The study of plant litter decomposition in terrestrial ecosystems commonly employs litter bags to compare the loss of mass among species, among sites, and under various experimental manipulations, or to investigate the process itself. Analysis of the resulting data is quite variable among investigators, and at times inappropriate. Two general analytical approaches to the examination of decomposition data are reviewed. Analysis of variance is useful if the intent is to compare treatment means, but does not directly test hypotheses regarding decomposition rates. If the intent is to determine rate constants, than fitting mathematical models to data is the more appropriate analysis. Single and double exponential models best describe the loss of mass over time with an element of biological realism.

Decay of balsam fir (Abies balsamea) boles was examined in an upper subalpine forest of the White Mountains, New Hampshire, USA. Fifty percent of the initial mass was los in 23 yr; 90% was lost in 77 yr. High decay rates were attributed to the small diameters of the boles, ample moisture, and a nitrogen—rich environment. Average dead woods mass in this forest was 4.9 kg/m2, representing 25% of the sum of dead wood, live plant biomass, and forest floor organic matter. Changes in density and moisture and in the concentrations and content of various chemical components of the boles were traced over the decay sequence. Changes in the content of cellulose, lignin, carbon and sodium followed loss of mass during decay. Contents of calcium, magnesium, potassium and phosphorus decreased faster than loss of mass in the early stages of decay. Much of this initial loss was ascribed to sloughing of nutrient—rich bark which in these small boles comprised 13% of dry mass. Later in decay, the loss rates of calcium, magnesium and potassium were about the same or slightly less than the loss rate of mass. After a steep initial drop, phosphorus content of the boles remained approximately constant between years 12 and 33. Thereafter the loss rate paralleled loss of mass. Nitrogen content was approximately constant in the first 33 yr after which it declined in parallel with loss of mass.

All trees @>2.5 cm dbh were censused on a 1.5—ha tract of 60—yr—old tropical moist forest in 1968 and again in 1978 to determine rates of tree mortality, recruitment, dbh increment, and canopy gap formation. Species composition changed very little. The pioneer of gap species Cordia alliodora, Luehea seemanii, and Spondias radlkoferi had no recruitment and accounted for most mortality in the larger size classes. Ninety percent of all mortality was for stems <10 cm dbh. Total tree density decline 11% (from 3112 to 2781 trees/ha), but basal area increased 22% (from 25.7 to 31.4 m2/ha). Growth in diameter was highly variable, both among species and among size classes. Trees in the 30—50 cm dbh class had a mean dbh increment of 0.9 cm/yr. Gaps occurred over an area equal to 7.3% of the plot during the 10—yr period, suggesting that about 137 yr would be required for the 1.5—ha plot to be affected by tree falls.

Vegetation structure and dynamics of the upper subalpine or fir zone were studied in the White Mountains of New Hampshire. The fir zone extends from 1220 m, the approximate upper limit of Picea rubens occurrence, to treeline (°1450 m) where it is usually represented by a low krummholz. Live tree density in the part of the zone in which trees are >2 m tall averages 5000 stems/ha. Eighty—four percent of these are Abies balsamea, and most of the rest are Betula papyrifera var. cordifolia. Basal area averages 30 m2/ha; canopy height averages 6.8 m. Stands tend to be dominated by discrete age classes. Tree ages range up to 111 yr (average 55 yr). Fir zone vegetation is subject to a number of processes that lead to a hierarchical set of overlapping patterns. Elevation and wind exposure produce general, or first—order, patterns over the landscape, e.g., canopy height decreases with elevation and exposure. Overlying this general pattern are a series of second—order patterns. Two acute disturbance factors–hurricanes and avalanches–have stamped discrete impressions over the general pattern. Other, more endogenously generated patterns, marks the landscape with more subtle textures. These patterns include fir waves, broken gaps, strips, and glades. A conceptualization of vegetation as a hierarchical series of overlapping patterns is an extension of Watt's view of vegetation dynamics and bears important ecological implications.

No abstract