Growth, tolerance and zinc and cadmium hyperaccumulation of Thlaspi caerulescens populations from three metal contaminated soils and three normal soils were compared under controlled conditions. Individuals of six populations were cultivated on five soils with increasing concentrations of zinc (50–25000 μg g−1) and cadmium (1–170 μg g−1). There was no mortality of normal soil populations in the four metal-contaminated soils, but plant growth was reduced to half that of populations from metal-contaminated soils. However, in noncontaminated soil, the growth of individuals from normal soils was greater than that of individuals from metal-contaminated soils. Individuals from normal soils concentrated three times more zinc in the aboveground biomass than those from metal-contaminated soils, but the latter accumulated twice as much cadmium. We conclude that populations of T. caerulescens from both normal and metal-contaminated soils are interesting material for phytoextraction of zinc and cadmium, but to optimize the process of phytoextraction it is necessary to combine the extraction potentials of both type of populations.

Interactive effects of elevated atmospheric CO2 and soil N availability on N2 fixation and biomass production were examined using Gliricidia sepium, a tropical leguminous tree species. Our objective was to determine if elevated CO2 alters the inhibitory effects of soil N on N2 fixation, and whether the response of Gliricidia to elevated CO2 was a function of N source originating from either substrate N fertilizer or N2 fixation. We hypothesized that CO2 enrichment would ameliorate the inhibitory effects of N fertilization on seedling nodulation and N2 fixation through increased C partitioning to nodules. Seedlings were grown from seed for 100 d in growth chambers at either 350 or 700 μmol mol−1 CO2. Seedlings were inoculated with Rhizobium spp. and grown either with 0, 1 or 10 mM N fertilizer. The δ15N isotope-dilution technique was used to determine N source partitioning between N2 fixation and inorganic N fertilizer uptake. The addition of 10 mM N fertilizer significantly reduced nodule number and mass, specific nitrogenase activity, the specific rate of N2 fixation, and the proportion of plant N derived from N2 fixation. Elevated CO2, however, strongly ameliorated the inhibitory effects of N fertilization, indicating that increased C availability for nodule activity may partially offset the inhibition of N2 fixation caused by substrate N, as nodule sugar concentrations were stimulated with CO2 enrichment. This study clearly shows that elevated CO2 enhanced plant productivity and net N content of Gliricidia tree seedlings by stimulating N2 fixation. In addition, seedling biomass production was greatly enhanced by elevated CO2, regardless of whether plant N was derived from the substrate or from the atmosphere. We conclude from this study that CO2 enrichment mitigates the inhibitory effects of substrate N on nodule initiation and development and specific N2 fixation, either through increased C allocation to nodule production and activity, or through increased N demand by the plant for biomass production. This experiment with Gliricidia provides evidence for a positive feedback between increased atmospheric CO2 concentrations, C allocation to symbiotic N2-fixing bacteria, and plant C and N accumulation that may occur when N2-fixing plants are grown under conditions where substrate N may typically inhibit N2 fixation.

The transcript (mRNA) level and the protein content (as determined by enzyme-linked immunosorbent assay) of thioredoxins (Trx) f and m, and of their targets, chloroplast fructose-1,6-bisphosphatase (FBPase) and NAD(P)- malate dehydrogenase (NADP-MDH), increase over the ontogeny of pea plants grown under normal conditions, showing their highest values before flowering (40 d growth). The clearest results appear in apical, but also in middle leaves. Enzyme activites of FBPase and NADP-MDH were lowest just before flowering. In the case of FBPase this was probably a mechanism to facilitate triose-phosphate export to the cytosol for sucrose synthesis. The likely function of NADP-MDH is to supply the cytosol, via the malate translocator, with the NAD(P)H necessary for UTP regeneration in the sucrose biosynthetic pathway. Both the Fv/Fm ratio and the net photosynthetic rate (IRGA) decreased at saturating irradiance (16 h photoperiod) and under sub-saturating continuous light. However, the Fv/Fm quotient recovered to normal values after several days adaptation to high light. A similar recovery was also observed in net photosynthesis, although normal levels were never obtained. Under light-stress conditions the concentration of Trxs f and m, and of the targets FBPase and NADP-MDH, were somewhat lower than those of unstressed plants. Even though the levels of the corresponding transcripts (mRNAs) are similar in upper leaves from control and light-stressed plants, those of the middle and basal leaves from plants grown under high light were substantially higher than those of the control plants. In addition to the well-documented transcriptional regulation of nuclear-coded chloroplast proteins, these results seem to indicate the existence of an additional post-transcriptional control.

The hydraulic conductance (L0) of detached, exuding root systems from melon (Cucumis melo cv. Amarillo oro) was measured. All plants received a half-strength Hoagland nutrient solution, and plants stressed either solely with NaCl (50 mM) or with NaCl (50 mM) following treatment (2 d) with CaCl2 (10 mM) were compared with controls and CaCl2-treated (10 mM) plants. The L0 of NaCl-treated plants was markedly decreased when compared to control and CaCl2-treated plants, but the decrease was smaller when NaCl was added to plants previously treated with CaCl2. A similar effect was observed when the flux of Ca2+ into the xylem and the Ca2+ concentration in the plasma membrane of the root cells were determined. In control, CaCl2- and NaCl + CaCl2-treated plants, HgCl2 treatment (50 μM) caused a sharp decline in L0 to values similar to those of NaCl-stressed roots, but L0 was restored by treatment with 5 mM DTT. However, in NaCl roots only a slight effect of Hg2+ and DTT were observed. The effect of all treatments on L0 was similar to that on osmotic water permeability (Pf) of individual protoplasts isolated from roots. The results suggest that NaCl decreased the passage of water through the membrane and roots by reducing the activity of Hg-sensitive water channels. The ameliorative effect of Ca2+ on NaCl stress could be related to water-channel function.

During the initial phases of elongation of pea internodes, oat and rice coleoptiles, oat mesocotyls, soybean hypocotyls and dandelion peduncles, net transverse orientation of cellulose wall microfibrils (Mfs) was found in the outer epidermal wall. This paper demonstrates that in all these axes, with the exception of rice coleoptile, net longitudinal orientation of microfibrils occurs in the outer epidermal wall in portions of the axes that were still elongating at the time of sampling. The timing of the transition to net longitudinal orientation and whether the transition proceeded acropetally or basipetally varied with the type of axis under study. The variability of the relationship between extension and the transition from net transverse to net longitudinal orientation suggests that factors other than extension are important in determining the transition. Layers of longitudinal wall microfibrils may be added to the extending epidermal wall to bolster its tensile strength commensurate with its function during and after extension. Attention is drawn to the parallels between the concept of tissue tension in growing axes and the concept that the epidermis functions as a stressed skin in the support of mature plant parts in primary growth.

The role of carbonic anhydrase in the carbon-concentrating-mechanism of bryophytes of the class Anthocerotae was investigated by comparing the gas-exchange characteristics of material which had been incubated in the membrane-permeable Carbonic Anhydrase inhibitor ethoxyzolamide, with those of untreated material and material which had been incubated in buffer solution. In Phaeoceros laevis (Anthocerotae), incubation in ethoxyzolamide caused a depression in the rate of gross assimilation and a decrease in CO2 affinity beyond that which could be attributed to increased diffusion limitation. A range of liverworts and mosses, in which a carbon-concentrating-mechanism is absent, were also investigated. These showed no depression of rates of gross assimilation after incubation in ethoxyzolamide relative to those of untreated material. The CO2 compensation point and CO2 uptake characteristics of Phaeoceros laevis were significantly affected by incubation in ethoxyzolamide. Values of CO2 compensation point for Phaeoceros laevis rose from 2.5 Pa, after incubation in buffer, to 20 Pa after incubation in ethoxyzolamide. The CO2 compensation point for the liverworts Pellia epiphylla and Marchantia polymorpha was not significantly affected by incubation in ethoxyzolamide. Measurements of the release of CO2 at the end of a short (15 min) period of illumination revealed that, after suppression of carbonic anhydrase activity, the rapid release of a CO2 pool occurred in Phaeoceros laevis but not in the liverworts. There were also significant differences between values for fractionation measured in units per mil (‰), measured instantaneously, for Phaeoceros laevis incubated in ethoxyzolamide, compared with fractionation values for this species after incubation in buffer. Incubation in ethoxyzolamide caused fractionation values to rise from 12.4–22.7‰, indicating that the carbon-concentrating-mechanism of this species had been inactivated. Incubation in ethoxyzolamide had no effect on fractionation values for the liverworts. The convexity of the light saturation curves of liverworts and Phaeoceros laevis was also investigated, but there were no differences between groups before or after the two treatments. The data indicate an important role for carbonic anhydrase in the functioning of the carbon-concentrating-mechanism in the Anthocerotae.

Root growth and physiological responses to elevated CO2 were investigated for three important Mojave Desert grasses: the C3 perennial Achnatherum hymenoides, the C4 perennial Pleuraphis rigida and the C3 annual Bromus madritensis ssp. rubens. Seeds of each species were grown at ambient (360 μl l−1) or elevated (1000 μl l−1) CO2 in a glasshouse and harvested at three phenological stages: vegetative, anthesis and seed fill. Because P. rigida did not flower during the course of this study, harvests for this species represent three vegetative stages. Primary productivity was increased in both C3 grasses in response to elevated CO2 (40 and 19% for A. hymenoides and B. rubens, respectively), but root biomass increased only in the C3 perennial grass. Neither above-ground nor below-ground biomass of the C4 perennial grass was significantly affected by the CO2 treatment. Elevated CO2 did not significantly affect root surface area for any species. Total plant nitrogen was also not statistically different between CO2 treatments for any species, indicating no enhanced uptake of N under elevated CO2. Physiological uptake capacities for NO3 and NH4 were not affected by the CO2 treatment during the second harvest; measurements were not made for the first harvest. However, at the third harvest uptake capacity was significantly decreased in response to elevated CO2 for at least one N form in each species. NO3 uptake rates were lower in A. hymenoides and P. rigida, and NH4 uptake rates were lower in B. rubens at elevated CO2. Nitrogen uptake on a whole root-system basis (NO3+NH4 uptake capacity × root biomass) was influenced positively by elevated CO2 only for A. hymenoides after anthesis. These results suggest that elevated CO2 may result in a competitive advantage for A. hymenoides relative to species that do not increase root-system N uptake capacity. Root respiration measurements normalized to 20 °C were not significantly affected by the CO2 treatment. However, specific root respiration was significantly correlated with either root C[ratio ]N ratio or root water content when all data per species were included within a simple regression model. The results of this study provide little evidence for up-regulation of root physiology in response to elevated CO2 and indicate that root biomass responses to CO2 are species-specific.

Ice-core records of the concentration of atmospheric CO2 and its stable isotope ratio (δ13Ca) have shown that the global C cycle has not remained in steady-state over the past 11000 yr, implying a possible change in vegetation activity over this period. Here we evaluated the ecophysiological responses of the dwarf willow (Salix herbacea) over the past 9000 yr by measuring the stable carbon isotope composition and stomatal characters of a unique, well dated, high-latitude (68 °N) sub-fossil leaf sequence. After correction for corresponding changes in δ13Ca, a 9000-yr record of variations in the ratio of intercellular (ci) to atmospheric (ca) CO2 concentration was established. Intercellular[ratio ]atmospheric CO2 concentration ratios provide a time-integrated indicator of the set-point of leaf gas exchange, and the historical variations revealed in this record have been interpreted as an impact of environmental changes on leaf gas exchange. The sequence shows a progressive fall in ci/ca 9000–3000 yr BP as well as the climatic effects of the Medieval Warm Period, the Little Ice Age and the post-industrial CO2 rise. Leaf stomatal index (proportion of epidermal cells as stomata), but not stomatal density, was significantly (P<0.01) correlated with Holocene atmospheric CO2 variations. A process-based interpretation of the changes in ci/ca was made using two different coupled photosynthesis-stomatal conductance models. Calculated in this way, S. herbacea photosynthetic rates were relatively stable throughout the Holocene whilst stomatal conductance progressively declined. Both, however, showed the marked effects of the Medieval Warm Period and the Little Ice Age. Overall, the results demonstrate that S. herbacea leaf metabolism, like the global C cycle, has not remained in steady state during the Holocene but has responded to changes in atmospheric CO2 concentration and short-term climatic oscillations.

Photosynthetic rates of plants grown in natural systems exhibit diurnal patterns often characterized by an afternoon decline, even when measured under constant light and temperature conditions. Since we thought changes in the carbohydrate status could cause this pattern through feedback from starch and sucrose synthesis, we studied the natural fluctuations in photosynthesis rates of plants grown at 36 and 56 Pa CO2 at a FACE (free-air-CO2-enrichment) research site. Light-saturated photosynthesis varied by 40% during the day and was independent of the light-limited quantum yield of photosynthesis, which varied little through the day. Photosynthesis did not correspond with xylem water potential or leaf carbohydrate build-up, but rather with diurnal changes in air vapor-pressure deficit and light. The afternoon decline in photosynthesis also corresponded with decreased stomatal conductance and decreased Rubisco carboxylation efficiency which in turn allowed leaf-airspace CO2 partial pressure to remain constant. Growth at elevated CO2 did not affect the afternoon decline in photosynthesis, but did stimulate early-morning photosynthesis rates relative to the rest of the day. Plants grown at 56 Pa CO2 had higher light-limited quantum yields than those at 36 Pa CO2 but, there was no growth–CO2 effect on quantum yield when measured at 2 kPa O2. Therefore, understory plants have a high light-limited quantum yield that does not vary through the day. Thus, the major diurnal changes in photosynthesis occur under light-saturated conditions which may help understory saplings maximize their sunfleck-use-efficiency.

Sorghum (Sorghum bicolor) synthesizes a complex mixture of 3-deoxyanthocyanidin phytoalexins in response to inoculation with the non-pathogenic fungus Bipolaris maydis. The anthocyanin cyanidin 3-dimalonyl glucoside, is also synthesized naturally in response to light. To determine the order and time of appearance of these compounds, etiolated sorghum mesocotyls were inoculated with B. maydis and tissue extracts were analysed by photodiode array-HPLC every 2 h for the first 24 h and at 48 h post inoculation (hpi). Uninoculated and inoculated etiolated mesocotyls were also allowed to incorporate L-[U-14C] phenylalanine. Apigeninidin appeared at 10 hpi, followed by luteolinidin and apigeninidin 5-O-arabinoside at 14 hpi. Luteolinidin 5-methylether was not detected until 18 hpi and apigeninidin 7-methylether not until 20 hpi. The concentrations of the primary phytoalexins, apigeninidin, luteolinidin and apigeninidin 5-O-arabinoside, rose steadily between 12 and 24 hpi, and the levels of apigeninidin and luteolinidin were approximately equivalent by 24 hpi. However, between 24 and 48 hpi luteolinidin and luteolinidin 5-methylether accumulated rapidly so that by 48 hpi the amounts of luteolinidin and luteolinidin 5-methylether had increased approximately twofold. Radiolabelling also showed that 14C was incorporated into the 3-deoxyanthocyanidins and cyanidin 3-dimalonyl glucoside. Several other unidentified phenolic compounds also accumulated radioactivity.

Five evergreen subtropical tree species growing under identical environmental conditions were investigated to establish which hydraulic properties are genotypically rigid and which show phenotypic plasticity. Maximum xylem-specific conductivity (ks) correlated well with the anatomical characteristics (conduit diameter and density) for the four angiosperms Tecomaria capensis, Trichilia dregeana, Cinnamomum camphora and Barringtonia racemosa; the anatomy of the gymnosperm Podocarpus latifolius was not assessed. Huber values (functional xylem cross-sectional area [ratio ] leaf area) varied inversely with ks among species. Maximum leaf-specific conductivity was similar in the five unrelated species. Vulnerability of xylem to cavitation differed between species, as did the relationship between transpiration and water potential. Models of these parameters and isolated midday readings confirm that these trees operate at similar maximum leaf-specific conductivity (kl) values. The data are consistent with the hypothesis that conductivity characteristics (kl, ks) are influenced by environment, whereas vulnerability to cavitation is genetically determined.

Fourier Transform laser Raman spectroscopy was used to generate diagnostic spectra for pigments and biodegradative calcium oxalate in situ in two yellow-pigmented species of the lichen genus Acarospora from contrasting sites in the Antarctic and the Mediterranean. This non-intrusive technique was used to identify the photoprotective pigments rhizocarpic acid and β-carotene by their unique Raman spectral fingerprints. The use of low energy near-IR excitation at 1064 nm eliminated interference from autofluorescence of photosynthetic pigments. The insensitivity of the technique to water permitted the use of field-fresh material. The dominant yellow pigment, rhizocarpic acid, gave a diagnostic pattern of corroborative bands at wavenumbers (ν) 1596, 1665, 1620 and 1000 cm−1. It was possible to discriminate between hydration states of calcium oxalate; the monohydrate (whewellite) featured a ν(CO) stretching band at 1493 cm−1 whereas the dihydrate (weddellite) had a contrasting ν(CO) stretching band at 1476 cm−1. Fourier Transform deconvolution and intensity measurements were used to obtain relative quantitative data for rhizocarpic acid by using its ν(CO) and ν(CONH) amide modes, for carotenoid pigment by its ν(C = C) band at 1520 cm−1 and for calcium oxalates by their ν(CO) bands. ν(CO), ν(CONH) and ν(C = C) are the vibrational stretching modes of the carbonyl C = O, protein amide 1 and alkenyl C = C moieties, respectively, in the pigments and metabolic products of the Acarospora lichens. The ability to determine the precise (20 μm spot diameter) spatial distribution of these key functional molecules in field-fresh thallus profiles and variegations has great potential for understanding the survival strategies of lichens, which receive high insolation, including elevated levels of UV-B, under extremes of desiccation and temperature in hot and cold desert habitats.

Nitrogen limitation compromises the realization of yield potential in cereals more than any other single factor. In rice, the world's most important crop species, the assumption has long been that only ammonium-N is efficiently utilized. Consequently, nitrate utilization has been largely ignored, although fragmentary data have suggested that growth could be substantial on nitrate. Using the short-lived radiotracer 13N, we here provide direct comparisons of root transmembrane fluxes and cytoplasmic pool sizes for nitrate- and ammonium-N in a major variety of Indica rice (Oryza sativa), and show that nitrate acquisition is not only of high capacity and efficiency but is superior to that of ammonium. We believe our results have implications for rice breeding and molecular genetics as well as the design of water-management and fertilization regimes. Potential strategies to harness this hitherto unexplored N-utilization potential are proposed.

Sloughing of root cap cells and exudation of mucilage plays an important role in the penetration of compacted soils by roots. For the first time we have quantified the rate of sloughing of root cap cells in an abrasive growth medium that was compacted to create mechanical impedance to root growth. The number of maize (Zea mays) root cap cells sloughed into sand increased as a result of compaction, from 1930 to 3220 d−1 per primary root. This represented a 12-fold increase in the number of cells sloughed per mm root extension (from 60 to >700). We estimated that the whole of the cap surface area was covered with detached cells in compacted sand, compared with c. 7% of the surface area in loose sand. This lubricating layer of sloughed cells and mucilage probably decreases frictional resistance to soil penetration. The total carbon deposited by the root was estimated at c. 110 μg g−1 sand d−1. Sloughed cells accounted for <10% of the total carbon, the vast majority of carbon being contained in mucilage exudates.

The extent of vertical migration of 33P in thalli of the heathland lichen Cladonia portentosa was investigated under field conditions. 33P-labelled orthophosphate was introduced into the bottom 25 mm of podetia cut to a length of 50 mm from the apices. The distribution of label was scanned using a molecular imager immediately after incubation, and after growing for 8 wk and 6 months. Differences in the relative distribution of label between podetia harvested at the beginning and the end of the experiment showed that there had been a significant migration of 33P upwards out of the labelled 25 mm stratum towards the apex. This was confirmed by statistically significant changes in the median (md) and the 90 percentile of total relative distribution of 33P label. In a control treatment in which label was introduced into the apical 25 mm of podetia, which were then grown inverted (top down), no upward movement of label was detected. By contrast, a statistically significant reduction in the md of the distribution indicated migration downwards towards the thallus apex. The results are consistent with the hypothesis that P is recycled within podetia of mat-forming lichens, migrating from degrading basal regions upwards to the growing apices following a source–sink relationship.

In shallow flooded parts of rich fens Mentha aquatica might thrive in deeper water than Epilobium hirsutum but previous experiments have provided no clear indication that the flooding tolerance of these species differs. In this study we investigated, by measuring growth, biomass allocation and vegetative reproduction, whether the impact of water level on vegetative reproduction might produce different lower boundaries on water level gradients. There was a striking contrast between biomass production at high water levels and the field distribution of both species. After 18 wk, the mean biomass of E. hirsutum grown in waterlogged and flooded conditions was 82% and 54%, respectively, of the mean biomass production of drained plants. Biomass of waterlogged and flooded M. aquatica was reduced to 57% and 37% in drained conditions. Waterlogged and flooded E. hirsutum had swollen stem bases and invested a high proportion of biomass in adventitious roots. Stems of M. aquatica did not swell, formed few adventitious roots and maintained an equal proportion of below-ground roots at all water levels. The effect of water level on vegetative reproduction corresponded well with the lower hydrological boundaries of both species. When waterlogged and flooded, most rhizomes of E. hirsutum emerged from above-ground parts of the stem base and were oriented in an upward direction. Plants in flooded soil allocated less biomass to rhizomes and also reduced the number and size of rhizomes. Rhizome formation of M. aquatica on the other hand was not directly affected by water level and only depended on plant size. These differences in vegetative reproduction are discussed in relation to the different abilities of both species to oxygenate their below-ground roots. It was concluded that the mode of adaptation to soil flooding might also affect vegetative reproduction and, therefore, a species' ability for long-term persistence in soil-flooded habitats.

Segmentation of the aeration system and its anatomy was studied in rhizomes of common reed (Phragmites australis). Segmentation is achieved by nodal diaphragms which allow the passage of pressurised gas-flow but which also form effective barriers against flooding of the internal space in case of local injury. The pressures required to force water through the diaphragms were measured and compared with the anatomical structure of the diaphragm. The fine hydrophilic stellate parenchyma of the diaphragms was shown to act as a matrix supporting menisci of air–water interfaces and therefore preventing water movement until the ‘limit’ pressure (measured values ranged from 18 to >40 kPa) was overcome. Although the surfaces of the large stellate parenchyma and sclerenchyma strands covering the diaphragm are hydrophobic these components do little to prevent the ingress of water.

Recent research in whole-plant stomatal physiology, conducted largely with potted plants in controlled environments, suggests that stomatal conductance (gs) might be more closely linked to plant chemical variables than to hydraulic variables. To test this in a field situation, seasonal gs was examined in relation to a number of plant and environmental variables in 11 temperate, deciduous forest tree species. Stomatal conductance was generally better correlated with environmental variables (air temperature, vapor pressure deficit, PPFD) than with plant variables, and slightly better correlated with plant hydraulic variables (shoot water and osmotic potentials) than with plant chemical variables (xylem sap ABA concentration, xylem sap pH). We examined a model, developed previously for maize, which describes regulation of gs by xylem sap ABA concentration with leaf water status acting to modify stomatal sensitivity to the ABA signal. This model explained slightly more variation in seasonal gs in the forest trees than did single plant variables but not more variation than most single environmental variables. Response surface models, especially those incorporating environmental variables, were more consistently successful at explaining gs across species.

The hydraulic architecture of the secondary hemiepiphyte Monstera acuminata was examined in native plants from Los Tuxtlas, Veracruz, Mexico, to determine how it compared to better-known growth forms such as trees, shrubs, lianas and primary hemiepiphytes. Monstera acuminata starts its life cycle as a prostrate herb. As it ascends a tree or other vertical support, the stem becomes thicker, produces larger leaves, and may die back from the base upwards until only aerial feeding roots serve to connect the stem to the soil. Unlike the pattern of vessel-size distribution along the stems of woody dicotyledons, M. acuminata has its wider vessels at the top of the stem, decreasing in diameter towards the base. Also peculiar is the fact that Huber values (axis area/distal leaf area) tend to increase exponentially at higher positions within the plant. Based on the hydraulic conductivity (kh) and leaf-specific conductivity (LSC, kh/distal leaf area), the base of the stem potentially acts as a severe hydraulic constriction. This constriction is apparently not limiting, as aerial roots are produced further up the stem. The plants have remarkably strong root pressures, up to 225 kPa, which may contribute to the maintenance of functional vessels by refilling them at night or during periods of very high atmospheric humidity, as in foggy weather and rain. In common with dicotyledonous plants, vessel length, vessel diameter, kh, specific conductivity (ks, kh/axis area) and LSCs were all positively correlated with axis diameter. The features of the hydraulic architecture of M. acuminata may be an evolutionary consequence of an anatomical constraint (lack of vascular cambium and therefore of secondary growth) and the special requirements of the hemiepiphytic growth form.

Adventitious rooting might be induced in recalcitrant woody genotypes by infection with Agrobacterium rhizogenes, and, in some cases, might also require exogenous auxin. The objective of the present study was to determine how agrobacteria trigger root formation in the stem of a recalcitrant woody microcutting, which cytological events result from the combined presence of infection and exogenous auxin, and which types of roots are induced by infection. Microcuttings of a recalcitrant walnut (Juglans regia), infected or not with A. rhizogenes strain 1855, were cultured with either indolebutyric acid (IBA), IAA, or without exogenous hormones, to induce rhizogenesis. They were cytohistologically and ultrastructurally investigated at various times in culture. Southern blot and PCR analyses were performed to verify the frequency of transgenic, chimeric and bacterium-containing roots. The infection was sufficient per se to stimulate rhizogenesis. Rooting on the infected cuttings was enhanced by exogenous IBA, which accelerated and increased root meristemoid formation, in comparison with without-hormone treatment. Meristemoids were organized both directly by the cambial cells and indirectly by the callus, and showed a pluricellular origin. Inter and intracellular bacteria were observed in the stem throughout the culture period (30 d). They were preferentially present in the vessels, and mainly in those showing polyphenol deposition. In the infected IAA-treated cultures, a high level of secondary xylem formation occurred instead of rhizogenesis. Nontransformed roots were preferentially produced by the infected cuttings treated with the auxins. Bacterium-containing and chimeric roots were produced by infected cuttings independently of the treatment. Thus, in a recalcitrant walnut, nontransformed root meristemoids are stimulated by combining infection and exogenous indolebutyric acid. Furthermore, the persistence of bacteria in the stem during the culture and the pluricellular origin of the meristemoids explain the presence of the bacterium-containing and chimeric roots.