Stem water storage capacity and diurnal patterns of water use were studied in five canopy trees of a seasonal tropical forest in Panama. Sap flow was measured simultaneously at the top and at the base of each tree using constant energy input thermal probes inserted in the sapwood. The daily stem storage capacity was calculated by comparing the diurnal patterns of basal and crown sap flow. The amount of water withdrawn from storage and subsequently replaced daily ranged from 4 kg d -1 in a 0·20-mdiameter individual of Cecropia longipes to 54 kg d -1 in a 1·02-m-diameter individual of Anacardium excelsum, representing 9-15% of the total daily water loss, respectively. Ficus insipida, Luehea seemannii and Spondias mombin had intermediate diurnal water storage capacities. Trees with greater storage capacity maintained maximum rates of transpiration for a substantially longer fraction of the day than trees with smaller water storage capacity. All five trees conformed to a common linear relationship between diurnal storage capacity and basal sapwood area, suggesting that this relationship was species-independent and size-specific for trees at the study site. According to this relationship there was an increment of 10 kg of diurnal water storage capacity for every 0·1 m 2 increase in basal sapwood area. The diurnal withdrawal of water from, and refill of, internal stores was a dynamic process, tightly coupled to fluctuations in environmental conditions. The variations in basal and crown sap flow were more synchronized after 1100 h when internal reserves were mostly depleted. Stem water storage may partially compensate for increases in axial hydraulic resistance with tree size and thus play an important role in regulating the water status of leaves exposed to the large diurnal variations in evaporative demand that occur in the upper canopy of seasonal lowland tropical forests.

The empirical calibration of Granier-type heat dissipation sap flow probes that relate temperature difference (DeltaT) to sap velocity (v) was reevaluated in stems of three tropical tree species. The original calibration was confirmed when the entire heated probe was in contact with conducting xylem, but mean v was underestimated when part of the probe was in contact with nonconducting xylem or bark. Analysis of the effects of nonuniform sap velocity profiles on heat dissipation estimates showed that errors increased as v and the proportion of the probe in nonconducting wood increased. If half of a 20-mm probe is in sapwood with a v of 0.15 mm s(-1) and the other half is in nonconducting wood, then mean v for the whole probe can be underestimated by as much as 50%. A correction was developed that can be used if the proportion of the probe in nonconducting wood is known. Even with the entire heated probe in contact with conducting xylem, v would be underestimated when radial velocity gradients are present. In this case, the error would be smaller except when velocity gradients are very steep, as can occur in species with ring-porous wood anatomy. Errors occur because the relationship between DeltaT and v is nonlinear. Mean DeltaT along the probe is therefore not a measure of mean v, and users of heat dissipation probes should not assume that v is integrated along the length of the probe. The same type of error can occur when DeltaT is averaged through time while v is changing, but the error is small unless there are sudden, step changes between zero and high sap velocity. It is recommended that relatively short probes (20 mm or less) be used and that probes longer than the depth of conducting sapwood be avoided. Multiple probes inserted to a range of depths should be used in situations where steep gradients in v are expected. If these conditions are met, heat dissipation probes remain useful and widely applicable for measuring sap flow in woody stems.

Little is known about partitioning of soil water resources in species-rich, seasonally dry tropical forests. We assessed spatial and temporal patterns of soil water utilization in several canopy tree species on Barro Colorado Island, Panama, during the 1997 dry season. Stable hydrogen isotope composition (δD) of xylem and soil water, soil volumetric water content (θ), and sap flow were measured concurrently. Evaporative fractionation near the soil surface caused soil water δD to decrease from about -15‰ at 0.1 m to -50 to -55‰ at 1.2 m depth. Groundwater sampled at the sources of nearby springs during this period yielded an average δD value of -60‰. θ increased sharply and nearly linearly with depth to 0.7 m, then increased more slowly between 0.7 and 1.05 m. Based on xylem δD values, water uptake in some individual plants appeared to be restricted largely to the upper 20 cm of the soil profile where θ dropped below 20% during the dry season. In contrast, other individuals appeared to have access to water at depths greater than 1 m where θ remained above 45% throughout the dry season. The depths of water sources for trees with intermediate xylem δD values were less certain because variation in soil water δD between 20 and 70 cm was relatively small. Xylem water δD was also strongly dependent on tree size (diameter at breast height), with smaller trees appearing to preferentially tap deeper sources of soil water than larger trees. This relationship appeared to be species independent. Trees able to exploit progressively deeper sources of soil water during the dry season, as indicated by increasingly negative xylem δD values, were also able to maintain constant or even increase rates of water use. Seasonal courses of water use and soil water partitioning were associated with leaf phenology. Species with the smallest seasonal variability in leaf fall were also able to tap increasingly deep sources of soil water as the dry season progressed. Comparison of xylem, soil, and groundwater δD values thus pointed to spatial and temporal partitioning of water resources among several tropical forest canopy tree species during the dry season.

Summary Crassulacean acid metabolism (CAM) is a specialized mode of photosynthesis that features nocturnal CO2 uptake, facilitates increased water‐use efficiency (WUE), and enables CAM plants to inhabit water‐limited environments such as semi‐arid deserts or seasonally dry forests. Human population growth and global climate change now present challenges for agricultural production systems to increase food, feed, forage, fiber, and fuel production. One approach to meet these challenges is to increase reliance on CAM crops, such as Agave and Opuntia, for biomass production on semi‐arid, abandoned, marginal, or degraded agricultural lands. Major research efforts are now underway to assess the productivity of CAM crop species and to harness the WUE of CAM by engineering this pathway into existing food, feed, and bioenergy crops. An improved understanding of CAM has potential for high returns on research investment. To exploit the potential of CAM crops and CAM bioengineering, it will be necessary to elucidate the evolution, genomic features, and regulatory mechanisms of CAM. Field trials and predictive models will be required to assess the productivity of CAM crops, while new synthetic biology approaches need to be developed for CAM engineering. Infrastructure will be needed for CAM model systems, field trials, mutant collections, and data management.

Tropical moist forests are notable for their richness in tree species. The presence of such a diverse tree flora presents potential problems for scaling up estimates of water use from individual trees to entire stands and for drawing generalizations about physiological regulation of water use in tropical trees. We measured sapwood area or sap flow, or both, in 27 co-occurring canopy species in a Panamanian forest to determine the extent to which relationships between tree size, sapwood area and sap flow were species-specific, or whether they were constrained by universal functional relationships between tree size, conducting xylem area, and water use. For the 24 species in which active xylem area was estimated over a range of size classes, diameter at breast height (DBH) accounted for 98% of the variation in sapwood area and 67% of the variation in sapwood depth when data for all species were combined. The DBH alone also accounted for > or = 90% of the variation in both maximum and total daily sap flux density in the outermost 2 cm of sapwood for all species taken together. Maximum sap flux density measured near the base of the tree occurred at about 1,400 h in the largest trees and 1,130 h in the smallest trees studied, and DBH accounted for 93% of the variation in the time of day at which maximum sap flow occurred. The shared relationship between tree size and time of maximum sap flow at the base of the tree suggests that a common relationship between diurnal stem water storage capacity and tree size existed. These results are consistent with a recent hypothesis that allometric scaling of plant vascular systems, and therefore water use, is universal.

Concurrent, independent measurements of stomatal conductance (g^), transpiration [E) and microenvironmental variables were used to characterize control of crown transpiration in four tree species growing in a moist, lowland tropical forest. Access to the upper forest canopy was provided by a construction crane equipped with a gondola. Estimates of boundary layer conductance (g,,) obtained with two independent methods permitted control of E to be partitioned quantitatively between gâ nd gh using ^ dimensionless decoupling coefficient (12) ranging from zero to 1. A combination of high g^ (c. 300-600 mmol m~^ s~') and low wind speed, and therefore relatively low g,, (c. 100-800 mmol m~^ s"'), strongly decoupled E from control by stomata in all four species (Q = 0-7-0-9). Photosyntbetic water-use efficiency was predicted to increase rather than decrease with increasing gs because g,, was relatively low and internal conductance to CO2 transfer was relatively high. Responses of g, to humidity were apparent only when the leaf surface, and not the bulk air, was used as the reference point for determination of external vapour pressure. However, independent measurements of crown conductance (g^.), a total vapour phase conductance that included stomatal and boundary layer components, revealed a clear decline in ge with increasing leaf-to-bulk air vapour pressure difference (V^), because the external reference points for determination of ge and V,, were compatible. The relationships between g^ and V,, and between g, and V, appeared to be distinct for each species. However, when gj, and gj were normalized by the branch-specific ratio of leaf area to sapwood area (LA/SA), a morphological index of potential transpirational demand relative to water transport capacity, a common relationship between conductance and evaporative demand for all four species emerged. Taken together, these results implied that, at a given combination of LA/SA and evaporative demand scaled to the appropriate reference point, the vapour phase conductance and therefore transpiration rates on a leaf area basis were identical in all four contrasting species studied.

The nutrient demands of regrowing tropical forests are partly satisfied by nitrogen-fixing legume trees, but our understanding of the abundance of those species is biased towards wet tropical regions. Here we show how the abundance of Leguminosae is affected by both recovery from disturbance and large-scale rainfall gradients through a synthesis of forest inventory plots from a network of 42 Neotropical forest chronosequences. During the first three decades of natural forest regeneration, legume basal area is twice as high in dry compared with wet secondary forests. The tremendous ecological success of legumes in recently disturbed, water-limited forests is likely to be related to both their reduced leaflet size and ability to fix N, which together enhance legume drought tolerance and water-use efficiency. Earth system models should incorporate these large-scale successional and climatic patterns of legume dominance to provide more accurate estimates of the maximum potential for natural nitrogen fixation across tropical forests.

Water uptake and transport were studied in eight liana species in a seasonally dry tropical forest on Barro Colorado Island, Panama. Stable hydrogen isotope composition (δD) of xylem and soil water, soil volumetric water content (θ v ), and basal sap flow were measured during the 1997 and 1998 dry seasons. Sap flow of several neighboring trees was measured to assess differences between lianas and trees in magnitudes and patterns of daily sap flow. Little seasonal change in θ v was observed at 90-120 cm depth in both years. Mean soil water δD during the dry season was −19‰ at 0-30 cm, −34‰ at 30-60 cm, and −50‰ at 90-120 cm. Average values of xylem δD among the liana species ranged from -28‰ to -44‰ during the middle of the dry season, suggesting that water uptake was restricted to intermediate soil layers (30-60 cm). By the end of the dry season, all species exhibited more negative xylem δD values (-41‰ to -62‰ ), suggesting that they shifted to deeper water sources. Maximum sap flux density in cooccurring lianas and trees were comparable at similar stem diameter (DBH). Furthermore, lianas and trees conformed to the same linear relationship between daily sap flow and DBH. Our observations that lianas tap shallow sources of soil water at the beginning of the dry season and that sap flow is similar in lianas and trees of equivalent stem diameter do not support the common assumptions that lianas rely