Net primary production of forests: a constant fraction of gross primary production?

Considerable progress has been made in our ability to model and measure annual gross primary production (GPP) by terrestrial vegetation. But challenges remain in estimating maintenance respiration (R(m)) and net primary production (NPP). To search for possible common relationships, we assembled annual carbon budgets from six evergreen and one deciduous forest in Oregon, USA, three pine plantations in New South Wales, Australia, a deciduous forest in Massachusetts, USA, and a Nothofagus forest on the South Island of New Zealand. At all 12 sites, a standard procedure was followed to estimate annual NPP of foliage, branches, stems, and roots, the carbon expended in synthesis of these organs (R(g)), their R(m), and that of previously produced foliage and sapwood in boles, branches, and large roots. In the survey, total NPP ranged from 120 to 1660 g C m(-2) year(-1), whereas the calculated fraction allocated to roots varied from 0.22 to 0.63. Comparative analysis indicated that the total NPP/GPP ratio was conservative (0.47 +/- 0.04 SD). This finding supports the possibility of greatly simplifying forest growth models. The constancy of the NPP/GPP ratio also provides an incentive to renew efforts to understand the environmental factors affecting partitioning of NPP above and belowground.

Evaluating a simple radiation/dry matter conversion model using data from Eucalyptus globulus plantations in Western Australia.

A simple model that describes growth in terms of physical and physiological processes is needed to predict growth rates and hence the productivity of trees at particular sites. The linear relationship expected between absorbed photosynthetically active radiation (phi(pa), MJ m(-2)) and dry mass production (G(t)); i.e., G(t) = epsilonphi(pa), where epsilon is the radiation utilization coefficient, was fitted to three years' data from five Western Australian Eucalyptus globulus Labill. plantations for which monthly growth measurements, leaf area indices, weather data and soil water measurements were available. Reductions in growth efficiency relative to absorbed photosynthetically active radiation were associated with high vapor pressure deficits (D, kPa) so the relationship between monthly aboveground biomass increments and D was used to calculate utilizable phi(pa). Plotting cumulative aboveground growth against utilizable phi(pa) gave strong linear relationships with slope epsilon. Values of epsilon ranged from 0.93 to 2.23 g dry mass MJ(-1) phi(pa). The variation could not be explained either in terms of soil water content in the root zones, because all plantations appeared to have access to groundwater, or in terms of soil chemistry. A value of epsilon approximately 2.2 is considered near the maximum likely to be applicable to Eucalyptus plantations. An interesting peripheral finding was a strong relationship between allometric ratios and soil phosphorus; this, if confirmed elsewhere, will be of considerable value in converting biomass increments to wood production. There was also a strong negative relationship between the average ratio of leaf/total aboveground biomass and soil nitrogen content.

Evaluating progress toward closed forest models based on fluxes of carbon, water and nutrients.

Closed system models are defined as mathematical models of systems having specified boundaries within which all flows into and out of the system are accounted for. Closure is obtained experimentally when we can measure all the flows and do not depend on residuals. The meeting on which this volume is based discussed a range of models and approaches to modeling, and the possibility of achieving closure. There was general agreement that we can develop closed system models of the water balance, carbon cycle and nutrient fluxes at the stand level. Confidence in our ability to account for all the flows is greatest for water, decreasing progressively for carbon and nutrients. The priority areas for research on the carbon balance are belowground processes, foliage dynamics and respiration. The problems requiring particular attention in relation to the water balance are the measurement of interception losses, lateral flow in the soil and evaporation from snow. Areas warranting particular research attention in relation to nutrient fluxes through forest stands are the rates, and the controls on rates, of nutrient uptake by trees, and rates of mineralization with emphasis on the importance of microbial processes at the ecosystem level. Most models are written for uniform conditions. Forests are not uniform so the problem of heterogeneity, and how to deal with it in models, requires considerable attention, as does the question of how to scale up, to deal with large areas. There are a great many forest models of all types and the continual development of new ones may not be an effective use of research resources. There is a need for some assessment of the range of models currently existing, or under development, and for moves toward a directed strategy of model structure and development.

Water stress and seedling growth of two eucalypt species from contrasting habitats.

Seedlings of Eucalyptus maculata Hook (mesic environment) and E. brockwayi C.A. Gardn. (arid environment) were supplied 100, 70 or 40% of their water requirements estimated from leaf area and the water used by well-watered seedlings. Restricting water supply caused large differences in growth rates, which were related to large differences in total leaf area. There was a fivefold range of variation in number of leaves per plant, and a reduction of up to 20% in average leaf size as a consequence of restricting water supply. Eucalyptus maculata seedlings produced more dry matter than E. brockwayi seedlings, but net assimilation rate was higher in E. brockwayi seedlings. Transpiration rates were also higher in E. brockwayi than in E. maculata. Leaf expansion was analyzed as a function of water stress integral (S(Psi)), which is the cumulative integral over time of predawn water potential below a datum. The leaf area achieved at any stress level was not uniquely dependent on total S(Psi), there was a secondary effect associated with reduced leaf growth caused by previous stress. At any value of S(Psi), reductions in leaf growth of water-stressed seedlings relative to leaf growth of well-watered control seedlings were greater in E. maculata seedlings than in E. brockwayi. Treatment differences in both species showed that, within the levels of stress applied, a moderate water stress over a long period of time was more detrimental to dry matter production than a severe stress for a short time.

Experimental approaches to the study of the effects of nutrients and water on carbon assimilation by trees.

This paper outlines the experiemental approaches needed to provide detailed information about the effects of nutrients and water on carbon assimilation, and hence growth of trees. Data from a large multidisciplinary experiment being conducted in Australia illustrate important aspects of the approach advocated. The system model consists of a series of submodels describing the main physiological processes considered to govern the growth of trees. It is argued that field experiments designed to investigate physiological processes and their interactions should involve treatments that cause large differences in the growth rates of trees and in the condition of the trees at any time. The biomass of experimental stands should be defined. Measurements of stem and leaf growth, litterfall, leaf photosynthetic characteristics, plant water status and soil water content, plant nutrient status and nutrient mineralization rates in the soil provide the information needed to explain observed differences in growth and to refine the models underlying the research.

A dynamic model for studying flow of water in single trees.

Flow of water in a single tree was modeled in terms of the Darcy equation using a catena of four compartments: root, stem (further divided into discs), branches and leaves. Within each compartment or disc, water content was related to both water potential and conductivity of the xylem tissue using power or logarithmic functions, thus introducing both capacitance and variable resistance to flow in the model. Transpiration from the leaves to the atmosphere was used as the upper boundary to the model, and the soil-root interface as the lower boundary. Parameters for the water content, water potential and conductivity functions, together with physical dimensions were obtained by direct measurement or from the literature. A sensitivity analysis showed that the largest changes in simulated water potential and flow were associated with changes in the parameters directly controlling conductivity. Simulation of both smoothed diurnal changes and stepwise changes showed a phase lag down the tree, with flow tending to approach a steady state, but with changes in the gradients of water potential, water content and conductivity. A preliminary test of the model was made against field data using the Penman-Monteith equation to estimate the transpiration rate in a well-watered Pinus contorta Dougl. stand. Stem flow, water potential and water content were measured directly on a representative tree, which was subsequently harvested to provide dimensions and laboratory estimations of the parameters in the functions by direct measurement.