Chamaegigas intrepidus Dinter (syn. Lindernia intrepidus (Dinter) Oberm.) is a poikilohydric aquatic plant that lives in rock pools on granitic outcrops in Central Namibia. The pools are only filled intermittently during the summer rains, and the plants can pass through 15–20 rehydration/dehydration cycles during a single wet season. Rehydrated plants also have to cope with substantial diurnal fluctuations in the pool pH as a result of photosynthetic CO2 uptake. We have used in vivo31P NMR spectroscopy to investigate the effect of external pH and dehydration (low water potential) on intracellular pH in the roots and submerged leaves of C. intrepidus. Increasing the external pH from 6 to 10 had no effect on the steady state cytoplasmic and vacuolar pH values of submerged leaves, but caused a slight alkalinization of the root cytoplasm. Similarly dehydration with PEG-600 at either pH 6 or pH 10 had no effect on the cytoplasmic pH of the leaves, but it did cause a small alkalinization of the leaf vacuoles at pH 10. These results imply an unusually effective regulation of intracellular pH, consistent with the adaptation of C. intrepidus to the extreme environmental conditions of its habitat. The NMR analysis also showed that dehydration had no effect on the inorganic phosphate and phosphocholine pools, and this was taken to indicate that the cell membranes were well protected from the effects of the low water potential.

New Phytologist welcomes this month, and introduces, its new Managing Editor, Jonathan Ingram. He comes to New Phytologist from Elsevier Science's Trends in Plant Science, where he has worked as Assistant Editor since the highly successful launch of that magazine in January 1996.

Jonathan's introduction to plant science, at Oxford over 10 years ago, was through the excellent tuition of Vernon Butt in a traditional botany degree. He then took a DPhil with Andrew Smith and Chris Leaver, learning biochemical and molecular research methods while studying malate decarboxylation associated with Crassulacean acid metabolism. He subsequently moved to the Max Planck Institute for Breeding Research in Cologne, Germany, where he did research with Dorothea Bartels into the molecular mechanism of drought tolerance in the resurrection plant Craterostigma plantagineum, specifically the role of sucrose-phosphate synthase.

These are challenging times in science publishing, particularly as technological advances force rapid change, but there are also many exciting new opportunities. Jonathan's experience and background will help New Phytologist remain in the vanguard of plant journals, while maintaining its traditions of scientific excellence and friendly service to authors and readers, traditions carefully nurtured by his predecessor David Stribley.

We wish David improved health and a happy retirement.

The range of processes regulated by gibberellins (GAs) covers all aspects of the life history of the plant from seed germination to vegetative growth and flowering. In seeds there has been an intensive search, using the techniques of both biochemistry and cell biology, for the regulatory molecules linking GA perception to gene regulation and the events of germination. Although a GA receptor has yet to be identified, the site of perception has been localized to the plasma membrane. Calmodulin, Ca2+ and cGMP have also been identified as elements of the GA signal transduction pathway. These regulators parallel many of the signalling elements identified in the transduction of other signals such as phytochrome and ABA. Studies of GA-regulated gene expression, principally of the α-amylases of cereal aleurone, have identified core GA-responsive promoter elements, such as the gibberellin response element (GARE), box-1 and pyrimidine boxes, as well as elements that may lend specificity to GA-regulated expression, such as the Opaque-2-similar element (O2S), and TRE and CRE motifs. One of the most striking features of all of these studies of the molecular basis of GA action is the interaction of GA-dependent regulatory elements with those of other factors such as ABA. GA-response elements also appear to be conserved between disparate GA-response systems. For example, Myb transcription factors appear to regulate a multitude of GA-induced genes in cereal aleurone as well as to alter GA responses when expressed in Arabidopsis. Thus the study of GA signal transduction and response systems is highlighting the conservation of regulatory elements used by plants. These common factors, used by distinct signal transduction systems, provide a molecular basis for the integration of the GA signal with other growth regulators that is the hallmark of plant growth and development.

Summary 599

I. Introduction 599

II. Concepts and terminology 600

III. Historical background 600

IV. Studies of experimental hybrids 601

1. Isolating mechanisms 601

2. Prezygotic barriers 602

(a) Gametic barriers to hybridization 602

3. Postzygotic barriers 603

(a) Chromosomal rearrangements 604

(b) Genic sterility or inviability 604

4. Hybrid vigour 605

5. Introgression 606

6. Hybrid speciation 607

V. Experimental manipulations of natural hybrid populations 609

1. Hybrid-zone formation 610

2. Pollinator-mediated selection 610

3. Habitat selection 612

VI. The biology of different classes of hybrids 612

1. Character expression 613

(a) Morphological characters 613

(b) Chemical characters 613

(c) Molecular characters 613

2. The fitness of different classes of hybrids 614

(a) The importance of variance 614

(b) Estimating hybrid fitness 615

3. Interactions with parasites and herbivores 616

4. Patterns of mating 617

(a) Outcrossing rate 617

(b) Hybridization frequency 618

(c) Mate choice 618

VII. Conclusions and future research 619

Acknowledgements 620

References 620

Most studies of plant hybridization are concerned with documenting its occurrence in different plant groups. Although these descriptive, historical studies are important, the majority of recent advances in our understanding of the process of hybridization are derived from a growing body of experimental microevolutionary studies. Analyses of artificially synthesized hybrids in the laboratory or glasshouse have demonstrated the importance of gametic selection as a prezygotic isolating barrier; the complex genetic basis of hybrid sterility, inviability and breakdown; and the critical role of fertility selection in hybrid speciation. Experimental manipulations of natural hybrid zones have provided critical information that cannot be obtained in the glasshouse, such as the evolutionary conditions under which hybrid zones are formed and the effects of habitat and pollinator-mediated selection on hybrid-zone structure and dynamics. Experimental studies also have contributed to a better understanding of the biology of different classes of hybrids. Analyses of morphological character expression, for example, have revealed transgressive segregation in the majority of later-generation hybrids. Other studies have documented a high degree of variability in fitness among different hybrid genotypes and the rapid response of such fitness to selection – evidence that hybridization need not be an evolutionary dead end. However, a full accounting of the role of hybridization in adaptive evolution and speciation will probably require the integration of experimental and historical approaches.

Littorella uniflora (L.) Ascherson is a small, perennial, amphibious rhizophyte of rosette life-form which is common along the margins of lakes, tarns and reservoirs where water-level fluctuations are often rapid and unpredictable. The majority of plants are continuously submersed and reproduce vegetatively, but a small proportion become completely emersed for variable lengths of time, when flowering and seed set occur. To find out how L. uniflora adjusts to sudden emersion we studied the plants at a reservoir where water level falls each spring and remains low throughout the summer; L. uniflora adjusted very quickly showing a degree of phenotypic plasticity not expected in a ‘stress tolerator’, including the production of a new set of terrestrial leaves with reduced lacunal volume and increased stomatal density, a rapid increase in leaf growth rate, and flowering within 3–4 wk. Comparison of terrestrial L. uniflora with aquatic plants growing permanently submersed in lake and tarn habitats showed that three to fourfold more carbon (C) and nitrogen (N) was incorporated into above-ground biomass by emersed plants. However, ramet production in the aquatic environment appeared to be more costly, in terms of C and N invested, than terrestrial flower and seed production. The combination of continuous, submersed vegetative spread with the capacity for a high degree of phenotypic plasticity allowing some flower and seed production to occur during brief periods of emersion seems to account for the success of this plant in the amphibious niche.

Following a precultivation with pedospheric nitrogen nutrition, Ricinus plants were supplied with nitrogen solely by spraying nitrate or ammonium solution onto the leaves during the experimental period. The chemical composition of tissues, xylem and phloem exudates was determined and on the basis of the previously determined nitrogen flows (Peuke et al., New Phytologist (1998), 138, 657–687) the flows of potassium, sodium, magnesium, calcium, chloride and ABA were modelled. These data, which permit quantification of net-uptake, transport in xylem and phloem, and utilization in shoot and root, were compared with results obtained in plants with pedospherically-supplied nitrate or ammonium and data in the literature. Although the overall effects on the chemical composition of supplying ammonium to the leaves were not as pronounced as in pedospherically supplied plants, there were some typical responses of plants fed with ammonium (ammonium syndrome). In particular, in ammonium-sprayed plants uptake and transport of magnesium decreased and chloride uptake was increased compared with nitrate-sprayed plants. Furthermore, acropetal ABA transport in the xylem in ammonium-sprayed Ricinus was threefold higher than in nitrate-sprayed plants. Additionally, concentrations of anions were more or less increased in tissues, particularly in the roots, and transport fluids. The overall signal from ammonium-sprayed leaves without a direct effect of ammonium ions on uptake and transport systems in the root is discussed.

This review considers the reasons for, and research governing, the regulation and monitoring of genetically engineered micro-organisms and viruses (GEMs) released into the environment. The hazards associated with releasing GEMs into the environment are the creation and evolution of new pests and diseases, and damage to the ecosystem and non target species. The similarities and differences between GEMs and conventional micro-organisms are discussed in relation to risk assessment. Other issues covered include the persistence of micro-organisms in the environment, transgene dispersal to non-engineered microbes and other organisms, the effects of transgenes and transformation on fitness, and the evolution of pests and pathogens that are given or acquire transgenes. Areas requiring further research are identified and recommendations for risk assessment made.

One of the most complex and contentious issues in Australian ecology concerns the environmental impact of Aboriginal landscape burning. This issue is not only important for the development of a comprehensive understanding of the dynamics and evolution of the Australian biota, but is central to the formulation of appropriate strategies for the conservation of the nation's biodiversity. Ethnographic evidence leaves little doubt that Aboriginal burning played a central role in the maintenance of the landscapes subsequently colonized by Europeans. Both 19th century European colonists and anthropologists in the 20th century documented the indispensability of fire as a tool in traditional Aboriginal economies, which have aptly been described as ‘fire-stick farming’. Aborigines used fire to achieve short-term outcomes such as providing favourable habitats for herbivores or increasing the local abundance of food plants, but it is not clear whether or not Aborigines had a predictive ecological knowledge of the long-term consequences of their use of fire. A large body of ecological evidence suggests that Aboriginal burning resulted in substantial changes in the geographic range and demographic structure of many vegetation types. Aboriginal burning was important in creating habitat mosaics that favoured the abundance of some mammal species and in the maintenance of infrequently burnt habitats upon which the survival of specialized fauna depends. Aboriginal fire regimes were probably critical for the maintenance of at least one species of tree (Callitris intratropica) in the monsoon tropics.

The question of the original impact of humans on the Australian environment is fundamentally speculative because of vague, disputed time frames proposed for the waves of colonization and shifting settlement patterns of Aborigines in the late Quaternary period. There is an inherent circular argument concerning the cause and effect of climate change, vegetation change, and burning through the late Quaternary. Charcoal and pollen evidence from long sedimentary cores is ambiguous and cannot be used to demonstrate unequivocally the initial impact of Aboriginal people on the landscapes of Pleistocene Australia. The sparse available evidence does not support the hypotheses that Aboriginal burning was primarily responsible for the extinction of Pleistocene megafauna; was critical for the maintenance of habitats of small mammals that have become extinct following European colonization; initiated widespread accelerated soil erosion rates in either the Pleistocene or Holocene; or forced the evolutionary diversification of the Australian biota. Burning may have caused the extinction of some fire-sensitive species of plants and animals dependent upon infrequently burnt habitats, and it must have maintained structurally open vegetation such as grasslands and also extended the range of fire-adapted species, such as Eucalyptus, into environments climatically suitable for rain forest. Palaeoecological research concerning prior impacts of Aborigines must give way to focused studies of the role of different anthropogenic fire regimes in contemporary ecosystems that have not been destroyed by European colonization. Such research is crucial for comprehending the role of Aboriginal burning in the maintenance of Australia's unique, rich biodiversity.

Circumstantial evidence suggests that plants that have evolved metal tolerance are at a disadvantage on normal soil, i.e. there is a cost of tolerance. One hypothesis for the cause of this cost is that individuals have a greater requirement for copper, and so suffer micronutrient deficiency on normal soils, as a result of a reduced uptake, distribution and/or utilization of copper. We provided highly and less copper-tolerant plants of Mimulus guttatus Fischer ex DC. (the common monkey flower) with sub-optimal copper, and demonstrated the importance of copper as an essential micronutrient during the reproductive phase, both in the production of viable pollen and in seed set. We also looked at the effect of sub-optimal copper supply on the growth of the microgametophyte, and the efficiency with which seed was set. No evidence was found that highly tolerant plants have an increased copper requirement during the reproductive phase. This is in agreement with earlier work on Mimulus guttatus, which investigated the copper requirement of highly tolerant plants during vegetative growth and found that any differences in copper requirement were small. The ‘metal requirement hypothesis’ is, therefore, not the sole explanation for the cost of copper tolerance in M. guttatus.

Responses of leaf gas exchange and above-ground growth of beech (Fagus sylvatica L.) and Norway spruce (Picea abies Karst.) to atmospheric CO2 enrichment (374 μl l−1 vs. 590 μl l−1) and increased wet deposition of N (5 vs. 50 kg N ha−1 a−1) in combination with two natural forest soil types (‘acidic’ and ‘calcareous’) were studied in large open-top chambers. Eight juvenile beech and spruce trees from different provenances, together with a ground cover composed of five understorey species, were established in each of 32 model ecosystems. Both beech and spruce showed sustained enhancement of photosynthesis in response to atmospheric CO2 enrichment during the first 2 yr of treatment. Nevertheless, switching measurement CO2 concentrations revealed partial downward adjustment of photosynthesis in trees grown in elevated CO2, beech generally showing more pronounced downward adjustment than spruce. The responsiveness of photosynthesis to CO2 enrichment did not vary significantly among trees from different provenances. Stomatal conductance was reduced under elevated CO2 in both tree species. In spruce, the radial growth of the main stem and the annual production of wood (shoot-wood dry mass of current-year lateral shoots), needle dry mass, and assimilation area per tree were stimulated both by CO2 enrichment and increased N deposition, but were not significantly affected by soil type by year 2. In contrast, in beech, the radial growth of the stem and the total leaf number, foliage dry mass, and assimilation area per tree were all not significantly affected by elevated CO2 and increased N deposition when responses of the two soil types were pooled, but were greater on calcareous than on acidic soil by year 2. However, CO2 interacted with soil type in beech: irrespective of the N deposition rate, saplings showed growth stimulation on the calcareous soil but responded negatively to CO2 enrichment on the acidic soil (where growth was slower). Our results suggest that complex interactions between CO2, species and soil quality need to be accounted for when attempting to predict forest development in a future CO2-rich world.

Quasi-steady-state measurements of root hydraulic conductance (KR) of Olea oleaster Hoffmgg. et Link potted seedlings were performed using a pressure chamber with the aim of: (a) measuring the impact of different water-stress levels on a KR; (b) measuring the kinetics of KR recovery several days after soil rewetting; (c) relating changes in KR to changes in root anatomy and morphology. Increasing water-stress was applied in terms of ratio of leaf water potential (ΨL) measured at midday to that at zero turgor (ΨTLP), i.e. ΨL/ΨTLP=0·5, 1·0, 1·2, 1·6; KR was measured initially and at 24, 48, 72, 96 h after irrigation.

Values of KR in seedlings stressed to ΨL/ΨTLP=1·2 increased for 48 h after irrigation from 0·23 to 0·97×10−5 kg s−1 m−2 MPa−1 i.e. from 16% to 66% of that measured in unstressed seedlings. A marked shift of the x-axis intercept of the straight line relating flow to pressure (zero flow at non-zero pressure) was recorded initially after irrigation and persisted up to 48 h. Recovery of KR occurred within 24 h after irrigation in seedlings at ΨL/ΨTLP=0·5 and 48 h later in those at ΨL/ΨTLP=1·0.

Severe drought stress (ΨL/ΨTLP=1·6) caused anatomical changes to roots which formed a two-layered exodermis with thicker suberized walls and a three- to four-layered endodermis with completely suberized tangential walls. Recovery of KR in these roots required resumed growth of root tips and emergence of new lateral roots.

The aim of the present study was to investigate the effects of elevated CO2 on the antioxidative systems and the contents of pigments, soluble protein and lipid peroxidation in leaves of adult oaks, Quercus pubescens and Quercus ilex, grown at naturally enriched CO2 concentrations. For this purpose, a field study was conducted at two CO2 springs in Central Italy. Measurements of the pre-dawn water potentials indicated less drought stress in trees close to CO2 springs than in those grown at ambient CO2 concentrations. Most leaf constituents investigated showed significant variability between sampling dates, species and sites. The foliar contents of protein and chlorophylls were not affected in trees grown close to the CO2 vents compared with those in ambient conditions. Increases in glutathione and other soluble thiols were observed, but these responses might have been caused by a low pollution of the vents with sulphurous gases. At CO2 vents, glutathione reductase was unaffected, and superoxide dismutase activity was significantly diminished, in both species. Generally, the activities of catalase, guaiacol peroxidase and ascorbate peroxidase as well as the sum of dehydroascorbate and ascorbate were decreased in leaves from trees grown in naturally CO2-enriched environments compared with those grown at ambient CO2 concentrations. The reduction in protective enzymes did not result in increased lipid peroxidation, but increased monodehydroascorbate radical reductase and dehydroascorbate reductase activities found in leaves of Q. pubescens suggest that the smaller pool of ascorbate was subjected to higher turnover rates. These data show that changes in leaf physiology persist, even after lifetime exposure to enhanced atmospheric CO2. The results suggest that the down-regulation of protective systems, which has also previously been found in young trees or seedlings under controlled exposure to elevated CO2 concentrations, might reflect a realistic response of antioxidative defences in mature trees in a future high-CO2 world.

Maternal-environmental effects on subsequent progeny life-history traits were evaluated in squash (Cucurbita pepo L.) in terms of the amount of time available for seed development, and the timing of fruit production. Progeny arising from three kinds of fruit were compared. Plants from which fruits were removed 3 d post-pollination throughout the growing season developed only ‘late’ fruits (during 10–15 d) at the end of the growing season; on control plants both ‘early’ and ‘late’ fruits developed (both types allowed to ripen fully). Seed from each type of fruit was weighed individually and categorized into three size classes, then germinated and raised to maturity, including regular harvesting of all fruits 3 d post-pollination. Maternal effects were evident for both vegetative and reproductive traits and carried over to later stages. In contrast, effects due simply to seed size disappeared by day 30 for leaf variables and day 60 for male flower production. Within a seed-size class, progeny arising from fruit of treated plants produced significantly more leaves, with greater size, and more male flowers than those arising from fruit of control plants, while the reverse was true for fruit number and fruit mass. This result is discussed in terms of possible gibberellic acid involvement. In control plants, progeny arising from seeds in the large, fully mature ‘early’ fruits produced significantly more, and larger leaves by day 30 than did those from late fruits (suggesting differential provisioning in seeds during development). Male flower production had a highly significant positive correlation with vegetative mass and a significant negative correlation with fruit production.

Arbuscular mycorrhizal (AM) fungi extensively colonize the root cortex under low-soil-phosphate (P) conditions, whereas infection is limited under high-P conditions. Fungal growth under both P conditions might be influenced by plant defence-related gene expression. In this study, we used in situ hybridization methods to compare the cellular localization of three defence-related mRNAs in non-infected bean roots and in relation to fungal infection units. In non-infected and infected roots, mRNAs encoding acidic and basic endochitinases were generally most abundant in the vascular cylinder. High-P-grown mycorrhizal roots showed localized accumulation of the acidic endochitinase mRNA in cortical cells containing arbuscules and in their immediate vicinity (one to five cell layers). The pattern of accumulation of the basic endochitinase mRNA was not affected by P or AM fungal infection. At the low P concentration, the β-1,3-glucanase mRNA accumulated predominantly in the vascular cylinder of non-infected roots. Suppression of β-1,3-glucanase mRNA accumulation in these tissues was observed in non-infected roots at the high-P and in mycorrhizal roots at both P concentrations. The observed suppression extends at least several mm from fungal infection units, characterizing a systemic effect. Beta-1,3-glucanase mRNA accumulated also around a number of cortical cells containing arbuscules only at the low P concentration. The localized accumulations of the endochitinase and β-1,3-glucanase mRNAs suggest that the encoded proteins might be involved in the control of intraradical fungal growth.

This paper analyses relationships between relative growth rate (rgr), seed mass, biomass allocation, photosynthetic rate and other plant traits as well as habitat factors (rainfall and altitude) in 20 wild species of Aegilops L. and one closely related species of Amblyopyrum (Jaub. & Spach) Eig., which differ in ploidy level (diploid, tetraploid and hexaploid). The plants were grown hydroponically for 20 d in a growth chamber. The relationships between parameters were calculated either using the phylogenetic information (phylogenetically independent contrasts, PIC) or without using the phylogenetic information (trait values of taxa, TIP). The results using the two approaches were very similar, but there were a few exceptions in which the results were different (e.g. rgr vs. seed mass). Specific leaf area (sla) was positively correlated with leaf area ratio (lar) and negatively correlated with net assimilation rate (nar), which together resulted in the absence of a correlation between sla and rgr. Leaf photosynthetic rates (expressed on a mass or area basis) showed no correlation with rgr. rgr was positively correlated with the stem mass ratio and negatively with root mass ratio. Species with a lower d. wt percentage have a higher rgr. Aegilops species from locations with higher annual rainfall invested less biomass in roots and more in shoots (leaves and stems) and had a higher rgr. Diploid species had a lower seed mass and initial mass than the hybrids (tetraploid and hexaploid species), but there was no correlation of rgr with ploidy level. Polyploid species, which have higher seed mass, occur at a higher altitude than diploid species. Our results show that variation in rgr in Aegilops and Amblyopyrum spp. is associated mainly with variation in biomass allocation (proportion of biomass in stems and roots) and d. wt percentage, and not with variation in sla, leaf photosynthetic rates or seed mass.

We tested the extent to which growth responses to elevated carbon dioxide (CO2) are temperature-dependent and change through early seedling ontogeny among boreal tree species of contrasting relative growth rates (rgr). Populus tremuloides Michx, Betula papyrifera Marsh, Larix laricina (Du Roi) K. Koch, Pinus banksiana Lamb., and Picea mariana (Mill.) B.S.P. were grown from seeds for 3 months in controlled-environment chambers at two CO2 concentrations (370 and 580 μmol mol−1) and five temperature regimes of 18/12, 21/15, 24/18, 27/21 and 30/24°C (light/dark). Growth increases in response to CO2 enrichment were minimal at the lowest temperature and maximal at 21/15°C for the three conifers and at 24/18°C or higher for the two broadleaved species, corresponding with differences in optimal temperatures for growth. In both CO2 treatments, rgr among species and temperatures correlated positively with leaf area ratio (lar) (r[ges ]0·90, P<0·0001). However, at a given lar, rgr was higher in elevated CO2, owing to enhanced whole-plant net assimilation rate. On average in all species and temperatures at a common plant mass, CO2 enrichment increased rgr (9%) through higher whole-plant net assimilation rate (22%), despite declines in lar in high CO2 (11%). Reductions in lar are thus an important feedback mechanism reducing positive plant growth responses to CO2. Proportional allocation of dry mass to roots did not vary between CO2 treatments. Early in the experiment, proportional increases in plant dry mass in elevated CO2 were larger in faster-growing Populus tremuloides and B. papyrifera than in the slower-growing conifers. However, growth increases in response to CO2 enrichment fell with time for broadleaved species and increased for the conifers. With increasing plant size over time, compensatory adjustments to CO2 enrichment in the factors that determine rgr, such as lar, were much larger in broadleaves than in conifers. Thus, the hypothesis that faster-growing species are more responsive to elevated CO2 was not supported, given contrasting patterns of growth response to CO2 with increasing plant size and age.

Three stands of Phragmites australis (Cav.) Trin. ex Steudel were investigated regarding the relationship between the number of efflux culms and convective ventilation efficiency affecting the hypoxic status of roots and rhizomes. The lack of old (efflux) culms after mowing the preceding winter caused a significantly higher counterpressure within the rhizome, thereby diminishing air flushing rate, i.e. oxygen supply, of rhizomes. The levels of alanine and c-aminobutyric acid in basal culm internodes increased significantly. Both amino acids indicate the hypoxic status of the root and rhizome metabolism of P. australis. Amino acid patterns of the basal culm internodes are discussed with respect to the maintenance of aerobic root metabolism and nutrient availability.

Swards of Dactylis glomerata cultivars (cvs) KM2 and Lutetia and of Lolium perenne cvs Aurora and Vigor were grown under full irrigation or prolonged summer drought (80 d) in a field experiment in the South of France.

After irrigation was withheld, leaf extension rates of all cvs fell by 90% within 9–12 d, and rapid scorching of laminae followed. Tiller mortality at the end of the drought was very different in the cocksfoot cvs (4% for KM2 and 76% for Lutetia) and intermediate (41%) for both ryegrass cvs. Following re-watering, rates of herbage regrowth were closely correlated with tiller survival. Measured minerals contributed c. 0·52 MPa to osmotic potential in all treatments, whereas water-soluble carbohydrates (WSC) contributed 0·25 MPa under irrigation and 0·46 MPa during drought.

There was no systematic difference between the two species for summer survival under severe drought, but large differences between the cocksfoot cvs. The traits most strongly associated with superior survival were: (a) a deep root system and greater water uptake at depth; (b) low water and osmotic potentials in surviving laminae, i.e. better tolerance to dehydration; (c) large pool-size of WSC reserves (fructans having degree of polymerization >4) in entire tiller bases (stubble); (d) low accumulation of proline in stubble; (e) rapid nitrogen uptake after rewatering.

Eragrostis pilosa (Linn.) P Beauv., a C4 grass native to east Africa, was grown at both ambient (350 μmol mol−1 and elevated (700 μmol mol−1) CO2 in either the presence or absence of the obligate, root hemi-parasite Striga hermonthica (Del.) Benth. Biomass of infected grasses was only 50% that of uninfected grasses at both CO2 concentrations, with stems and reproductive tissues of infected plants being most severely affected. By contrast, CO2 concentration had no effect on growth of E. pilosa, although rates of photosynthesis were enhanced by 30–40% at elevated CO2. Infection with S. hermonthica did not affect either rates of photosynthesis or leaf areas of E. pilosa, but did bring about an increase in root[ratio ]shoot ratio, leaf nitrogen and phosphorus concentration and a decline in leaf starch concentration at both ambient and elevated CO2. Striga hermonthica had higher rates of photosynthesis and shoot concentrations of soluble sugars at elevated CO2, but there was no difference in biomass relative to ambient grown plants. Both infection and growth at elevated CO2 resulted in an increase in the Δ13C value of leaf tissue of E. pilosa, with the CO2 effect being greater. The proportion of host-derived carbon in parasite tissue, as determined from δ13C values, was 27% and 39% in ambient and elevated CO2 grown plants, respectively. In conclusion, infection with S. hermonthica limited growth of E. pilosa, and this limitation was not removed or alleviated by growing the association at elevated CO2.

Calluna vulgaris L. (Hull) is not one of the usual hosts of the winter moth, Operophtera brumata L., but outbreaks have caused extensive damage to heather moorland in Scotland in recent years. This study investigated the potential role of environmental change in such outbreaks by rearing O. brumata larvae on C. vulgaris plants grown in open-top chambers for 20 months with enriched CO2 (600 ppm) and nitrogen supply (average 52·5 kg N ha−1 yr−1) in factorial combination. This prolonged exposure to elevated CO2 caused no change in shoot growth, photosynthesis or foliar C[ratio ]N ratio of C. vulgaris, even with increased N supply, indicating that the absence of response was not due to N limitation. Increased N supply itself resulted in increased shoot growth and a decrease in tissue C[ratio ]N ratio. Phenolic content did not change in response to either CO2 or N enrichment, contrary to the predictions of the carbon/nutrient balance hypothesis. In line with the absence of plant response, there was no effect of CO2 on the development of Operophtera brumata on C. vulgaris, and so continued increase in atmospheric CO2 concentration is unlikely to affect directly O. brumata outbreaks on heather moorland. Operophtera brumata showed increased larval development, growth rate and pupal weight on N-treated plants, correlated both to the decrease in foliar C[ratio ]N ratio, and to the increase in shoot extension which was predictive of survivorship. Thus, increased atmospheric N deposition, or increased rates of mineralization in a warmer environment, might increase the severity of O. brumata outbreaks on C. vulgaris. Since the combination of high N availability and disturbance of heather canopy by herbivory is known to result in increased dominance of grasses, it is suggested that this could lead to further degradation of moorland in upland Britain.