experiment. Both experiments used native Quercus ellipsoidalis E.J. Hill and Andropogon gerardii Vitman leaf litter and either root litter or wooden dowels.

Results

Leaf litter decay rates generally increased with soil N availability in both experiments while belowground litter decayed more slowly with increasing soil N. Litter chemistry differed among litter types, and these differences had significant effects on belowground (but not aboveground) decay rates and on aboveground litter N dynamics during decomposition. Fungal colonization of detritus was positively correlated with soil fertility and decay rates.

Conclusions

Higher soil fertility associated with low fire frequency was associated with greater leaf litter production, higher rates of fungal colonization of detritus, more rapid leaf litter decomposition rates, and greater N release in the root litter, all of which likely enhance soil fertility. During decomposition, both greater mass loss and litter N release provide mechanisms through which the plant and decomposer communities provide positive feedbacks to soil fertility as ultimately driven by decreasing fire frequency in N-limited soils and vice versa.     

火烧对森林土壤有机碳的影响研究进展

对国内外火烧影响森林土壤有机碳动态的研究成果进行了综合述评。较多研究表明低强度火烧不会造成土壤有机碳贮量的明显变化,但火烧非常强烈而彻底,土壤有机碳明显减少。有限研究表明火烧对森林土壤呼吸的影响结果有增加、降低或无影响,因火烧强度、火后观测时间、森林类型、火烧迹地上植被恢复进程和气候条件等而异。同时,火烧对土壤有机碳组分(活性有机碳和黑碳)也具有不同程度的影响。随着全球变化研究的深入,火烧作为森林主要管理措施对大气CO2浓度影响亦愈来愈受重视,今后应着重开展以下几方面研究:(1)扩大气候和经营管理的变化对森林土壤有机碳贮量时空动态影响研究;(2)深入探讨火烧影响土壤CO2释放的过程及机理;(3)加强火烧历史和频率对黑碳影响的研究;(4)从广度和深度上加强火烧等经营措施对亚热带森林土壤碳动态影响的研究。     

Frequent fire affects soil nitrogen and carbon in an African savanna by changing woody cover

When tropical and sub-tropical ecosystems burn, considerable amounts of N present in the biomass fuel may be released. This ultimately results in a loss of fixed N to the atmosphere. It is often assumed that this volatilization loss of N with frequent fire will result in a reduction of plant-available N and total system N. By changing the amount of woody biomass fire may, however, also have indirect effects on N and C dynamics. Here we consider the effects of 50 years of frequent fire on total soil N and soil organic C (SOC) and total soil N in a mesic savanna in the Kruger National Park, South Africa. We also determine how changes in woody biomass may affect total soil N and SOC. We measured soil and fine root N and C concentrations as well as total soil N and SOC pools in four burning treatments, including fire exclusion, of a long-term fire experiment. Our results show that regardless of soil depth, fire treatment had no significant effect on total soil N and SOC. Our results also show that under trees total soil N and SOC concentrations of the surface soil increase, and pools of N and SOC increase to a depth of 7 cm. However, the extent to which soil N and C dynamics differed under canopies and away from canopies was dependent on fire treatment. Our results show that the effect of fire on soil N and C is mediated both through the indirect effect of changes in woody cover and the direct effects of fire (volatilization losses of nutrients). We suggest that woody thickening in this mesic savanna will have pronounced effects on long-term N and C dynamics.     

Fire-mediated interactions between shrubs in a South American temperate savannah

We examined spatial patterns of fire-caused mortality and after-fire establishment of two dominant shrub species, Baccharis dracunculifolia and Eupatorium buniifolium in a humid temperate South American savannah. Our objective was to determine whether fires mediate in interactions between these two species. After a natural fire burned a large tract of savannah, we established two plots (respectively 550 and 500 m²) within which we mapped all surviving and dead shrubs as well as all individuals of shrub species that recruited in the following year. We used techniques of point-pattern analysis to test specific null hypotheses about spatial associations in the distribution, mortality, and establishment of shrubs. Results support the notions that fire mediates interactions between these two species. Fire-caused death of E. buniifolium tended to occur selectively in the vicinities of Baccharis individuals, and recruitment of B. dracunculifolia tended to be concentrated in the places of dead shrubs. These responses, however, were contingent on local abundances of shrubs which depend in part from the recent fire history. Anthropogenic perturbation of the natural fire regime would have therefore distorted the role of fire mediated interactions as drivers of the dynamics of the vegetation of this temperate savannah.     

Vegetation cover regulates the quantity,quality and temporal dynamics of dissolved organic carbon and nitrogen in Antarctic soils

Populations of the two native Antarctic vascular plant species (Deschampsia antarctica and Colobanthus quitensis) have expanded rapidly in recent decades, yet little is known about the effects of these expansions on soil nutrient cycling. We measured the concentrations of dissolved organic carbon (DOC) and nitrogen (DON), amino acids and inorganic N in soils under these two vascular plant species, and under mosses and lichens, over a growing season at Signy Island in the maritime Antarctic. We recorded higher concentrations of nitrate, total dissolved nitrogen, DOC, DON and free amino acids in soil under D. antarctica and C. quitensis than in lichen or moss dominated soils. Each vegetation cover gave a unique profile of individual free amino acids in soil solution. Significant interactions between soil type and time were found for free amino acid concentrations and C/N ratios, indicating that vascular plants significantly change the temporal dynamics of N mineralization and immobilization. We conclude that D. antarctica and C. quitensis exert a significant influence over C and N cycling in the maritime Antarctic, and that their recent population expansion will have led to significant changes in the amount, type and rate of organic C and N cycling in soil.     

Influence of smoke,heat and fire on germination of woody species occurring in the dry valleys of southwest China

Aims Savannahs depend on fire for their persistence. Fire influences regeneration from seeds in several ways: it converts the environment into a more open space which can benefit the establishment of seedlings, and fire itself can also enhance germination by chemical and physical cues, such as smoke and heat. There is limited information as to how seed of Asian savannah species respond to fire, even though Asia has several dry vegetation types that are associated with fire. Our main question was whether fire enhances or triggers the germination of woody species occurring in southwest Chinese dry valleys, which have savannah vegetation.     

Alteration and Recovery of Slash Pile Burn Sites in the Restoration of a Fire‐Maintained Ecosystem

Restoration practices incorporating timber harvest (e.g. to remove undesirable species or reduce tree densities) may generate unmerchantable wood debris that is piled and burned for fuel reduction. Slash pile burns are common in longleaf pine ecosystem restoration that involves hardwood removal before reintroduction of frequent prescribed fire. In this context, long‐lasting effects of slash pile burns may complicate restoration outcomes due to unintended alterations to vegetation, soils, and the soil seed bank. In this study, our objectives were to (1) examine alterations to the soil seed bank, soil physical and chemical characteristics, and initial vegetation recolonization following burn and (2) determine the rate of return of soil and vegetation characteristics to pre‐burn conditions. We found that burning of slash piles (composed of scores of whole trees) results in elevated nutrient levels and significant impacts on vegetation and the soil seed bank, which remain evident for at least 6 years following burn. In this ecosystem, formerly weakly acidic soils become neutral to basic and levels of P remain significantly higher. Following an initial decrease after burn, total soil N increases with time since burn. These changes suggest that not only does pile burning create a fire scar initially devoid of biota, but it also produces an altered soil chemical environment, with possible consequences for long‐term ecosystem restoration efforts in landscapes including numerous fire scars. To facilitate restoration trajectories, further adaptive management to incorporate native plant propagules or suppress encroaching hardwoods within fire scars may be warranted in fire‐dependent ecosystems.     

Effects of different nitrogen:phosphorus levels on the growth and ecological stoichiometry of Glycyrrhiza uralensis

Aims The increase in atmospheric nitrogen (N) deposition has accelerated N cycling of ecosystems, probably resulting in increases in phosphorus (P) demand of ecosystems. Studies on the effects of artificial N:P treatment on the growth and carbon (C), N, P ecological stoichiometry of desert steppe species could provide not only a new insight into the forecasting of how the interaction between soils and plants responses to long-term atmospheric N deposition increase, but also a scientific guidance for sustainable management of grassland in northern China under global climate change. Methods Based on a pot-cultured experiment conducted for Glycyrrhiza uralensis (an N-fixing species) during 2013 to 2014, we studied the effects of different N:P supply ratios (all pots were treated with the same amount of N but with different amounts of P) on aboveground biomass, root biomass, root/shoot ratio, and C:N:P ecological stoichiometry both in G. uralensis (leaves and roots) and in soils. Additionally, through the correlation analyses between biomass and C:N:P ecological stoichiometry in leaves, roots, and soils, we compared the differences among the C:N:P ecological stoichiometry of the three pools, and discussed the indication of C:N:P ecological stoichiometry in soils for the growth and nutrient uptake of G. uralensis. Important findings The results showed that, reducing N:P decreased C:P and N:P ratios both in G. uralensis (leaves and roots) and in soils but increased aboveground biomass and root biomass of G. uralensis, indicating that low to moderate P addition increased P availability of soils and P uptake of G. uralensis. However, excessive low N:P (high P addition) led to great decreases in soil C:P and N:P ratios, thus hindering N uptake and the growth of G. uralensis. C:N:P ratios in the two pools of G. uralensis (especially in leaves) had close correlations with soil C:N:P ratio, indicating that the change in soil C:N:P ratio would have a direct influence on plants. Our results suggest that, through regulating C:N:P ratio in leaves and soils, appropriate amounts of P addition could balance soil P supply and plant P demand and compensate the opposite influences of long-term atmospheric N deposition increase on the structure of desert steppe.     

Nitrogen deposition and plant species interact to influence soil carbon stabilization

Anthropogenic nitrogen (N) deposition effects on soil organic carbon (C) decomposition remain controversial, while the role of plant species composition in mediating effects of N deposition on soil organic C decomposition and long‐term soil C sequestration is virtually unknown. Here we provide evidence from a 5‐year grassland field experiment in Minnesota that under elevated atmospheric CO₂ concentration (560 ppm), plant species determine whether N deposition inhibits the decomposition of soil organic matter via inter‐specific variation in root lignin concentration. Plant species producing lignin‐rich litter increased stabilization of soil C older than 5 years, but only in combination with elevated N inputs (4 g m^?2 year^?1). Our results suggest that N deposition will increase soil C sequestration in those ecosystems where vegetation composition and/or elevated atmospheric CO₂ cause high litter lignin inputs to soils.     

Effects of different nitrogen:phosphorus levels on the growth and ecological stoichiometry of Glycyrrhiza uralensis北大核心CSCD

Aims The increase in atmospheric nitrogen (N) deposition has accelerated N cycling of ecosystems, probably resulting in increases in phosphorus (P) demand of ecosystems. Studies on the effects of artificial N:P treatment on the growth and carbon (C), N, P ecological stoichiometry of desert steppe species could provide not only a new insight into the forecasting of how the interaction between soils and plants responses to long-term atmospheric N deposition increase, but also a scientific guidance for sustainable management of grassland in northern China under global climate change. Methods Based on a pot-cultured experiment conducted for Glycyrrhiza uralensis (an N-fixing species) during 2013 to 2014, we studied the effects of different N:P supply ratios (all pots were treated with the same amount of N but with different amounts of P) on aboveground biomass, root biomass, root/shoot ratio, and C:N:P ecological stoichiometry both in G. uralensis (leaves and roots) and in soils. Additionally, through the correlation analyses between biomass and C:N:P ecological stoichiometry in leaves, roots, and soils, we compared the differences among the C:N:P ecological stoichiometry of the three pools, and discussed the indication of C:N:P ecological stoichiometry in soils for the growth and nutrient uptake of G. uralensis. Important findings The results showed that, reducing N:P decreased C:P and N:P ratios both in G. uralensis (leaves and roots) and in soils but increased aboveground biomass and root biomass of G. uralensis, indicating that low to moderate P addition increased P availability of soils and P uptake of G. uralensis. However, excessive low N:P (high P addition) led to great decreases in soil C:P and N:P ratios, thus hindering N uptake and the growth of G. uralensis. C:N:P ratios in the two pools of G. uralensis (especially in leaves) had close correlations with soil C:N:P ratio, indicating that the change in soil C:N:P ratio would have a direct influence on plants. Our results suggest that, through regulating C:N:P ratio in leaves and soils, appropriate amounts of P addition could balance soil P supply and plant P demand and compensate the opposite influences of long-term atmospheric N deposition increase on the structure of desert steppe.     

N and P Cycling in Tanzanian Humid Savanna: Influence of Herbivores, Fire, and N2-Fixation

Availabilities of nitrogen (N) and phosphorus (P) have a strong influence on plant growth and the species composition of savannas, but it is not clear how these availabilities depend on factors such as fire, N₂-fixation, and activities of wild herbivores and cattle. We quantified soil N and P availabilities in various ways (extractable pools, mineralization, resin adsorption) along vegetation gradients within a recently abandoned cattle ranch and a former game reserve in Tanzania (both areas now part of the Saadani National Park). We also assessed annual N and P balances to evaluate how long-term availabilities of N and P are affected by large herbivores, symbiotic N₂-fixation, and fire. The results show that cattle ranching led to a spatial re-distribution of nutrients, with the local accumulation of P being stronger and more persistent than that of N. In the former game reserve, intensively grazed patches of short grass tended to have elevated soil N and P availabilities; however, because quantities of nutrients removed through grazing exceeded returns in dung and urine, the nutrient balances of these patches were negative. In dense Acacia stands, N₂-fixation increased N availability and caused a net annual N input. Fire was the major cause for nutrient losses from tallgrass savanna, and estimated N inputs from the atmosphere and symbiotic N₂-fixation were insufficient to compensate for these losses. Our results call into question the common assumption that N budgets in annually burned savanna are balanced; rather, these ecosystems are a mosaic of patches with both N enrichment and impoverishment, which vary according to the vegetation type.     

Intraspecific adoption and foster feeding of fledglings in the North Island robin

We sampled soils and vegetation within and outside two sheep and rabbit exclosures, fenced in 1979, on steep sunny and shady slopes at 770 m altitude on seasonally-dry pastoral steeplands. The vegetation of sunny aspects was characterised by higher floristic diversity, annual species, and low plant cover. Here the exotic grass Anthoxanthum odoratum dominated on grazed treatments, and the exotic forb Hieracium pilosella on ungrazed. Shady aspects supported fewer, and almost entirely perennial, species. Here Hieracium pilosella dominated grazed treatments, but co-dominated with the exotic forb H. praealtum and the native grass Festuca novae-zelandiae on ungrazed treatments. There was 43% more biomass in exclosures (P < 0.01). Most of the biomass difference (4285 kg/ha) was from greater root mass (2400 kg/ha). 1385 kg/ha of the difference was from herbage and the remainder (500 kg/ha) from litter. Exclosures had 50 to 100% more Ca, Mg, K and P in the biomass (P < 0.05), but the effect on soils was limited to significantly higher concentrations of total N (P < 0.05) and exchangeable Mg (P < 0.01) in 0-7.5 cm soils. We conclude that stopping grazing for 16 years on seasonally-dry steeplands results in greater plant cover, approximately double the biomass of standing vegetation, greater biomass in roots, and more biomass nutrients relative to grazed areas. However, it does not favour native species and has little effect on soil nutrients or soil carbon. Stopping grazing alone therefore cannot be regarded as a comprehensive short- or medium-term vegetation or soil rehabilitation option.