Deborah A. Clark

MEASURING NET PRIMARY PRODUCTION IN FORESTS: CONCEPTS AND FIELD METHODS

There are pressing reasons for developing a better understanding of net primary production (NPP) in the worlds forests. These ecosystems play a large role in the worlds carbon budget, and their dynamics, which are likely to be responding to global changes in climate and atmospheric composition, have major economic implications and impacts on global biodiversity. Although there is a long history of forest NPP studies in the ecological literature, current understanding of ecosystem-level production remains lim- ited. Forest NPP cannot be directly measured; it must be approached by indirect methods. To date, field measurements have been largely restricted to a few aspects of NPP; methods are still lacking for field assessment of others, and past studies have involved confusion about the types of measurements needed. As a result, existing field-based estimates of forest NPP are likely to be significant underestimates. In this paper we provide a conceptual framework to guide efforts toward improved estimates of forest NPP. We define the quantity NPP* as the summed classes of organic material that should be measured or estimated in field studies for an estimate of total NPP. We discuss the above- and belowground components of NPP* and the available methods for measuring them in the field. We then assess the implications of the limitations of past studies for current understanding of NPP in forest ecosystems, discuss how field NPP* measurements can be used to complement tower-based studies of forest carbon flux, and recommend design criteria for future field studies of forest NPP.

Reconciling carbon-cycle concepts, terminology, and methods

Recent projections of climatic change have focused a great deal of scientific and public attention on patterns of carbon (C) cycling as well as its controls, particularly the factors that determine whether an ecosystem is a net source or sink of atmospheric carbon dioxide (CO2). Net ecosystem production (NEP), a central concept in C-cycling research, has been used by scientists to represent two different concepts. We propose that NEP be restricted to just one of its two original definitions—the imbalance between gross primary production (GPP) and ecosystem respiration (ER). We further propose that a new term—net ecosystem carbon balance (NECB)—be applied to the net rate of C accumulation in (or loss from [negative sign]) ecosystems. Net ecosystem carbon balance differs from NEP when C fluxes other than C fixation and respiration occur, or when inorganic C enters or leaves in dissolved form. These fluxes include the leaching loss or lateral transfer of C from the ecosystem; the emission of volatile organic C, methane, and carbon monoxide; and the release of soot and CO2 from fire. Carbon fluxes in addition to NEP are particularly important determinants of NECB over long time scales. However, even over short time scales, they are important in ecosystems such as streams, estuaries, wetlands, and cities. Recent technological advances have led to a diversity of approaches to the measurement of C fluxes at different temporal and spatial scales. These approaches frequently capture different components of NEP or NECB and can therefore be compared across scales only by carefully specifying the fluxes included in the measurements. By explicitly identifying the fluxes that comprise NECB and other components of the C cycle, such as net ecosystem exchange (NEE) and net biome production (NBP), we can provide a less ambiguous framework for understanding and communicating recent changes in the global C cycle.

SPACING DYNAMICS OF A TROPICAL RAIN FOREST TREE: EVALUATION OF THE JANZEN-CONNELL MODEL

As a partial explanation for the maintenance of high tree diversity in wet tropical forests, Janzen (1970) and Connell (1971) independently hypothesized that natural enemies act to increase spacing within these tree populations through disproportionately high attack on progeny near adults. Both authors also hypothesized a minimum critical distance effect, because of 100% progeny mortality within a given distance of adults. We describe the necessary and sufficient conditions for testing these hypotheses, and show that attempts to evaluate them have been hampered by use of the inappropriate standard of regular spacing. Data describing the spacing dynamics of Dipteryx panamensis, a rain forest canopy tree, support both hypotheses. From 7 mo to 2 yr postgermination, seedling survival was positively correlated with distance to adult and negatively correlated with local conspecific seedling density. Partial correlation was used to separate the effects of density and distance, and it was shown that seedling density was the only significant factor in this case. Older juveniles and saplings occurred at greater distances from the nearest conspecific adult than did 1980 seedlings. No seedlings or juveniles survived within 8 m of an adult bole. A review of 24 data sets on tropical woody plants showed that most evidence indicates either density-dependence or distance-dependence in progeny mortality, as hypothesized by Janzen and Connell. Some positive evidence also exists for the minimum critical distance effect for tropical trees. In most of the cases involving seedling mortality, however, alternative causal factors such as intracohort competition or allelopathy were not ruled out. Before generalizations can be made about this process in tropical forests, carefully designed studies are needed on more populations of tropical trees.

NET PRIMARY PRODUCTION IN TROPICAL FORESTS: AN EVALUATION AND SYNTHESIS OF EXISTING FIELD DATA

Information on net primary production in tropical forests is needed for the development of realistic global carbon budgets, for projecting how these ecosystems will be affected by climatic and atmospheric changes, and for evaluating eddy covariance mea- surements of tropical forest carbon flux. However, a review of the database commonly used to address these issues shows that it has serious flaws. In this paper we synthesize the data in the primary literature on NPP in old-growth tropical forests to produce a consistent data set on NPP for these forests. Studies in this biome have addressed only a few NPP com- ponents, all aboveground. Given the limited scope of the direct field measurements, we sought relationships in the existing data that allow estimation of unmeasured aspects of production from those that are more easily assessed. We found a predictive relationship between annual litterfall and aboveground biomass increment. For 39 diverse tropical forest sites, we then developed consistent, documented estimates of the upper and lower bounds around total NPP to serve as benchmarks for calibrating and validating biogeochemical models with respect to this biome. We developed these estimates based on existing field measurements, current understanding of aboveground consumption and biogenic volatile organic carbon emissions, and our judgment that belowground production is bounded by the range 0.2-1.2 3 ANPP (aboveground NPP). Across this broad spectrum of tropical forests (dry to wet, lowland to montane, nutrient-rich to nutrient-poor soils), our estimates of lower and upper bounds on total NPP range from 1.7 to 11.8 Mg C·ha 21 ·yr 21 (lower bounds) and from 3.1 to 21.7 Mg C·ha 21 ·yr 21 (upper bounds). We also showed that two relationships that have been used for estimating NPP (the Bray-Gorham relationship based on leaf litterfall and the Miami model based on temperature or precipitation) are not valid for the tropical forest biome.

Landscape-scale variation in forest structure and biomass in a tropical rain forest

A better understanding of the reasons for variation in tropical rain forest (TRF) structure is important for quantifying global above-ground biomass (AGBM). We used three data sets to estimate stem number, basal area, and AGBM over a 600-ha oldgrowth TRF landscape (La Selva, N.E. Costa Rica). We analyzed the effects of soil type, slope angle, topographic position, and different sample designs and measurement techniques on these estimates. All three data sets were for woody stems 10 cm in diameter. Estimated AGBM was determined from stand-level measurements using Brown’s (Brown, 1997) allometric equation for Tropical Wet Forest trees. One data set was from three subjectively-sited 4-ha plots (the ‘OTS plots’), another was based on 1170 0.01 ha plots spaced on a regular grid (the ‘Vegetation map plots’), and the third was from 18 0.5 ha plots (the ‘Carbono plots’) sited to provide unbiased samples of three edaphic conditions: flat inceptisol old alluvial terraces; flat ultisol hill-tops; and steep ultisol slopes. Basal area, estimated AGBM and the contributions of major life forms were similar among studies, in spite of the differences in sampling design and measurement techniques. Although the Carbono plots on flat inceptisols had significantly larger and fewer trees than those on ultisols, AGBM did not vary over the relatively small edaphic gradient in upland areas at La Selva. On residual soils, the largest trees were on the flattest topographic positions. Slope angle per se was not correlated with basal area or AGBM within the residual soils. Errors introduced by palm and liana life forms, as well as hollow trees, did not significantly affect AGBM estimates. In contrast, the methods used to measure buttressed trees had a large impact. Plot sizes of 0.35‐0.5 ha were sufficient to achieve coefficients of variation of <12% for basal area with only six replicates in a given edaphic type. AGBM estimates ranged from 161 to 186 Mg/ha. These low values appear to be mainly due to the Tropical Wet Forest allometry equation used. This in turn may be indicative of a real and substantially lower ratio of biomass/basal area in Tropical Wet Forest than in Tropical Moist, as previously noted by Brown (1996). Our results indicate that for upland TRF landscapes with levels of environmental variation similar to La Selva, AGBM will be relatively insensitive to soil type and topography. However, because topography and soil type had much stronger effects on stem size, stand density, and spatial heterogeneity of stems, stand dynamics may be more sensitive than AGBM to this range of conditions. We recommend that future studies of landscape-scale forest structure employ stratified sampling designs across major environmental gradients. Unbiased sampling with replication, combined with consistent and well-documented measurement techniques, will lead to a greatly improved understanding of the magnitude of and reasons for variation in forest structure and AGBM within TRF landscapes. # 2000 Elsevier Science B.V. All rights reserved.

Belowground carbon allocation in forests estimated from litterfall and IRGA-based soil respiration measurements

Allocation of C to belowground plant structures is one of the most important, yet least well quantified fluxes of C in terrestrial ecosystems. In a literature review of mature forests worldwide, Raich and Nadelhoffer (1989) suggested that total belowground carbon allocation (TBCA) could be estimated from the difference between annual rates of soil respiration and aboveground litterfall. Here we analyze new measurements of soil respiration and litterfall, including data from the Ameriflux network. Our results generally agree with Raich and Nadelhoffer’s previous work. A regression analysis of data from mature forests produced the following relationship: annual soil respiration = 287 + 2.80 × annual litterfall. This regression slope indicates that, on average, soil respiration is roughly three times ab oveground litterfall-C, which further implies that TBCA is roughly twice annual aboveground litterfall-C. These inferences are based on the uncertain assumption of soil C stocks being at steady state. Nevertheless, changes in soil C would have to be very large to modify the conclusion that TBCA is generally much larger than litterfall. Among only mature temperate hardwood forests, however, the correlation between litterfall and soil respiration was poor, and the correlation among years for a single site was also poor. Therefore, the regression cannot be relied upon to provide accurate estimates of soil respiration or TBCA for individual sites. Moreover, interannual variation in TBCA, short-term changes in C stocks, or different temporal scales controlling leaf litter production and soil respiration may cause important deviations from the global average. The regression slope for data from young forests is steeper, possibly indicating proportionally greater TBCA, but the steady-state assumption is more problematic for young forests. This method

Distribution and effects on tree growth of lianas and woody hemiepiphytes in a Costa Rican tropical wet forest

We evaluated occurrence and abundance of lianas and woody hemiepiphytes on canopy and emergent tree species in primary tropical wet forest at the La Selva Biological Station, Costa Rica. Two pioneers, Cecropia obtusifolia and C. insignis, lacked both lianas and hemiepiphytes. The seven non-pioneer species differed significantly in their loads of lianas and hemiepiphytes. For all non-pioneer species, two measures of liana and hemiepiphyte loads (the percentage of the crown occupied and the combined basal area of descending hemiepiphyte and liana roots and stems) increased significantly with tree diameter. In all non-pioneer species, most trees >70 cm diameter (50-97%) were colonized. Lianas occupied more trees and had a smaller mean host diameter than did hemiepiphytes; however, basal area of descending roots or stems was equivalent for the two life forms in occupied trees. We used partial correlations con- trolling for tree diameter to evaluate the relationship between annual tree diameter growth and loads of hemiepiphytes and lianas for six non-pioneer species. Five of the six species showed a significant negative correlation between loads and diameter growth. Existing published data show that the high incidence of lianas and hemiepiphytes at La Selva is paralleled in most other Neotropical wet forests.

ASSESSING THE GROWTH OF TROPICAL RAIN FOREST TREES: ISSUES FOR FOREST MODELING AND MANAGEMENT

Growth performance was assessed for a diverse suite of canopy and emergent tree species in a lowland neotropical rain forest (the La Selva Biological Station, northeastern Costa Rica). Species were evaluated based on annual diameter measurements of large samples of individuals in all post-seedling size classes, over a 12-yr period. The study species were seven non-pioneers (Minquartia guianensis, Lecythis ampla, Hymenolobium mesoamericanum, Simarouba amara, Dipteryx panamensis, Pithecellobium elegans, and Hyeronima alchorneoides) and two pioneers (Cecropia obtusifolia and C. insignis). For each species, inherent growth capacity through ontogeny was estimated as the mean of the five largest annual increments (from different individuals) in each juvenile and adult size class. At all tree sizes, species differed highly significantly in this measure. In the small sapling sizes (≤4-cm diameter), the two pioneers showed markedly higher maximum growth than all the non-pioneers, in line with current understanding...

Abundance, growth and mortality of very large trees in neotropical lowland rain forest.

Very large trees, arbitrarily defined as those over 70 cm diameter above buttresses, account for a major portion of the above-ground biomass in neotropical rain forests. Owing to the scarcity of individuals of a given species and the difficulty of accurate measurement, there are few species-level data on the growth, mortality, and abundance of species that regularly reach emergent status. We report such data for very large individuals from old-growth tropical wet forest at the La Selva Biological Station in the Atlantic lowlands of the Republic of Costa Rica. The landscape-scale abundance of all species reaching over 70 cm diameter was assessed using 515 0.01-ha quadrats located at grid points in a 500 ha area of old-growth forest. In the total sample of 2301 stems 10 cm or more in diameter, very large individuals accounted for 2% of the stems, 23% of the basal area, and 27% of the estimated above-ground biomass. Growth and survival for five species that regularly attain emergent status were measured in a 150 ha area within the 500 ha plot. Survival of 282 very large individuals of the five species was measured over 6 years. The mean annual mortality rate of the total sample was only 0.6% year-1. Mean annual diameter growth increments varied from 1.9 to 5.2 mm year-1 among species, and were negatively correlated with diameter in four of the five species. For a sample of 193 individuals measured over 7 years, growth almost exactly equalled losses in basal area and biomass due to mortality. Because all of these species are regularly recruiting new trees into the over 70 cm diameter class, the amount of biomass in the large-individual size class is increasing over the 150 ha old-growth study area. Historic disturbance and/or current climatic change are hypothesized to account for the increase. We identify lack of standard diameter measurement criteria, and small and potentially unrepresentative plot locations as two problems in assessing the role of very large trees in other neotropical forests. Future studies should sample larger areas; this will increase the generality of the conclusions and will make possible a species-level comparison of the ecology of very large tropical trees.

Seedling Dynamics of a Tropical Tree: Impacts of Herbivory and Meristem Damage

The effects of herbivory and apical meristem damage on seedlings of the tropical wet forest tree Dipteryx panamensis (Leguminosae: Papilionoideae) were evaluated for natural and trans- planted seedlings at the La Selva Biological Station, Costa Rica. Original seedling leaves were long lived, up to 21 mo, and new leaf production was slow. Damage to the original leaves was gradual and continued to accumulate well after the leaves matured. The amount of leaf area lost was serially correlated between census periods for individual seedlings. Seedling longevity was highly correlated with the percentage of original leaf area present at 1 mo after germination, and with the number of leaves present at 7 mo of age. Leaf damage and terminal meristem damage were positively related to seedling density. Onset of herbivory was earlier for seedlings in a dense natural population than for seedlings transplanted to sites with no nonspecific seedlings within 10 m. In a 1-ha population, the number of leaves on 7-mo- old seedlings was negatively correlated with the number of seedling neighbors within 10 m. Incidence of terminal meristem damage was positively correlated with seedling density. Leaf number and mer- istem damage were also correlated with distance from the nearest Dipteryx adult. Only 19% of the seedlings survived the 1 st yr without losing at least 1 apical meristem. Meristem damage was highly correlated with amount of rainfall in the preceding 60 d. We detected no effect of meristem damage on seedling survivorship. Processes acting on very young seedlings can have major effects on tree regeneration. Herbivory on the first seedling leaves may be one such important process.