Decomposition rates ofleaflitter have been predicted from the leaves' lignin or nutrient (N or P) contents, the C:N ratio, and more recently the lignin:N ratio. But tests of these predictors have been based on groups of substrates each spanning only part of the natural range oflignin contents, and neglecting low-lignin ( < 10%) species. We allowed leaf litter from eight species of tree, shrub, or herb, ranging in lignin content from 3.4 to 20.5%, to decompose in laboratory microcosms for up to 4 mo (equivalent to 1.5-2 yr decay in the field) to test two hypotheses: (l) that the lignin: nitrogen ratio would have a better correlation with mass loss rates than would the C:N ratio, nutrient content, or other substrate quality indexes, and (2) that correlations of mass loss with initial N content would decrease, while correlations with lignin content would increase, as decay proceeded.Contrary to the first hypothesis, nitrogen content and the C:N ratio were the best predictors of mass loss rate, and were substantially better than the lignin:N ratio. We could find no better predictor of decomposition rate than the C:N ratio, and no better regression model than the simple linear one. However, when regressions were tested using pine needles (lignin content 26.2%), the C:N ratio and N content badly underestimated mass remaining (by 10-16%), while lignin content and the lignin:N ratio overestimated it by <2%.In accordance with the second hypothesis, regressions of initial lignin content or lignin: N ratio on mass remaining improved (higher R 2 ) from 2 to 4 mo decomposition, while those of N content grew worse, illustrating succession of nitrogen to lignin control of decomposition rate. Reported correlations of the lignin:N ratio with decomposition rate for some litter types arise as a special case of this two-phase mechanism of control by nutrients and lignin. For substrates low in lignin, or where a broad range oflignin contents is being considered, the C:N ratio is a better predictor of decomposition rate than the lignin:N ratio.

The effects of converting lowland tropical rainforest to pasture, and of subsequent succession of pasture lands to secondary forest, were examined in the Atlantic Zone of Costa Rica. Three replicate sites of each of four land—use types representing this disturbance—recovery sequence were sampled for changes in vegetation, pedological properties, and potential nitrogen mineralization and nitrification. The four land—use types included primary forest, actively grazed pasture (10—36 yr old), abandoned pasture (abandoned 4—10 yr) and secondary forest (abandoned 10—20 yr). Conversion and succession had obvious and significant effects on canopy cover, canopy height, species composition, and species richness; it appeared that succession of secondary forests was proceeding toward a floristic composition like that of the primary forests. Significant changes in soil properties associated with conversion of forest to pasture included: (1) a decrease in acidity and increase in some base exchange properties, (2) an increase in bulk density and a concomitant decrease in porosity, (3) higher concentrations of NH4+, (4) lower concentrations of NO3—, (5) lower rates of N—mineralization, and (6) in some cases, lower rates of nitrification. Chemical changes involving cations associated with conversion from forest to pasture indicated increases in soil fertility under the pasture regimes, while changes associated with nitrogen indicated decreases in fertility. Physical changes in density and porosity were deleterious with respect to infiltration, percolation, aeration, and, ultimately, erodability. Beyond the practical aspects of land management, many of these changes are very important to carbon and nitrogen cycling and to the emission and consumption of biogenic trace gases.

Summary1. The impacts of elevated atmospheric CO 2 and/or O 3 have been examined over 4 years using an open-air exposure system in an aggrading northern temperate forest containing two different functional groups (the indeterminate, pioneer, O 3 -sensitive species Trembling Aspen, Populus tremuloides and Paper Birch, Betula papyrifera , and the determinate, late successional, O 3 -tolerant species Sugar Maple, Acer saccharum ). 2. The responses to these interacting greenhouse gases have been remarkably consistent in pure Aspen stands and in mixed Aspen/Birch and Aspen/Maple stands, from leaf to ecosystem level, for O 3 -tolerant as well as O 3 -sensitive genotypes and across various trophic levels. These two gases act in opposing ways, and even at low concentrations (1·5 × ambient, with ambient averaging 34 -36 nL L − 1 during the summer daylight hours), O 3 offsets or moderates the responses induced by elevated CO 2 . 3. After 3 years of exposure to 560 µ mol mol − 1 CO 2 , the above-ground volume of Aspen stands was 40% above those grown at ambient CO 2 , and there was no indication of a diminishing growth trend. In contrast, O 3 at 1·5 × ambient completely offset the growth enhancement by CO 2 , both for O 3 -sensitive and O 3 -tolerant clones. Implications of this finding for carbon sequestration, plantations to reduce excess CO 2 , and global models of forest productivity and climate change are presented.

Belowground responses to aboveground disturbance were studied in experimental gaps created in a 95—yr—old stand of Pinus contorta in southeastern Wyoming. One—, 5—, 15—, and 30—tree clusters were felled to create a series of gaps in the root mat, and solution—phase N was monitored over two consecutive snow—melt periods via tension—tube water collectors. We hypothesized that dissolved and extractable nitrogen concentrations would not exceed predisturbance levels until a threshold canopy gap size had been achieved. As predicted, NOx—N attained significantly higher solution N concentrations (2—5 mg/L) only with the death of 15 trees or more. However, dissolved organic nitrogen decreased gradually with increasing gap size. Net mineralization and nitrification were studied using 30—d in situ incubation assays in each gap. Extractable nitrate routinely was negligible until the 30—tree gaps had been attained. Predicting the effects of disturbance on nutrient cycling, including timber—harvesting practices, requires information on belowground responses to gap formation. Our experiments suggest that gap size is important; removal of 15—30 tree clusters represented a threshold above which significant losses of available N to the groundwater may be incurred, at least in Rocky Mountain coniferous forests.

No abstract

No abstract

Summary1 Environmental variability can occur over various spatial scales, ranging from small patches at the scale of individual plants to long gradients over hundreds of metres. 2 In the New Jersey Pinelands, different species in the diverse shrub understorey of pitch pine (Pinus rigida Mill.) forests are patterned at these various scales. 3 Soil moisture, extractable NH 4 -N and N mineralization rate vary in complex ways, with the scale of spatial patterning changing over time and with depth in the soil profile. Moisture in both mineral and organic horizons, and NH 4 -N in the organic horizon, have patterns that are more stable over time than the mineralization rate in either horizon, or the NH 4 -N concentrations in the mineral horizon. 4 Vegetation patterns, as captured in principal components analysis, were poorly explained by any of the soil properties. Only the more temporally stable properties showed any relationship with vegetation patterns. 5 These results suggest that environmental gradients reflect patterns of environmental variation in four dimensions. Variation in the vertical dimension and over time is as pronounced and important as variation in the horizontal dimensions. 6 Many methods used to analyse vegetation implicitly assume temporal and spatial stability of environmental properties. Our results suggest that a more complex, fourdimensional assessment of environmental variation should be incorporated into models of vegetation-environment relationships.

Decomposition of a slow-decaying litter type is expected to be faster in the presence of a nutrient-rich, fast-decaying litter type, but this effect has never been conclusively demonstrated for deciduous leaves. In a Rocky Mountain aspen forest, we followed decomposition of leaf litter of trembling aspen (Populustremuloides), a relatively slow-decomposing, nutrient-poor species, and green alder (Alnuscrispa), a nutrient-rich, faster-decomposing species, as well as a mixture of the two, for 2 years. Mass losses over the first winter were greatest for aspen alone, probably as a result of loss of solubles, but the mass loss rate overall was lowest for aspen (k = −0.191/year) and greatest for alder (k = −0.251/year). Mass loss rate for mixed litter (k = −0.245/year) was much closer to the rate for alder than for aspen, demonstrating a marked acceleration of mass loss rates in the mixed-litter bags. At these rates, 95% mass loss would be achieved by aspen, alder, and mixed litter in 14.5, 11.5, and 11.6 years, respectively.