Notes on the feeding of Ectemnorrhinus similis waterhouse (Curculionidae) adults on Marion Island

Summary Little is known regarding the role of arthropods as herbivores on subantarctic islands. In grassland and drainage line communities on Marion Island maximum densities (up to 220/m²) of Ectemnorrhinus similis adults occur in March to April, representing a maximum biomass of ca. 1 g m^-2. Captured beetles ingest approximately 14% of body weight per day of Acaena magellanica leaf material and 37% of body weight per day of Brachythecium rutabulum fronds. No significant feeding on Poa cookii leaves was observed. It is proposed that E. similis be regarded as a significant primary consumer on the island.     

Estimated Impact of feral house mice on sub-Antarctic invertebrates at Marion Island

Summary The energy metabolism of feral house mice Mus musculus was established on sub-Antarctic Marion Island, using the doubly-labelled water turnover technique. Mean water influx was 565 ml kg^-1 day^-1 and mean CO₂ production was 5.41 ml g^-1 h^-1, i.e. 3375 kJ kg^-1 day^-1. From the energy content of the main items (Lepidoptera larvae, Curculionidae) in the diet of the mice it was estimated that the dry mass of food consumed was 3.5 g mouse^-1 day^-1. The overall impact of mice on invertebrates, based on mean mouse density and the mean percentage invertebrates in the diet, was estimated at 108 g ha^-1 day^-1 or 39.4 kg ha^-1 y^-1 (dry mass). Greatest predation pressure was on larvae of the flightless moth Pringleophaga marioni: 65 g ha^-1 day^-1 or 23.7 kg ha^-1 y^-1. Insect biomass is lower on Marion Island than on nearby Prince Edward Island, which is mouse-free. It is suggested that populations of certain insects on Marion Island are depressed by the alien mice.     

Biology and ecology of theDusmoecetes Jeannel (Col. Curculionidae) species complex on Marion Island

Summary TheDusmoecetes species complex on sub-Antarctic Marion Island comprises two species. The larger species,D. similis (C.O. Waterhouse), feeds on angiosperms as adults and detritus as larvae, whereas the smaller species,D. marioni Jeannel feeds on bryophytes in all stages.D. similis has seven larval instars and a generation time of one year or longer.D. marioni has between five and seven larval instars, depending on the plant community it inhabits, and a generation time of one year or less. Immatures ofD. similis have a theoretical null point of development of -0.62°C and aQ ₁₀ of 3.57. In both species there are overlapping generations in the field, although in the case ofD. similis there is a distinct seasonal emergence of adults starting in September. Mean standing crop of larvae on the coastal plain is approximately 5.5 kg (dry mass)·ha^-1, but can be as high as 11.85 kg (dry mass)·ha^-1 inAzorella selago dominated communities. This study indicates that both species are important herbivores and/or detritivores on Marion Island.     

The diet and impact of house mice on a sub-Antarctic island

An analysis of the stomach contents of house mice (Mus musculus L.) at three sites on Marion Island (47°S, 38°E) over a 1-year period showed that the mice feed mostly on terrestrial macroinvertebrates but that plants (mainly seed) are an important component of the diet in mid- to late summer. Larvae of a flightless moth, weevil larvae and adults and, at one of the sites, earthworms were the invertebrate items that showed the highest importance value over the year. Diet diversity was slightly lower in summer and autumn than in winter and spring. Diet variety was lowest in midsummer and highest in either autumn or winter, depending on the site. Mean stomach-content mass was significantly higher in midwinter than during the rest of the year. When offered prey of different types and sizes, mice selected moth larvae first in 92% of the trials; earthworms or weevil adults were most frequently selected second, and weevil larvae fourth. An introduced slug species was selected only once, as the last choice and only part of it was eaten. With both moth larvae and earthworms, the heaviest individual offered was almost always chosen first. Surprisingly, mice never consumed more than about half of the seed offered to them and their condition deteriorated severely during the trials with seed; in a third of the trials with two types of seed, the mice completely ignored the seed. The impact of mice predation on invertebrates was assessed at two of the sites - a dry mire and a coastal biotic herbfield that is influenced by seabird and seal manuring. Moth larvae and adults, weevil larvae and adults, earthworms, spiders and flies made up >90% of the animal remains in the stomach contents over the year and, on average, the mice daily consumed 45 g (dry mass) ha^-1 of those invertebrates at the mire and 194 g ha^-1 at the biotic site. Moth larvae made up a substantial proportion of these amounts; average daily consumption was 30 g ha^-1 at the mire and 107 g ha^-1 at the biotic site. In terms of the amount consumed in relation to biomass, the heaviest impact of mice at both sites was on weevil adults; at the biotic site mice daily consumed 13% of weevil adult biomass in autumn and nearly 6 times the annual average biomass over the year. At the mire, slightly more than the annual average weevil adult biomass was consumed over the year. The consumption of invertebrates by mice found in this study (1992/1993) was about 20% greater than in 1979/1980. The most striking changes in annual average consumption rate between 1979/1980 and 1992/1993 were for weevil larvae (increased nearly fourfold) and earthworms (increased sixfold at the biotic site and threefold at the mire). In contrast, the consumption of spiders at both sites decreased between 1979/1980 and 1992/1993.     

Litterfall from two eucalypt woodlands in central Queensland

The mean annual litterfall at two dry woodland sites in central Queensland was 1129 kg ha^–1 for an open E. populnea F. Muell. woodland (n = 2 years), and 2318 kg ha^–1 for a woodland dominated by E. cambageana Maiden (n = 1 year). Leaves formed the largest component of total litterfall, which varied seasonally with a spring–summer maximum. Annual litterfall at these sites conformed with a pattern of decreasing litter production with declining annual rainfall, consistent with a range of eucalypt-dominated communities.     

Effect of typhoon disturbance on fine litterfall and related nutrient input in a subtropical forest on Okinawa Island, Japan

Typhoons are frequent on Okinawa Island, southwestern Japan. The effects of typhoon disturbance on the patterns of fine litterfall and related nutrient inputs in a subtropical evergreen broad-leaved forest were studied over 5 years from May 1996 to April 2001. Annual fine litterfall averaged 7558 kg ha⁻¹ (range from 6188 to 9439 kg ha⁻¹) for six sampling plots over 5 years, which differed significantly among years (p<0.001) but not among plots (p=0.122). A seasonal maximum was most evident for leaf litter component. Woody litter fell more irregularly through the year, and peak fall varied with typhoon and windstorm. The mean ratio of annual litterfall mass of sexual organs to leaves was 0.06, much lower than that in other tropical and subtropical rain forests. Nutrient concentrations varied in litterfall components, but were not significantly different among plots. The lowest concentrations of N and P in leaf litter were observed in March, which is also the month with the greatest leaf fall. However, the highest concentrations were recorded in typhoon season. Nitrogen and P concentrations were 34% and 106% greater in the green leaves that fell during typhoons than in senescent leaves. Mean nutrient inputs by litterfall were: N 83, P 3.2, K 25, Ca 71, Mg 19, Al 12, Na 10, Fe 0.86 and Mn 3.9 kg ha⁻¹ yr⁻¹, and differed significantly among years for all elements (p<0.0005) and among plots only for K (p<0.05) and Mn (p<0.0001). Typhoon disturbance strongly affected annual fine litterfall and related nutrient inputs, which contributed an average of 30% of the annual litterfall mass, and from 30% to 39% (for different nutrient elements) of annual total nutrient inputs. The results from this study suggest that typhoon-driven maintenance of rapid cycling of P and N and their high availability in soil appears to be an important mechanism to maintain productivity in the subtropical forest on Okinawa Island.     

Symbiotically fixed nitrogen from field- grown white and red clover mixed with ryegrasses at low levels of15N-fertilization

A field study was carried out near Zürich (Switzerland) to determine the yield of symbiotically fixed nitrogen (¹⁵N dilution) from white clover (Trifolium repens L.) grown with perennial ryegrass (Lolium perenne L) and from red clover (Trifolium pratense L.) grown with Italian ryegrass (Lolium multiflorum Lam.). A zero N fertilizer treatment was compared to a 30 kg N/ha per cut regime (90 to 150 kg ha⁻¹ annually). The annual yield of clover N derived from symbiosis averaged 131 kg ha⁻¹ (49 to 227 kg) without N fertilization and 83 kg ha⁻¹ (21 to 173 kg) with 30 kg of fertilizer N ha⁻¹ per cut in the seeding year. Values for the first production year were 308 kg ha⁻¹ (268 to 373 kg) without N fertilization and 232 kg ha⁻¹ (165 to 305 kg) with 30 kg fertilizer N ha⁻¹ per cut. The variation between years was associated mainly with the proportion of clover in the mixtures. Apparent clover-to-grass transfer of fixed N contributed up to 52 kg N ha⁻¹ per year (17 kg N ha⁻¹ on average) to the N yield of the mixtures. Percentage N derived from symbiosis averaged 75% for white and 86% for red clover. These percentages were affected only slightly by supplemental nitrogen, but declined markedly during late summer for white clover. It is concluded that the annual yield of symbiotically fixed N from clover/grass mixtures can be very high, provided that the proportion of clover in the mixtures exceeds 50% of total dry mass yield.     

Grass production and decomposition in Southern Guinea savanna,Nigeria

Summary Annual grass production in ungrazed plots was 2,731 kg ha^-1, litter production was 1,619 kg ha^-1 and decomposition was 1,789 kg ha^-1. In grazed plots the corresponding figures were 3,157 kg ha^-1, 1,440 kg ha^-1, and 1,475 kg ha^-1 respectively; cattle consumed 1,405 kg ha^-1. Litter disappearance was greatest in the dry season: 1,226 kg ha^-1 (69% of the annual total) disappearing in the 4 months of December to March in the ungrazed plots, largely due to consumption (790 kg ha^-1 in December to March) by fungus-growing termites (Macrotermitinae). A positive linear relationship was found between maximum grass biomass and annual rainfall in West Africa.     

Litter fall and litter layer decay in coastal scrub at Sydney,Australia

Summary Seasonal litter fall and changes in dry weight and minerals within the litter layer were sampled throughout one year. The annual total litter fall was 4.9 t per hectare of which 70% was leaf fall. Litter fall was highest in spring and early summer, the falls of each component (leaf, wood, etc.) and species being markedly seasonal but not all in phase. The annual amounts of minerals cycling as litter fall ranged from 44 kg N to 2 kg P per hectare. Mineral concentrations of the litter fall underwent seasonal variation and in the case of potassium were related to rainfall. The litter layer, mean biomass 19 t ha^-1, declined in weight at a rate of 2.4% per week between spring and summer. The mean turnover time for the dry weight of the litter layer was 3.8 years and for the mineral approx. 1 year (K, Cl) or in excess of 3 years (P, Ca, S, Mg and N).     

Respiratory Metabolism of sub-Antarctic insects from different habitats on Marion Island

Oxygen uptake of the foliage-dwelling larvae ofEmbryonopsis halticella Eaton (Lepidoptera: Yponomeutidae) and adults ofEctemnorhinus marioni Jeannel (Coleoptera: Curculionidae), the litter-dwelling larvae ofPringleophaga marioni (Lepidoptera: Tineidae) and the wrack-dwellingParactora dreuxi Séguy (Diptera: Helcomyzidae) was examined over the range of temperatures experienced by these insects in their microhabitats. With the exception of the kelp fly,P. dreuxi, Q₁₀s and activation energies were generally lower than those found in temperate and Arctic insects, but were similar to values found in beetles from sub-Antarctic South Georgia Island. Q₁₀ and activation energy of each species reflected the temperature regime found in its microhabitat. Activation energies of the Marion Island species were intermediate between those found in temperate and polar arthropods, but towards the polar end of the range. The hypothesis that insects are capable of showing respiratory adaptation to temperature is supported.     

The impact of a small, alien invertebrate on a sub-Antarctic terrestrial ecosystem: Limnophyes minimus (Diptera, Chironomidae) at Marion Island

Density and biomass of the larvae of a small, alien chironomid midge, Limnophyes minimus, whose parthenogenetic adult females do not feed, were quantified for ten major lowland plant communities at sub-Antarctic Marion Island (46°52′S 37°51′E) and compared with the density and biomass of indigenous macro-invertebrates in the same communities. An estimate of litter consumption by larvae of this midge was also made. L. minimus reached high densities in most of the plant communities sampled, with the highest density being recorded in the Cotula plumosa biotically influenced community (annual mean of 4,365 individuals m⁻²) and the lowest in the Crassula moschata salt spray community (annual mean of 41 individuals m⁻²). Estimates of litter ingestion indicated that L. minimus larvae are capable of consuming between 0.07 and 8.54 g_{(dry mass)} m⁻² per year, depending on the community. In some communities this litter consumption amounted to an order of magnitude more than that consumed by Pringleophaga marioni (Lepidoptera, Tineidae). Although the larvae of this moth species are thought to represent the bottleneck to nutrient recycling on the island, this study showed that midge larvae may also contribute substantially to this process. As a consequence, the considerable changes that have been predicted to occur in Marion Island's terrestrial ecosystem as a consequence of enhanced predation by mice on P. marioni larvae may be retarded or obscured by the contribution of the midge larvae to nutrient cycling. Hence, it is suggested that greater attention be given to the small and inconspicuous elements of the alien sub-Antarctic faunas because such species may have profound consequences for ecosystem functioning on these islands. Received: 17 November 1997 / Accepted: 23 February 1998     

Contingent absences account for range limits but not the local abundance structure of an invasive springtail

Understanding spatial variation in abundance is essential for forecasts of responses to environmental change impacts on diversity and the consequent conservation actions. However, few studies have sought to distinguish the causal basis of abundance structure and range limits. Here we do so for an invasive springtail species, Pogonognathellus flavescens, in a cold temperate island setting. Local microclimate variables and physiological tolerances of this habitat generalist suggest that it should be widely distributed across a range of habitats below 200 m elevation on the island. By contrast, island‐wide and local abundance surveys show that it is restricted to indigenous Poa cookii tussock grassland habitats in the southeast. Preference for this habitat is correlated with the presence of vertebrates and a threshold response to soil Ca and pH, with preference for low values of both. Habitat specificity may be the consequence of a founder effect because the species is characterized by only a single mitochondrial COI haplotype. Nonetheless, P. flavescens is absent from many other areas of suitable P. cookii habitat around the island. The most plausible explanation appears to be dispersal limitation (i.e. a contingent absence). Despite high locomotion speeds measured in the laboratory at optimum and mean annual soil temperatures (0.97 and 0.42 cm s^?1 at 26°C and 6.5°C, respectively), dispersal in the field indicated that >100 yr would be required to reach all available habitat in the absence of jump dispersal, for which few vectors exist. Thus, current range limits are set by dispersal limitation (i.e. contingent absences) whilst abundance structure is a function of variation in soil substrate quality. Edaphic variables both in this species and other soil invertebrates may be more significant than climatic factors in determining abundance and occurrence, indicating that they should be routinely included in species distribution models. Low genetic diversity and high habitat preference suggest that in the absence of introduction of additional individuals, the species will not spread rapidly at the island. However, over time, the widening distribution of its preferred habitat, P. cookii, as a consequence of a major management intervention (the eradication of feral cats), may enable it to colonize all suitable areas.     

Phenology of stranded kelp degradation by the kelp fly Paractora dreuxi mirabilis (Helcomyzidae) at Marion island

Summary During heavy seas, bull-kelp Durvillaea antarctica (Cham.) Har. plants are stranded either singly or in wrack beds on sub-Antarctic Marion Island's (46°54S, 37°45E) few pebbled beaches. In the supralittoral zone, Durvillaea wrack is consumed chiefly by kelp fly larvae (the endemic Paractora dreuxi mirabilis Séguy), whilst two marine amphipods attack wrack in the eulittoral and sometimes in the supralittoral zone. During the summer of 1984 on Marion Island, artificial Durvillaea deposits were established in the supralittoral zone of a beach close to the meteorological station and were closely monitored for six weeks thereafter. Eight days after deposition, temperatures inside the protected wrack bed reached 10°C above ambient, and P. dreuxi larvae attained a maximum biomass of 27 g/kg (dry mass) of kelp in the bed. After 30 days, wrack strings and a protected wrack bed had lost 80 and 65%, respectively, of their original dry mass. A bed left exposed to trampling by seals had disappeared. Over 30 days, P. dreuxi larvae accounted for 35% of the dry mass loss in the wrack bed. P. dreuxi has an opportunistic and flexible ecological strategy, facilitated by a stable, interstitial biotope of particulate kelp detritus from which it colonizes fresh wrack.     

Nitrogen transfer from arrowleaf clover to ryegrass in field plantings

Arrowleaf clover (Trifolium vesiculosum Savi) and annual ryegrass Lolium multiflorum Lam.) commonly are overseeded in dormant bermudagrass (Cynodon dactylon L. Pers.) sod on coastal plain soils in the southeastern United States. Two field experiments were conducted in consecutive years at different sites to estimate the amount of N transferred from the clover to the annual grass. Nitrogen treatments included 50 kg N ha^-1 as ¹⁵N depleted ammonium nitrate applied in either February or April, and a check (no N applied). Three clippings were made during the cool-season from March to June. In both experiments, less than 5 kg N ha^-1 were transferred from the clover to the grass. Ryegrass yields of dry matter and total N were not increased by growing with clover. Clover growth was typical for the region; average dry matter yield in pure stand was 2,615 kg ha^-1 over the two-year period. Clover in mixed stand fixed between 20 and 60 kg N/ha. Less than 13% of N contained in ryegrass was transferred from arrowleaf clover to ryegrass at any clipping while clover was actively growing. The quantity of N transferred over the entire season was not statistically significant.     

Feeding ecology and emergence production of annual cicadas (Homoptera: Cicadidae) in tallgrass prairie

The emergence phenology and feeding ecology of annual cicadas in tallgrass prairie are poorly documented. However, these large insects are abundant, and their annual emergence represents a potentially important flux of energy and nutrients from belowground to aboveground. We conducted a study at Konza Prairie Research Natural Area in eastern Kansas to characterize and quantify cicada emergence and associated energy and nutrient fluxes. We established emergence trap transects in three habitat types (upland prairie, lowland prairie, and riparian forest), and collected cicadas every 3 days from May to September. A subset of trapped cicadas was used for species- and sex-specific mass, nutrient, and stable isotope analyses. Five species were trapped during the study, of which three were dominant. Cicadetta calliope and Tibicen aurifera exhibited significantly higher emergence production in upland prairie than in lowland prairie, and were not captured in forested sites at all. T. dorsata emerged from all three habitat types, and though not significant, showed a trend of greater abundance in lowland grasslands. Two less abundant species, T. pruinosa and T. lyricen, emerged exclusively from forested habitats. Nitrogen fluxes associated with total cicada emergence were estimated to be ∼4 kg N ha^–1 year^–1 in both grassland habitats, and 1.01 kg N ha^–1 year^–1 in forested sites. Results of stable isotope analyses showed clear patterns of resource partitioning among dominant cicada species emerging from grassland sites. T. aurifera and C. calliope had δ¹³C and δ¹⁵N signatures indicative of feeding on shallowly rooted C₄ plants such as the warm-season grasses dominant in tallgrass prairie ecosystems, whereas T. dorsata signatures suggested preferential feeding on more deeply rooted C₃ plants. Received: 20 September 1999 / Accepted: 9 December 1999     

Dry matter and nutrient inputs through litter fall in a dry tropical forest of India

The present paper elucidates the pattern of leaf and non-leaf fall and quantifies of the total annual input of litter in a dry tropical forest of India. In addition, concentration of selected nutrients in various litter species and their annual return to the forest floor are examined. Total annual input of litter measured in litter traps ranged between 488.0–671.0 g m^-2 of which 65–72% was leaf litter fall and 28–35% wood litter fall. 73–81% leaves fall during the winter season. Herbaceous litter fall ranged between 80.0–110.0 g m^-2 yr^-1. The annual nutrient return through litter fall amounted (kg ha^-1): 51.6–69.6 N, 3.1–4.3 P, 31.0–40.0 Ca, 14.0–19.0 K and 3.7–5.0 Na, of which 71–77% and 23–29% were contributed by leaf and wood litter fall, respectively for different nutrients. Input of nutrients through herbaceous litter was: 13.0–16.6 for N, 1.0–1.4 for P, 4.0–5.0 for Ca, 7.9–10.5 for K and 0.8–1.0 kg ha^-1 yr^-1 for Na.     

Miscanthus × giganteus leaf senescence,decomposition and C and N inputs to soil

Energy crops are currently promoted as potential sources of alternative energy that can help mitigate the climate change caused by greenhouse gases (GHGs). The perennial crop Miscanthus × giganteus is considered promising due to its high potential for biomass production under conditions of low input. However, to assess its potential for GHG mitigation, a better quantification of the crop's contribution to soil organic matter recycling under various management systems is needed. The aim of this work was to study the effect of abscised leaves on carbon (C) and nitrogen (N) recycling in a Miscanthus plantation. The dynamics of senescent leaf fall, the rate of leaf decomposition (using a litter bag approach) and the leaf accumulation at the soil surface were tracked over two 1‐year periods under field conditions in Northern France. The fallen leaves represented an average yearly input of 1.40 Mg C ha^?1 and 16 kg N ha^?1. The abscised leaves lost approximately 54% of their initial mass in 1 year due to decomposition; the remaining mass, accumulated as a mulch layer at the soil surface, was equivalent to 7 Mg dry matter (DM) ha^?1 5 years after planting. Based on the estimated annual leaf‐C recycling rate and a stabilization rate of 35% of the added C, the annual contribution of the senescent leaves to the soil C was estimated to be approximately 0.50 Mg C ha^?1yr^?1 or 10 Mg C ha^?1 total over the 20‐year lifespan of a Miscanthus crop. This finding suggested that for Miscanthus, the abscised leaves contribute more to the soil C accumulation than do the rhizomes or roots. In contrast, the recycling of the leaf N to the soil was less than for the other N fluxes, particularly for those involving the transfer of N from the tops of the plant to the rhizome.     

The role of substrate feeding earthworms (Lumbricidae) for bioturbation in a beechwood soil

Summary Burrow formation (burrow length and faeces production) of the substrate feeding earthworm Aporrectodea caliginosa (Savigny) in relation to three temperatures (5, 10 and 15°C) and soil moisture contents (48, 60 and 73% water of dry wt) was determined. Soil moisture content affected the burrowing activity of A. caliginosa only below a distinct threshold (60% water of dry wt). At sufficient moisture an, increase in temperature by 5° C approximately doubled the amount of egesta produced. Another substrate feeding earthworm (Octolasion lacteum (Örley)) showed a similar pattern of burrowing activity.The data for the dependence of the faeces production of A. caliginosa on soil temperature and moisture content were combined with field data. The amount of faeces produced by the population of this earthworm species in a beechwood on limestone was calculated to be about 4.23 kg ha^-1 a^-1. The amount of egesta produced by all substrate feeding species is assumed to exceed 6 kg ha^-1 a^-1, which is equivalent to a soil layer of approximately 9 mm. This group of earthworms is therefore considered to be of greater importance for bioturbation and the formation of mull than previously assumed.     

Fungal biomass and decomposition in Spartina maritima leaves in the Mondego salt marsh (Portugal)

Spartina maritima (Curtis) Fernald is a dominant species in the Mondego salt marsh on the western coast of Portugal, and it plays a significant role in estuarine productivity. In this work, leaf litter production dynamics and fungal importance for leaf decomposition processes in Spartina maritima were studied. Leaf fall was highly seasonal, being significantly higher during dry months. It ranged from 42 g m^-2 in June to less than 6 g m^-2 during the winter. Fungal biomass, measured as ergosterol content, did not differ significantly between standing-decaying leaves and naturally detached leaves. Fungal biomass increased in wet months, with a maximum of 614 g g^-1 of ergosterol in January in standing-decaying leaves, and 1077 g g^-1 in December, in naturally detached leaves, decreasing greatly in summer. Seasonal pattern of fungal colonization was similar in leaves placed in litterbags on the marsh-sediment surface. However, ergosterol concentrations associated with standing-decaying and naturally detached leaves were always much higher than in litterbagged leaves, suggesting that fungal activity was more important before leaf fall. Dry mass of litterbagged leaves declined rapidly after 1 month (about 50%), mostly due to leaching of soluble organic compounds. After 13 months, Spartina leaves had lost 88% of their original dry weight. The decomposition rate constant (k) for Spartina maritima leaves was 0.151 month^-1.     

Effects of nitrogen supply on the anatomy and chemical composition of leaves of four grass species belonging to the genus Poa, as determined by image-processing analysis and pyrolysis–mass spectrometry

Previous experiments have shown that the anatomy and chemical composition of leaves of inherently fast- and slow-growing grass species, grown at non-limiting nitrogen supply, differ systematically. The present experiment was carried out to investigate whether these differences persist when the plants are grown at an intermediate or a very low nitrogen supply. To this end, the inherently fast-growing Poa annua L. and Poa trivialis L., and the inherently slow-growing Poa compressa L. and Poa pratensis (L.) Schreb. were grown hydroponically at three levels of nitrate supply: at optimum (RGR_max) and at relative addition rates of 100 and 50 mmol N (mol N)^?1 d^?1 (RAR₁₀₀ and RAR₅₀), respectively. As expected, at the lowest N supply, the potentially fast-growing species grew at the same rate as the inherently slow-growing ones. Similarly, the differences in leaf area ratio (LAR, leaf area:total dry mass), specific leaf area (SLA, leaf arear:leaf dry mass) and leaf mass ratio (LMR, leaf dry mass:total dry mass) disappeared. Under optimal conditions, the fast-growing species differed from the slow-growing ones in that they had a higher N concentration. There were no significant differences in C concentration. With decreasing N supply, the total N concentration decreased and the differences between the species disappeared. The total C concentration increased for the fast-growing species and decreased for the slow-growing ones, i.e. the small, but insignificant, difference in C concentration between the species at RGR_max increased with decreasing N supply. The chemical composition of the leaves at low N supply, analysed in more detail by pyrolysis–mass spectrometry, showed an increase in the relative amounts of guaiacyl lignin, cellulose and hemicellulose, whereas those of syringyl lignin and protein decreased. The anatomy and morphology of the leaves of the four grass species differing in RGR_max were analysed by image-processing analysis. The proportion of the total volume occupied by mesophyll plus intercellular spaces and epidermis did not correlate with the amount of leaf mass per unit leaf area (specific leaf mass, SLM) at different N supply. The higher SLM at low N supply was caused partly by a high proportion of non-veinal sclerenchymatic cells per cross-section and partly by the smaller volume of epidermal cells. We conclude that the decrease in relative growth rate (and increase in SLM) at decreasing N supply is partly due to chemical and anatomical changes. The differences between the fast- and slow-growing grass species at an optimum nutrient supply diminished when plants were growing at a limiting nitrogen supply.