Christopher Uhl

Vegetation Dynamics in Amazonian Treefall Gaps

In a 1 -ha plot of old-growth tierra firme forest near San Carlos de Rio Negro, Venezuela (northwest Amazon Basin), 88 trees - 10 cm dbh out of 786 died over a 10-yr period. Most deaths resulted in the formation of small (5-100 m2) canopy openings (gaps). Occasionally, large gaps are formed in this region when strong winds topple many trees together. In five small (single-treefall) gaps and one large (multiple-treefall) gap we studied changes in soil fertility and nutrient leaching, and also plant establishment, mortality, and growth during the first 4 yr following gap formation. Gaps were divided into four zones, or mi- crohabitats: a trunk zone, an open zone (between bole and forest edge), a crown zone, and a root-pit zone. Sampling was conducted in each microhabitat. Soil nutrient levels in single-treefall gaps did not differ in a predictable fashion in response to microhabitat within gap or gap age. Moreover, except for a small, short-term increase in NO3-N, leaching losses from single-treefall gaps did not differ from forest levels. Advance regeneration plays a dominant role in treefall gap succession at San Carlos. Under closed forest, advance regeneration has a mean annual survivorship of - 80%; height growth is only a few centimetres a year, and leaf retention times frequently exceed 4 yr. Four years after gap formation, advance regeneration accounted for 97% of all trees ?-1 m tall in the single-treefall gaps and 83% of all trees in the multiple-treefall gap. Almost all trees in both gap types were of primary forest species; pioneer trees comprised only a small fraction of the regrowth. In general, microhabitat within gaps did not influence plant density, or plant estab- lishment and mortality patterns. However, mortality was much higher for individuals that germinated after gap formation than for individuals present as advance regeneration. Plant growth within treefall gaps was influenced by gap size, gap microhabitat type, gap age, and plant size. Seedlings and saplings of forest trees in the multiple-treefall gap grew three times as fast as those in the single-treefall gaps. Within single-treefall gaps, height growth was greater in the trunk and open zones than in the crown zone, and trees generally grew more slowly as gaps aged. Furthermore, tree growth was positively correlated with tree size, causing the size differential between trees of different heights to expand as gaps aged. At San Carlos, tree seedlings do not grow to the canopy during a single gap event; rather, canopy closure occurs by growth of larger pole-sized trees that survived treefalls or by lateral expansion of bordering subcanopy trees. Our results indicate that neither gap size, microhabitat within gaps, nor gap age have measurable effects on nutrient loss, nor do they appear to affect plant density, plant estab- lishment, or plant mortality. Size, microhabitat, and temporal affects are minimized, in large part, because of the great importance of advance regeneration in gap succession.

SUCCESSION AND NUTRIENT DYNAMICS FOLLOWING FOREST CUTTING AND BURNING IN AMAZONIA

Plant establishment, plant mortality, vegetation productivity, nutrient accumulation, and nutrient leaching were studied during the first 5 yr of succession following the cutting and burning of a forest plot in the upper Rio Negro region of the Amazon Basin. The speed and vigor of forest regeneration following disturbance of this type provide a measure of the recuperative potential of this ecosystem. The study plot (0.09 ha) was dominated by forbs during the 1 st yr and then by the pioneer tree species Cecropiaficifolia Snethlage during year two. Tree mortality exceeded establishment during the 3rd yr because the C. ficifolia trees died nearly in unison. During the 4th yr, tree density increased sharply as successional and forest tree species grew in the space vacated by C. ficifolia. The canopy was dominated at that time by the pioneer species Vismia japurensis Reich. and Vismia lauriformis (Lam.) Choisy. By the 5th yr, all establishment space had apparently been preempted: there were only 0.07 new establishments/M2 during year five, while 95% of the trees present at the end of year four survived through the 5th yr. There were 56 tree species (? 2 m tall) present on the site after 5 yr. More than half of these were primary forest species. Primary forest species were most common in the understory. These trees averaged > 1 m in growth height each year and should eventually grow into the canopy. Hence, it appears that most individuals that will participate in the aggrading phase of forest development were already present on the site within 5 yr of burning. Biomass was only 66 g/m2 at the end of the 1 st yr, but g1000 g/m2 were added in each subsequent year. The standing crop of biomass (shoot and root) at 5 yr was 4840 g/m2, a value 16% that of the mature forest which occupied the site prior to cutting. Approximately 27% of the dead trunk and branch biomass from the pre-existing forest remained on the site after 5 yr. Litter production increased each year from 39 g/m2 (year one) to 825 g/m2 (year five) and tended to be inversely related to rainfall. Total aboveground production was highest for the 5th yr (1940 g/m2) and was probably close to a maximum for the site. Total live-plant nutrient stocks after 5 yr, as a percentage of the precut forest stocks, were 15% for N, 23% for P, 39% for K, 48% for Ca, and 45% for Mg. A large increase in the leachate concentrations of K, Mg, and NO3 occurred during the first 2 yr following burning. By the 5th yr, concentrations were not significantly different from those in the undisturbed forest. After 5 yr soil nutrient levels were also similar to those of the preburn forest. Overall, nutrient losses apparently were not great enough to destroy the sites ability to return eventually to a state similar to that which existed prior to disturbance. This does not mean that a more severe treatment, such as conversion of forest to pasture, would not affect the regrowth ability of a site.

Photosynthesis-nitrogen relations in Amazonian tree species - I. Patterns among species and communities

Among species, photosynthetic capacity (Amax) is usually related to leaf nitrogen content (N), but variation in the species-specific relationship is not well understood. To address this issue, we studied Amax-N relationships in 23 species in adjacent Amazonian communities differentially limited by nitrogen (N), phosphorus (P), and/or other mineral nutrients. Five species were studied in each of three late successional forest types (Tierra Firme, Caatinga and Bana) and eight species were studied on disturbed sites (cultivated and early secondary successional Tierra Firme plots). Amax expressed on a mass basis (Amass) was correlated (p<0.05) with Nmass in 17 of 23 species, and Amax on an area basis (Aarea) was correlated (p<0.05) with Narea in 21 of 23 species. The slopes of Amax-N relationships were greater and intercepts lower for disturbance adapted early successional species than for late successional species. On a mass basis, the Amax-N slope averaged ≈15 μmol CO2 [g N]-1 s-1 for 7 early secondary successional species and ≈4 μmol CO2 [g N]-1 s-1 for 15 late successional species, respectively. Species from disturbed sites had shorter leaf life-span and greater specific leaf area (SLA) than late successional species. Across all 23 species, the slope of the Amass-Nmass relationship was related (p<0.001) positively to SLA (r2=0.70) and negatively to leaf life-span (r2=0.78) and temporal niche during secondary succession (years since cutting-and-burning, r2=0.90). Thus, disturbance adapted early successional species display a set of traits (short leaf life-span, high SLA and Amax and a steep slope of Amax-N) conducive to resource acquisition and rapid growth in their high resource regeneration niches. The significance and form of the Amax-N relationship were associated with the relative nutrient limitations in the three late successional communities. At species and community levels, Amax was more highly dependent on N in the N-limited Caatinga than in the P-and N-limited Bana and least in the P-and Ca-limited Tierra Firme on oxisol-and differences among these three communities in their massbased Amax-N slope reflects this pattern (6.0, 2.4, and 0.7 μmol CO2 [g N]-1 s-1, respectively). Among all 23 species, the estimated leaf Nmass needed to reach compensation (net photosynthesis ≈ zero) was positively related to the Amass-Nmass slope and to dark respiration rates and negatively related to leaf life-span. Variation among species in the Amax-N slope was well correlated with potential photosynthetic N use efficiency, Amax per unit leaf N. The dependence of Amax on N and the form of the relationship vary among Amazonian species and communities, consistent with both relative availabilities of N, P, and other mineral nutrients, and with intrinsic ecophysiological characteristics of species adapted to habitats of varying resource availability.

Logging damage during planned and unplanned logging operations in the eastern Amazon

Abstract In the Paragominas region of eastern Amazonia, we compared the damage in unplanned and planned logging operations associated with each of five logging phases: (1) tree felling, (2) machine maneuvering to attach felled boles to chokers, (3) skidding boles to log landings, (4) constructing log landings and (5) constructing logging roads. Planned logging was conducted with a rubber-wheeled skidder or a bulldozer equipped with a winch, while unplanned logging was conducted with a bulldozer not equipped with a winch. We found, for each commercial tree felled, unplanned logging methods damaged 16 more trees ≥ 10 cm dbh and affected a ground area that was more than 100 m2 greater than in planned operations. In the tree felling phase, the number of trees experiencing severe crown damage (i.e. topped) was significantly greater (Students t-test; p Per hectare logged, we found that tree felling damaged more trees ≥ 10 cm dbh than the other logging phases, with unplanned felling damaging nearly twice as mary trees per hectare as planned felling (124 vs. 64). Furthermore, the unplanned area had a lower percentage of logging gaps containing one tree in them and a higher percentage of gaps containing two, three and four felled trees. The mean area of all gaps considered together was significantly larger in the unplanned operation (Studens t-test, p We estimate that the profit margins of companies that switched to planned logging methods would actually increase. The costs associated with planned logging are offset by the benefits of this approach, namely reduced machine operating time and labor per m3 of timber extracted and less waste. Furthermore, by implementing these techniques, more than 80 trees ≥ 10 cm dbh per hectare were spared damage during logging operations. By examining the way in which individual trees were damaged and in which the ground area was disturbed in five distinet logging phases, we are able to recommend planned logging measures that can reduce overall damage to the remaining stand by 25–33% and, thus, improve the likelihood of managing these tropical forests on a 30–40 year cutting cycle

EARLY PLANT SUCCESSION AFTER CUTTING AND BURNING IN THE UPPER RIO NEGRO REGION OF THE AMAZON BASIN

(1) Changes in vegetation structure, species composition, and species dominance were studied over the first twenty-two months of succession after the cutting and burning of several mature tierra firme (i.e. non-flooded) forest sites in south Venezuela in the upper Rio Negro region of the Amazon Basin. To explain the observed succession, additional field investigations were undertaken. (2) Four months after burning the density of plants was less than 1 m-2. The density of successional woody plants, forbs, and grasses increased rapidly from the fourth to the tenth month of succession. There was little change in plant density between the tenth and twenty-second month, but a large change in vegetation height occurred during this time. By the twenty-second month a loose canopy of Cecropia spp. was present at 5 m height. (3) Many of the forest tree species sprouted from the stump when cut, but burning killed sprouts and significantly reduced the size of the seed bank. Nevertheless, the first successional woody colonizers established from seed which survived the burn. The first forb and grass colonizers established from seeds dispersed onto the site after the burn. (4) The common successional woody species had bator bird-dispersed seeds. Many forbs and grasses had wind-dispersed seeds. (5) Population densities were low during the first months of succession because of the low density of propagules and because the low rainfall during this time was not favourable for seedling germination or establishment. The rapid increase in plant density after about the fourth month occurred because the early colonizers had then begun to produce, and disperse locally, large numbers of seeds. (6) There were six distinct microhabitat types present on the sites after burning. Most species tested showed distinct microhabitat preferences.

Amazon Rain-Forest Fires

Charcoal is common in the soils of mature rain forests within 75 kilometers of San Carlos de Rio Negro in the north central Amazon Basin. Carbon-14 dates of soil charcoal from this region indicate that numerous fires have occurred since the mid-Holocene epoch. Charcoal is most common in tierra firme forest Oxisols and Ultisols and less common in caatinga and igapo forest soils. Climatic changes or human activities, or both, have caused rain-forest fires.

FIRE IN AMAZONIAN SELECTIVELY LOGGED RAIN FOREST AND THE POTENTIAL FOR FIRE REDUCTION

Approximately 8000 km2 of Brazilian Amazon forest are selectively logged each year. Although virgin forest in eastern Amazonia is generally immune to fire, selectively logged forests are susceptible to fire. In eastern Amazonia we employed permanent-plot studies, forest fuel moisture measurements, and hemispheric canopy photographs to study the impacts of fire on a selectively logged forest, the microclimatic conditions that foster forest fires, and the measures that loggers might take to reduce fire incidence. Significant tree mortality followed a typical ground fire in a selectively logged forest. Forty-four percent of all trees ≥10 cm in diameter at breast height died in a burned plot while only 3% died in an unburned plot. In large logging gaps the density of regenerating pioneer species increased by >60% in burned plots 15 mo after the fire, while it decreased by >40% in unburned plots. The rate of fuel drying in selectively logged forest was influenced by photon flux density (PFD), time since loggin...

Abandoned Pastures in Eastern Amazonia. II. Nutrient Stocks in the Soil and Vegetation

(1) Soil nutrient content was studied in relation to vegetation development on thirteen abandoned pastures differing in age from two to eight years and in intensity of pasture use from light to heavy. The surface soils of these sites had higher cation concentration and pH than those of undisturbed mature forest of the area, but similar phosphorus, organic matter and total nitrogen concentrations. (2) Soil nutrient concentration generally was independent of age or prior intensity of use in the pastures. Pasture use has a significant effect on soil nutrient concentration when the sites are first abandoned, but this does not persist. Nutrient uptake by successional vegetation may substantially reduce soil nutrient stocks. (3) Soil nutrient concentration was not significantly correlated with vegetation biomass, primary tree biomass or species richness, either among or within sites. (4) Nutrient depletion of pastures abandoned eight years previously relative to total nutrient stocks in mature forest depended on pasture disturbance. Sites previously subjected to low use were only depleted of N, moderate-use sites were depleted of N and K, and heavy-use sites were depleted of N, K, Mg and possibly Ca and P. There was less nutrient depletion than the relatively small biomass of the successional sites would indicate. This was due to a higher proportion of nutrient-rich leaves and fine branches, enriched soil nutrient stocks relative to mature forest and, on the lightly disturbed sites, a large pool of nutrients in unburned woody residue. (5) Decomposition of unburned woody residue may provide one-half of the nutrients taken up by vegetation during eight years of recovery following light pasture use, but only 15% following moderate use, and none following the most intense use. (6) General schemes of succession following deforestation for pasture development are presented. Recovery after disturbance is complex, and follows different patterns depending on the intensity of pasture use.

Logging impacts and prospects for sustainable forest management in an old Amazonian frontier: The case of Paragominas

The Brazilian wood industry is highly mobile. Over the past 20 years the eastern Amazon developed from a logging backwater to the principal hardwood processing center in Brazil. This occurred because of a decline in hardwood stocks in the south of Brazil coupled with the development of good transport, energy, and communications systems in eastern Amazonia. We studied the structure and economy of the wood industry along a 340 km stretch of the Belem-Brasilia Highway in eastern Amazonia. Of the 238 sawmills present in this study region in late 1989, 79% were installed in the 1980s. Ninety-seven percent of the mill owners came from outside Amazonia. Most (63%) of the mill establishments were vertically integrated, engaging in both forest mill processing and forest logging. Logging establishments realize substantial profits. A typical sawmill with one band saw produces, on average, 4300 m3 of sawnwood year−1 from 9200 m3 of roundwood. The value of this sawn production is estimated at

Ecological Impacts of Selective Logging in the Brazilian Amazon: A Case Study from the Paragominas Region of the State of Para

670 800 or