Search

Population structure (size, age, spatial patterns) and radial growth patterns are used to analyze regeneration patterns of Abies faxoniana, Betula albosinensis, Betula utilis, Larix potaninii, Picea purpurea, and Sabina saltuaria and reconstruct disturbance history in 8 subalpine forest stands in Wang Lang Natural Reserve, Sichuan, China. In old-growth stands tree regeneration occurs in tree-fall gaps whereby A. faxoniana, Betula sp., P. purpurea, and S. saltuaria persist at stand scales by gap-phase regeneration. Clump sizes of young populations are similar to canopy gap sizes but clumps sizes vary among species. Young Betula patches are larger than those of A. faxoniana suggesting that gap-partitioning by size contributes to species coexistence in mixed stands. Picea purpurea and S. saltuaria are longer lived than A. faxoniana which may compensate for lower recruitment and prevent their replacement by A. faxoniana. Tree regeneration and community structure are also influenced by the understory bamboo Fargesia denudata. Seedlings, saplings, and shrub density all decline with an increase in bamboo cover. Species that regenerate in old-growth forest also regenerate after flooding as do species that establish only on bare substrates (i.e. Larix potaninii, Prunus sp.). Structural and compositional patterns in Wang Lang forests are a reflection of disturbance history, canopy species life history attributes such as dispersal ability, shade tolerance, growth rates, and longevity, and competition of trees and shrubs with understory bamboos.

Culm (ramet) demography of two bamboos, Fargesia robusta and Bashaniafangiana, was studied over 7 yr in southwest China to determine if: 1) ramet numbers remain relatively constant; 2) ramet mortality was constant; 3) density response was greater in spreading vs. compact clones; 4) extrinsic factors (e.g., predation, climate) influenced ramet dynamics; and 5) ramets that die are smaller than ones that live. Culm numbers remained relatively constant (?20%) in most plots, but culm numbers increased (> 20%) on some sites. Mortality of culms > 6 mo old over 7 yr was low and constant, but mortality of younger culms (

Giant panda Ailuropoda melanoleuca are obligate bamboo grazers that inhabit montane and subalpine forests in south-western China. Knowledge of factors influencing bamboo abundance is crucial for panda ecology and identifying how human activities change panda habitat. This study follows bamboo regeneration from seed over seven years in panda habitat in clearcut and closed forest. Bamboo seedling density was significantly lower in clearcuts in 1984, 1985, and 1990 than beneath forest, and few seedlings remain in clearcuts. It appears that bamboo will not replace itself in clearcuts, reducing the extent of bamboo forest for pandas. Beneath forest, bamboo seedlings take 15–20 years to grow back to full height.

Gaps created by disturbance in the forest canopy are important sites for tree regeneration from seed but plants already established in gaps may slow gap-filling. This study deals with consequences of bamboo die-back for tree regeneration and the dynamics of Abies-Betula forests in southwest China. Bamboo dominates the forest understory impeding tree regeneration when in its vegetative phase. Populations of tree seedlings were sampled in 1984–85 and 1990 in two sets of permanent plots where bamboo had died back in 1983. Both Abies and Betula density increased after bamboo die-back, Betula more so than Abies, especially in gaps. Before bamboo die-back, seedlings were established on raised surfaces such as logs but afterwards seedlings became common on the forest floor. This reduced the intensity of clumping of seedling populations between 1984 and 1990. A tree by tree replacement model predicts an increase in Abies and a decrease in Betula after bamboo die-back. Life histories of tree species, gap characteristics, and the bamboo growth cycle (mature/die-back/building) interact to promote fluctuating dominance of Abies and Betula in old-growth forests.

BY DONALD G. REID*?, ALAN H. TAYLORt, HU JINCHUI AND QIN ZISHENG* * Department of Biological Sciences, University of Calgary, Calgary, Alberta T2N 1N4, Canada and WWF-International, CH-1196 Gland, Switzerland; t Department of Geography, Pennsylvania State University, University Park, PA 16802, U.S.A.; and t Department of Biology, Nanchong Normal College, Nanchong, Sichuan, People's Republic of China

TAYLOR, A. H. (Department of Geography, The Pennsylvania State University, University Park, PA 16802 USA), D. G. REID (Department of Zoology, University of British Columbia, Vancouver, BC, Canada V6T 2A9), QIN ZISHENG AND Hu JINcHU (Department of Biology, Nanchong Teachers College, Nanchong, Sichuan, People's Republic of China). Spatial patterns and environmental associates of live bamboo (Bashania fangiana Yi) after mass-flowering in southwestern China. Bull. Torrey Bot. Club 118: 247-254. 1991.-The spatial pattern of live patches of Bashaniafangiana after most stands flowered in 1983 in Wolong Natural Reserve, China were analyzed. Live patches were intensely clumped at small spatial scales and remained clumped up to scales of 14.4 km2. Discriminant function analysis of 281 randomly selected points with live or dead bamboo described by six environmental variables identified elevation and percentage forest cover as the most important variables separating sites with live and dead bamboo. Live bamboo was associated with sites >3200 m and with clearcut forest. These data suggest that stress at high altitudes and in open habitats retarded flowering by altering resource allocation patterns within clones. The life history of B. fangiana is also compared to expected life history traits of bamboos responding to evolutionary mechanisms that may promote massflowering in bamboos.

Giant pandas Ailuropoda melanoleuca David inhabit cool temperate, montane, and subalpine conifer forests with a bamboo understorey in Sichuan, Gansu, and Shaanxi provinces, China. Clearcutting of subalpine conifer forest in panda habitat stimulates vegetative growth of bamboos. This impedes tree regeneration and hardwoods dominate sites after clearcutting, with conifers rarely becoming established. This study assesses the effects of selective cutting on patterns of forest regeneration through analyses of the age and size structure of tree populations in selectively cut and uncut forest. Conifer regeneration after selective cutting was adequate to produce a forest similar in structure and composition to uncut subalpine conifer forest. A selective cutting regime in production forest with pandas would strike a balance between panda conservation and needs for wood products, and would contribute to management initiatives aimed at preserving existing panda populations outside the protection of natural reserves.

Vegetation in canopy gaps of two old-growth Abies-Betula forest stands, one with bamboo the other without, was measured. The structure of gap vegetation at each site was used to derive tree replacement probabilities. Transition probabilities indicate different tree replacement trends in forests with bamboo compared to those without. Projected compositions show Betula to be the most abundant species in bamboo stands while Abies remains most abundant where bamboo was absent. A dense bamboo sward seems to reduce the probability of Abies filling gaps by inhibiting establishment and growth of seedlings. Bamboo preempts space after canopy gap formation by increasing shoot production which reduces opportunities for establishment and growth of other woody species. Differences in dispersal ability and longevity of Abies and Betula appear to be important factors contributing to their coexistence forests with a small canopy gap disturbance regime.

(1) Regeneration patterns of Abiesfaxoniana, Betula utilis and B. albosinensis in central Sichuan, China were interpreted by stand structural analysis (age, size, horizontal patterning) of tree populations in old-growth ( > 250 years) stands. (2) Seedlings, saplings and young trees were scarce in stands where bamboo (Sinarundinaria fangiana) impeded tree regeneration. Shrub density and woody plant species richness were lower in stands with high bamboo cover. (3) Regeneration patterns indicate that A. faxoniana and B. utilis relative densities were similar where bamboo was dense and that A. faxoniana was more than twice as abundant as B. utilis where bamboo cover was low or absent. Bamboos seem to impede A.faxoniana regeneration more than that of B. utilis. Both species regenerated in canopy gaps, but A. faxoniana less so than B. utilis where bamboos were dense. This produces higher Betula density in bamboo stands. Sizes of young patches of B. utilis tended to be larger than patches of A.faxoniana. Partitioning of gaps by size may contribute to coexistence of A. faxoniana and B. utilis in old-growth stands.

(1) Bamboos dominate the understorey of montane and subalpine forests in central Sichuan, China where they are the sole food source for giant pandas. (2) Standing crop biomass and net production were estimated for three species Fargesia spathacea Franchet., Sinarundinariafangiana (A. Camus) Keng ex Keng f. and Fargesia scabrida Yi. (3) Above-ground biomass averaged 23 7 t ha-' for F. spathacea, 9 3 t ha-' for F. scabrida and 7 4 t ha-' for S. fangiana. (4) Average annual net production of F. spathacea and S. fangiana was 3 6 t haand 1 5 t ha-', respectively. (5) Insects, giant pandas, rodents and other predators killed 46%, of all F. spathacea shoots and 45%, of all F. scabrida shoots. (6) Giant pandas selected large shoots and insects small ones of both F. spathacea and F. scabrida favouring development of stands composed of intermediate sized culms. (7) Average annual culm recruitment for F. scabrida, S. fingiana and F. spatIhacea were 13 7%,, 10 0%,, and 8 2%Y,, respectively. (8) Average annual mortality rates for F. scabrida, S. fiangiana and F. spathacea were 8 5%,, 9 1%,, and 10 6%,, respectively. (9) Sinarundinaria fangiana flowered and died in 1983. Culm selection by pandas in remaining live patches is expected to change to use in proportion to availability. Pandas will shift their diet to F. spathacea if remaining patches of S. facngiana flower and die. (10) The most effective means of preventing food shortages for pandas during periods of bamboo flowering is by preserving the existing mosaic of bamboo species.