Buttress form of the central African rain forest tree Microberlinia bisulcata, and its possible role in nutrient acquisition

Buttressing is a trait special to tropical trees but explanations for its occurrence remain inconclusive. The two main hypotheses are that they provide structural support and/or promote nutrient acquisition. Studies of the first are common but the second has received much less attention. Architectural measurements were made on adult and juvenile trees of the ectomycorrhizal species Microberlinia bisulcata, in Korup (Cameroon). Buttressing on this species is highly distinctive with strong lateral extension of surface roots of the juveniles leading to a mature buttress system of a shallow spreading form on adults. This contrasts with more vertical buttresses, closer to the stem, found on many other tropical tree species. No clear relationship between main buttress and large branch distribution was found. Whilst this does not argue against the essential structural role of buttresses for these very large tropical trees, the form on M. bisulcata does suggest a likely second role, that of aiding nutrient acquisition. At the Korup site, with its deep sandy soils of very low phosphorus status, and where most nutrient cycling takes place in a thin surface layer of fine roots and mycorrhizas, it appears that buttress form could develop from soil-surface root exploration for nutrients by juvenile trees. It may accordingly allow M. bisulcata to attain the higher greater competitive ability, faster growth rate, and maximum tree size that it does compared with other co-occurring tree species. For sites across the tropics in general, the degree of shallowness and spatial extension of buttresses of the dominant species is hypothesized to increase with decreasing nutrient availability.

Plurality of tree species responses to drought perturbation in Bornean tropical rain forest

Drought perturbation driven by the El Niño Southern Oscillation (ENSO) is a principal stochastic variable determining the dynamics of lowland rain forest in S.E. Asia. Mortality, recruitment and stem growth rates at Danum in Sabah (Malaysian Borneo) were recorded in two 4-ha plots (trees ≥ 10 cm gbh) for two periods, 1986–1996 and 1996–2001. Mortality and growth were also recorded in a sample of subplots for small trees (10 to <50 cm gbh) in two sub-periods, 1996–1999 and 1999–2001. Dynamics variables were employed to build indices of drought response for each of the 34 most abundant plot-level species (22 at the subplot level), these being interval-weighted percentage changes between periods and sub-periods. A significant yet complex effect of the strong 1997/1998 drought at the forest community level was shown by randomization procedures followed by multiple hypothesis testing. Despite a general resistance of the forest to drought, large and significant differences in short-term responses were apparent for several species. Using a diagrammatic form of stability analysis, different species showed immediate or lagged effects, high or low degrees of resilience or even oscillatory dynamics. In the context of the local topographic gradient, species’ responses define the newly termed perturbation response niche. The largest responses, particularly for recruitment and growth, were among the small trees, many of which are members of understorey taxa. The results bring with them a novel approach to understanding community dynamics: the kaleidoscopic complexity of idiosyncratic responses to stochastic perturbations suggests that plurality, rather than neutrality, of responses may be essential to understanding these tropical forests. The basis to the various responses lies with the mechanisms of tree-soil water relations which are physiologically predictable: the timing and intensity of the next drought, however, is not. To date, environmental stochasticity has been insufficiently incorporated into models of tropical forest dynamics, a step that might considerably improve the reality of theories about these globally important ecosystems.

Transient dominance in a central African rain forest

The large-crowned emergent tree Microberlinia bisulcata dominates rain forest groves at Korup National Park, Cameroon, along with two codominants, Tetraberlinia bifoliolata and T. korupensis. M. bisulcata has a pronounced modal size frequency distribution around 110 cm stem diameter: its recruitment potential is very poor. It is a long-lived light-demanding species, one of many found in African forests. Tetraberlinia species lack modality, are more shade tolerant, and recruit better. All three species are ectomycorrhizal. M. bisulcata dominates grove basal area, even though it has similar numbers of trees (≥50 cm stem diameter) as each of the other two species. This situation presented a conundrum that prompted a long-term study of grove dynamics. Enumerations of two plots (82.5 and 56.25 ha) between 1990 and 2010 showed mortality and recruitment of M. bisulcata to be very low (both rates 0.2% per year) compared with Tetraberlinia (2.4% and 0.8% per year), and M. bisulcata grows twice as fast as the Tetraberlinia. Ordinations indicated that these three species determined community structure by their strong negative associations while other species showed almost none. Ranked species abundance curves fitted the Zipf-Mandelbrot model well and allowed “overdominance” of M. bisulcata to be estimated. Spatial analysis indicated strong repulsion by clusters of large (50 to <100 cm) and very large (≥100 cm) M. bisulcata of their own medium-sized (10 to <50 cm) trees and all sizes of Tetraberlinia. This was interpreted as competition by M. bisulcata increasing its dominance, but also inhibition of its own replacement potential. Stem coring showed a modal age of 200 years for M. bisulcata, but with large size variation (50–150 cm). Fifty-year model projections suggested little change in medium, decreases in large, and increases in very large trees of M. bisulcata, accompanied by overall decreases in medium and large trees of Tetraberlinia species. Realistically increasing very-large-tree mortality led to grove collapse without short-term replacement. M. bisulcata most likely depends on climatic events to rebuild its stands: the ratio of disturbance interval to median species' longevity is important. A new theory of transient dominance explains how M. bisulcata may be cycling in abundance over time and displaying nonequilibrium dynamics.

Relaxation of species-specific neighborhood effects in Bornean rain forest under climatic perturbation

Evidence of negative conspecific density dependence (NDD) operating on seedling survival and sapling recruitment has accumulated recently. In contrast, evidence of NDD operating on growth of trees has been circumstantial at best. Whether or not local NDD at the level of individual trees leads to NDD at the level of the community is still an open question. Moreover, whether and how perturbations interfere with these processes have rarely been investigated. We applied neighborhood models to permanent plot data from a Bornean dipterocarp forest censused over two 10-11 year periods. Although the first period was only lightly perturbed, a moderately strong El Nino event causing severe drought occurred in the first half of the second period. Such events are an important component of the environmental stochasticity affecting the region. We show that local NDD on growth of small-to-medium-sized trees may indeed translate to NDD at the level of the community. This interpretation is based on increasingly negative effects of bigger conspecific neighbors on absolute growth rates of individual trees with increasing basal area across the 18 most abundant overstory species in the first period. However, this relationship was much weaker in the second period. We interpreted this relaxation of local and community-level NDD as a consequence of increased light levels at the forest floor due to temporary leaf and twig loss of large trees in response to the drought event. Mitigation of NDD under climatic perturbation acts to decrease species richness, especially in forest overstory and therefore has an important role in determining species relative abundances at the site.

Herbivores equalize the seedling height growth of three dominant tree species in an African tropical rain forest

Determining the impact of insect herbivores on forest tree seedlings and saplings is difficult without experimentation in the field. Moreover, this impact may be heterogeneous in time and space because of seasonal rainfall and canopy disturbances, or ‘gaps’, which can influence both insect abundance and plant performance. In this study we used fine netting to individually protect seedlings of Microberlinia bisulcata, Tetraberlinia bifoliolata and Tetraberlinia korupensis trees (Fabaceae = Leguminosae) from insects in 41 paired gap-understorey locations across 80 ha of primary rain forest (Korup, Cameroon). For all species, growth in height and leaf numbers was negligible in the understorey, where M. bisulcata had the lowest survival after c. 2 years. In gaps, however, all species responded positively with pronounced above-ground growth across seasons. When exposed to herbivores their seedling height growth was similar, but in the absence of herbivores, M. bisulcata significantly outgrew both Tetraberlinia species and matched their leaf numbers. This result suggests that insect herbivores might play an important role in maintaining species coexistence by mitigating sapling abundance of the more palatable M. bisulcata, which in gaps was eaten the most severely. The higher ratio in static leaf damage of control-to-caged M. bisulcata seedlings in gaps than understorey locations was consistent with the Plant Vigour Hypothesis. This result, however, did not apply to either Tetraberlinia species. For M. bisulcata and T. korupensis, but not T. bifoliolata (the most shade-tolerant species), caging improved relative seedling survival in the understory locations compared to gaps, providing restricted support for the Limiting Resource Model. Approximately 2.25 years after treatments were removed, the caged seedlings were taller and had more leaves than controls in all three species, and the effect remained strongest for M. bisulcata. We conclude that in this community the impact of leaf herbivory on seedling growth in gaps is strong for the dominant M. bisulcata, which coupled to a very low shade-tolerance contributes to limiting its regeneration. However, because gaps are common to most forests, insect herbivores may be having impacts upon functionally similar tree species that are also characterized by low sapling recruitment much more widely than currently appreciated. An implication for the restoration and management of M. bisulcata populations in forests outside of Korup is that physical protection from herbivores of new seedlings where the canopy is opened by gaps, or by harvesting, should substantially increase its subcanopy regeneration, and thus, too, its opportunities for adult recruitment.

Herbivores differentially limit the seedling growth and sapling recruitment of two dominant rain forest trees

Resource heterogeneity may influence how plants are attacked and respond to consumers in multiple ways. Perhaps a better understanding of how this interaction might limit sapling recruitment in tree populations may be achieved by examining species’ functional responses to herbivores on a continuum of resource availability. Here, we experimentally reduced herbivore pressure on newly established seedlings of two dominant masting trees in 40 canopy gaps, across c. 80 ha of tropical rain forest in central Africa (Korup, Cameroon). Mesh cages were built to protect individual seedlings, and their leaf production and changes in height were followed for 22 months. With more light, herbivores increasingly prevented the less shade-tolerant Microberlinia bisulcata from growing as tall as it could and producing more leaves, indicating an undercompensation. The more shade-tolerant Tetraberlinia bifoliolata was much less affected by herbivores, showing instead near to full compensation for leaf numbers, and a negligible to weak impact of herbivores on its height growth. A stage-matrix model that compared control and caged populations lent evidence for a stronger impact of herbivores on the long-term population dynamics of M. bisulcata than T. bifoliolata. Our results suggest that insect herbivores can contribute to the local coexistence of two abundant tree species at Korup by disproportionately suppressing sapling recruitment of the faster-growing dominant via undercompensation across the light gradient created by canopy disturbances. The functional patterns we have documented here are consistent with current theory, and, because gap formations are integral to forest regeneration, they may be more widely applicable in other tropical forest communities. If so, the interaction between life-history and herbivore impact across light gradients may play a substantial role in tree species coexistence.

Evidence of species-specific neighborhood effects in the dipterocarpaceae of a Bornean rain forest

Although accumulating evidence indicates that local intraspecific density-dependent effects are not as rare in species-rich communities as previously suspected, there are still very few detailed and systematic neighborhood analyses of species-rich communities. Here, we provide such an analysis with the overall goal of quantifying the relative importance of inter- and intraspecific interaction strength in a primary, lowland dipterocarp forest located at Danum, Sabah, Malaysia. Using data on 10 abundant overstory dipterocarp species from two 4-ha permanent plots, we evaluated the effects of neighbors on the absolute growth rate of focal trees (from 1986 to 1996) over increasing neighborhood radii (from 1 to 20 m) with multiple regressions. Only trees 10 cm to < 100 cm girth at breast height in 1986 were considered as focal trees. Among neighborhood models with one neighbor term, models including only conspecific larger trees performed best in five out of 10 species. Negative effects of conspecific larger neighbors were most apparent in large overstory species such as those of the genus Shorea. However, neighborhood models with separate terms and radii for heterospecific and conspecific neighbors accounted for more variability in absolute growth rates than did neighborhood models with one neighbor term. The conspecific term was significant for nine out of 10 species. Moreover, in five out of 10 species, trees without conspecific neighbors had significantly higher absolute growth rates than trees with conspecific neighbors. Averaged over the 10 species, trees without conspecific neighbors grew 32.4 cm(2) in basal area from 1986 to 1996, whereas trees with conspecific neighbors only grew 14.7 cm(2) in basal area, although there was no difference in initial basal area between trees in the two groups. Averaged across the six species of the genus Shorea, negative effects of conspecific larger trees were significantly stronger than for heterospecific larger neighbors. Thus, high local densities within neighborhoods of 20 m may lead to strong intraspecific negative and, hence, density-dependent, effects even in species rich communities with low overall densities at larger spatial scales. We conjecture that the strength of conspecific effects may be correlated with the degree of host specificity of ectomycorrhizae.

Secondary succession and dipterocarp recruitment in Bornean rain forest after logging

To understand succession in dipterocarp rain forest after logging, the structure, species composition and dynamics of primary (PF) and secondary (SF) forest at Danum were compared. In 10 replicate 0.16-ha plots per forest type trees >= 10 cm gbh (3.2 cm dbh) were measured in 1995 and 2001. The SF had been logged in 1988, which allowed successional change to be recorded at 8 and 13 years. In 2001, saplings (1.0-3.1 cm dbh) were measured in nested quadrats. The forest types were similar in mean radiation at 2 m height, and in density, basal area and species number of all trees. Among small (10 <= 31.4) and large ( >= 31.4 cm gbh) trees, in both 1995 and 2001, there were 10- and 3-fold more dipterocarps in SF than PF respectively; and averaging over the two dates, there were correspondingly ca. 10- and 18-fold more pioneers. Mortality was ca. 60% higher in SF than PF, largely due to a seven-fold difference for pioneers: for dipterocarps there was little difference. Recruitment was similar in PF and SE Stem growth rates were 37% higher in SF than PF for all trees, although dipterocarps showed the opposite trend. Among saplings, dipterocarps dominated SF with a 10-fold higher density than in PF. For dipterocarps, the light (LH) and medium-heavy (MHH) canopy hardwoods, and the shade-tolerant, smaller-stature other (OTH) species (e.g. Hopea and Vatica) were in the ratios ca. 40:15:45 in SF and 85: < 1:15 in PF. LHs had higher mortality than OTHs in SE In PF ca. 80% of the saplings were LH: in SF ca. 70% were OTH. The predominance of OTHs in SF is explained by the logging of primary rain forest which was in a likely late stage of recovery from natural disturbance, plus the continuing shaded conditions in the understorey promoted by dense pioneer vegetation. At 13 years after logging succession appeared to be inhibited: LHs were being suppressed but MHHs and OTHs persisted. Succession in lowland dipterocarp, rain forests may therefore depend on the successional state of the primary forest when it is logged. A review of logged versus unlogged studies in Borneo highlights the need for more detailed ecological comparisons.

Resistance of a lowland rain forest to increasing drought intensity in Sabah, Borneo

Occasional strong droughts are an important feature of the climatic environment of tropical rain forest in much of Borneo. This paper compares the response of a lowland dipterocarp forest at Danum, Sabah, in a period of low (LDI) and a period of high (HDI) drought intensity (1986-96, 9.98 y;1996-99, 2.62 y). Mean annual drought intensity was two-fold higher in the HDI than LDI period (1997 v. 976 mm), and each period had one moderately strong main drought (viz. 1992, 1998). Mortality of `all' trees greater than or equal to 10 cm gbh (girth at breast height) and stem growth rates of `small' trees 10less than or equal to50 cm gbh were measured in sixteen 0.16-ha subplots (half on ridge, half on lower slope sites) within two 4-ha plots. These 10-50-cm trees were composed largely of true understorey species. A new procedure was developed to correct for the effect of differences in length of census interval when comparing tree mortality rates. Mortality rates of small trees declined slightly but not significantly between the LDI and HDI periods (1.53 to 1.48% y(-1)): mortality of all trees showed a similar pattern. Relative growth rates declined significantly by 23% from LDI to HDI periods (11.1 to 8.6 mm m(-1) y(-1)): for absolute growth rates the decrease was 28% (2.45 to 1.77 mm y(-1)). Neither mortality nor growth rates were significantly influenced by topography. For small trees, across subplots, absolute growth rate was positively correlated in the LDI period, but negatively correlated in the HDI period, with mortality rate. There was no consistent pattern in the responses among the 19 most abundant species (n greater than or equal to 50 trees) which included a proposed drought-tolerant guild. In terms of tree survival, the forest at Danum was resistant to increasing drought intensity, but showed decreased stem growth attributable to increasing water stress.

Structure and inferred dynamics of a large grove of Microberlinia bisulcata trees in central African rain forest: the possible role of periods of multiple disturbance events

A 272-ha grove of dominant Microberlinia bisulcata (Caesalpinioideae) adult trees greater than or equal to 50 cm stem diameter was mapped in its entirety in the southern part of Korup National Park, Cameroon. The approach used an earlier-established 82.5-ha permanent plot with a new surrounding 50-m grid of transect lines. Tree diameters were available from the plot but trees on the grid were recorded as being greater than or equal to 50 cm. The grove consisted of 1028 trees in 2000. Other species occurred within the grove. including the associated subdominants Tetraberlinia bifoliolata and T. korupensis. Microberlinia bisulcata becomes adult at a stein diameter of c. 50 cm and at an estimated age of 50 y. Three oval-shaped subgroves with dimensions c. 8 50 in x 13 50 in (90 ha) were defined. For two of them (within the plot) tree diameters were available. Subgroves differed in their scales and intensities of spatial tree patterns, and in their size frequency distributions, these suggesting differing past dynamics. The modal scale of clumping was 40-50 m. Seed dispersal by pod ejection (to c. 50 in) was evident from the semi-circles of trees at the grove's edge and from the many internal circles (100-200 m diameter). The grove has the capacity. therefore, to increase at c. 100 m per century. To form its present extent and structure. it is inferred that it expanded and infilled from a possibly smaller area of lower adult-tree density. This possibly happened in three waves of recruitment, each one determined by a period of several intense disturbances. Climate records for Africa show that 1740-50 and 1820-30 were periods of drought, and that 1870-1895 was also regionally very dry. Canopy openings allow the light-demanding and fast-growing ectomycorrhizal M. bisulcata to establish, but successive releases are thought to be required to achieve effective recruitment. Nevertheless, in the last 50 y there were no major events and recruitment in the grove was very poor. This present study leads to a new hypothesis of the role of periods of multiple extreme events being the driving factor for the population dynamics of many large African tree species such as M. bisulcata.

Rainfall input, throughfall and stemflow of nutrients in a central African rain forest dominated by ectomycorrhizal trees

Incident rainfall is a major source of nutrient input to a forest ecosystem and the consequent throughfall and stemflow contribute to nutrient cycling. These rain-based fluxes were measured over 12 mo in two forest types in Korup National Park, Cameroon, one with low (LEM) and one with high (HEM) ectomycorrhizal abundances of trees. Throughfall was 96.6 and 92.4% of the incident annual rainfall (5370 mm) in LEM and HEM forests respectively; stemflow was correspondingly 1.5 and 2.2%. Architectural analysis showed that ln(funneling ratio) declined linearly with increasing ln(basal area) of trees. Mean annual inputs of N, P, K, Mg and Ca in incident rainfall were 1.50, 1.07, 7.77, 5.25 and 9.27 kg ha(-1), and total rain-based inputs to the forest floor were 5.0, 3.2, 123.4, 14.4 and 37.7 kg ha-1 respectively. The value for K is high for tropical forests and that for N is low. Nitrogen showed a significantly lower loading of throughfall and stemflow in HEM than in LEM forest, this being associated in the HEM forest with a greater abundance of epiphytic bryophytes which may absorb more N. Incident rainfall provided c. 35% of the gross input of P to the forest floor (i. e., rain-based plus small litter inputs), a surprisingly high contribution given the sandy P-poor soils. At the start of the wet season leaching of K from the canopy was particularly high. Calcium in the rain was also highest at this time, most likely due to washing off of dry-deposited Harmattan dusts. It is proposed that throughfall has an important `priming' function in the rapid decomposition of litter and mineralization of P at the start of the wet season. The contribution of P inputted from the atmosphere appears to be significant when compared to the rates of P mineralization from leaf litter.

Drought avoidance and the effect of local topography on trees in the understorey of Bornean lowland rain forest

The water relations of two tree species in the Euphorbiaceae were compared to test in part a hypothesis that the forest understorey plays an integral role in drought response. At Danum, Sabah, the relatively common species Dimorphocalyx muricatus is associated with ridges whilst another species, Mallotus wrayi, occurs widely both on ridges and lower slopes. Sets of subplots within two 4 -ha permanent plots in this lowland dipterocarp rain forest, were positioned on ridges and lower slopes. Soil water potentials were recorded in 1995-1997, and leaf water potentials were measured on six occasions. Soil water potentials on the ridges (-0.047 MPa) were significantly lower than on the lower slopes (-0.012 MPa), but during the driest period in May 1997 they fell to similarly low levels on both sites (-0.53 MPa). A weighted 40-day accumulated rainfall index was developed to model the soil water potentials. At dry times, D. muricatus (ridge) had significantly higher pre-dawn (-0.21 v. -0.57 MPa) and mid-day (-0.59 v. -1.77 MPa) leaf water potentials than M. wrayi (mean of ridge and lower slope). Leaf osmotic potentials of M. wrayi on the ridges were lower (-1.63 MPa) than on lower slopes (-1.09 MPa), with those for D. muricatus being intermediate (-1.29 MPa): both species adjusted osmotically between wet and dry times. D. muricatus trees were more deeply rooted than M. wrayi trees (97 v. 70 cm). M. wrayi trees had greater lateral root cross-sectional areas than D. muricatus trees although a greater proportion of this sectional area for D. muricatus was further down the soil profile. D. muricatus appeared to maintain relatively high water potentials during dry periods because of its access to deeper water supplies and thus it largely avoided drought effects, but M. wrayi seemed to be more affected yet tolerant of drought and was more plastic in its response. The interaction between water availability and topography determines these species' distributions and provides insights into how rain forests can withstand occasional strong droughts.

Litter breakdown and mineralization in a central African rain forest dominated by ectomycorrhizal trees

Based on litter mass and litterfall data, decomposition rates for leaves were found to be fast (k = 3.3) and the turnover times short (3.6 mo) on the low-nutrient sandy soils of Korup. Leaf litter of four ectomycorrhizal tree species (Berlinia bracteosa, Didelotia africana, Microberlinia bisulcata and Tetraberlinia bifoliolata) and of three non-ectomycorrhizal species (Cola verticillata, Oubanguia alata and Strephonema pseudocola) from Korup were left to decompose in 2-mm mesh bags on the forest floor in three plots of each of two forest types forest of low (LEM) and high (HEM) abundance of ectomycorrhizal (caesalp) trees. The litter of the ectomycorrhizal species decayed at a significantly slower rate than that of the non-ectomycorrhizal species, although the former were richer in P and N concentrations of the start. Disappearance rates of the litter layer showed a similar trend. Ectomycorrhizal species immobilized less N, but mineralized more P, than non-ectomycorrhizal species. Differences between species groups in K, Mg and Ca mineralization were negligible. Effect of forest type was clear only for Mg: mineralization of Mg was faster in the HEM than LEM plots, a pattern repeated across all species. This difference was attributed to a much more prolific fine root mat in the HEM than LEM forest. The relatively fast release of P from the litter of the ectomycorrhizal species suggests that the mat must allow an efficient uptake to maintain P in the forest ecosystem.

Reproductive investment and seedling survival of the mast-fruiting rain forest tree, Microberlinia bisulcata A. chev

In the southern part of Korup National Park, Cameroon, the mast fruiting tree Microberlinia bisulcata occurs as a codominant in groves of ectomycorrhizal Caesalpiniaceae within a mosaic of otherwise species-rich lowland rain forest. To estimate the amount of carbon and nutrients invested in reproduction during a mast fruiting event, and the consequential seed and seedling survival, three related field studies were made in 1995. These provided a complete seed and seedling budget for the cohort. Seed production was estimated by counting woody pods on the forest floor. Trees produced on average 26,000 (range 0-92,000) seeds/tree, with a dry mass of 16.6 kg/tree. Seeds were contained in woody pods of mass 307 kg/tree. Dry mass production of pods and seeds was 1034 kg ha(-1), equivalent to over half (55%) of annual leaf litterfall for this species, and contained 13% of the nitrogen and 21% of the phosphorus in annual leaf litterfall. Seed and young-seedling mortality was investigated with open quadrats and cages to exclude vertebrate predators, at two distances from the parent tree. The proportion of seeds on the forest floor which disappeared in the first 6 wk after dispersal was 84%, of which 26.5% was due to likely vertebrate removal, 36% to rotting, and 21.5% to other causes. Vertebrate predation was greater close to the stem than 5 m beyond the crown (41 vs 12% of seeds disappearing) where the seed shadow was less dense. Previous studies have demonstrated an association between mast years at Korup and high dry-season radiation before flowering, and have shown lower leaf-litterfall phosphorus concentrations following mast fruiting. The emerging hypothesis is that mast fruiting is primarily imposed by energy limitation for fruit production, but phosphorus supply and vertebrate predation are regulating factors. Recording the survival of naturally-regenerating M. bisulcata seedlings (6-wk stage) showed that 21% of seedlings survived to 31 mo. A simple three-stage recruitment model was constructed. Mortality rates were initially high and peaked again in each of the next two dry seasons, with smaller peaks in the two intervening wet seasons, these latter coinciding with annual troughs in radiation. The very poor recruitment of M. bisulcata trees in Korup, demonstrated in previous investigations, appears not to be due to a limitation in seed or young-seedling supply, but rather by factors operating at the established-seedling stage.