Forest Ecology and Management 216 (2005) 149–165 www.elsevier.com/locate/foreco Subsistence harvesting of pole-size understorey species from Ongoye Forest Reserve, South Africa: Species preference, harvest intensity, and social correlates S. Boudreau, M.J. Lawes *, S.E. Piper, L.J. Phadima School of Biological and Conservation Sciences, Forest Biodiversity Research Unit, University of KwaZulu-Natal, P/Bag X01, Scottsville 3209, South Africa Received 12 October 2004; received in revised form 11 April 2005; accepted 10 May 2005 Abstract We investigate the effect of subsistence harvesting on the ecological status of the pole-size tree component of the understorey at the Ongoye Forest Reserve (OFR; 2611 ha), KwaZulu-Natal province, South Africa. Using generalised linear modelling (GLM) we examine the ecological and social correlates of species preference and harvest intensity. Data were collected from 22 strip transects (5 m 300 m; 0.15 ha). Only 11.6% of the available pole-size trees (2 cm < DBH < 15 cm) were harvested, mostly for building materials. No instance of canopy tree logging was recorded. Sixty-eight species were identified; however, only seven species (82% of harvested stems) were preferred: Englerophytum natalense (33%), Garcinia gerrardii (19%), Drypetes gerrardii (9%), Tabernaemontana ventricosa (9%), Rinorea angustifolia (4%), Oxyanthus speciosus (4%), and Chrysophyllum viridifolium (4%). Size-class distributions for these seven species were inverse J-shaped, typical of fine-grained species that regenerate over small spatial scales, suggesting that current harvesting intensities may be sustainable. Pole-size stem density was significantly similar among residual stands in harvested areas (2014 31 stems ha1) suggestive of a stem-density harvest threshold below which further effort was unprofitable. The number of harvested stems increased with increasing stem availability across species and stem size-classes. Small size-classes (2–5 cm DBH) were harvested most intensely, followed by the intermediate (5–10 cm) and the large (10–15 cm) size-classes, for all species except C. viridifolium. For the latter, the harvesting intensity was greatest for the 10–15 cm size-class. Lastly, harvest intensity was greatest in those areas closest to households near the reserve boundary but was not affected by household density. Although subsistence harvesting at the OFR appears to be sustainable at current levels, we note that similar harvest intensities of pole-sized stems in studies from smaller forests (<60 ha) led to local extinction of tree species. In addition, because the dominant understorey species at OFR are almost exclusively harvested, the mid- to long-term effects of this harvesting preference on forest dynamics must be assessed to develop sound ecological forest management policies. # 2005 Elsevier B.V. All rights reserved. Keywords: Harvest intensity model; Poles; Species preference; Subsistence harvesting; Sustainable use; Subtropical forest; Understorey species * Corresponding author. Tel.: +27 33 2605 443; fax: +27 33 2605 105. E-mail address: lawes@ukzn.ac.za (M.J. Lawes). 0378-1127/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2005.05.029 150 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 1. Introduction Modern forest management systems are focussed on balancing the needs (rate of resource use) of users against the regeneration ecology and growth of the resource base to ensure the sustainable use and conservation of forest resources (McGregor, 1994; Hartshorn, 1995). Understanding the effects of harvesting on the composition and structure of the residual stand is essential for developing optimum harvesting systems (Cannon et al., 1994). Optimum management systems should only marginally alter the natural demography and biomass of the standing crop, and harvest levels should not exceed rates of resource regeneration or severely depress recruitment potential (Lawes et al., 2004). However, in most African forests subsistence harvesting is not effectively managed (Oates, 1999) and typically unsustainable harvesting rates are defined by various factors, such as short-term needs of consumers, power of traditional and formal authorities, size of consumer community, availability of suitable tree stem sizes, and forest size and accessibility (Burgess et al., 2000). In this study we examine the rates of harvesting of pole-sized timber and the important social and ecological correlates of subsistence harvesting intensity in Ongoye Forest Reserve (OFR), KwaZulu-Natal province, South Africa. In Africa, formal tree species selection and harvest strategies have to date focussed on large-scale and/or large-tree logging processes (West and Central Africa: Poorter et al., 1996; Struhsaker, 1997; South Africa: mortality-preemption methods; Seydack, 1995, 2000; Seydack et al., 1995). These strategies arise from, and are designed to meet, the demand for commercial timber. However, they overlook the intensive use and potentially insidious ecological effects of intensive subsistence harvesting of pole-size trees (Hall and Rodgers, 1986; Peden et al., 1996; Vermeulen, 1996; Vermeulen et al., 1996; Semesi, 1998; Ham and Theron, 1999; Burgess et al., 2000; Obiri et al., 2002). For example, Obiri et al. (2002) demonstrated that mostly pole-sized stems were harvested by subsistence users and furthermore, that a harvesting intensity of 10% of pole-size trees from relatively small forest patches in South Africa resulted in the local extinction of some understorey species. Of concern is that harvesting intensities as high as 50% of the available pole-size trees are common from easily accessible forests in Africa (Hall and Rodgers, 1986; Burgess et al., 2000). Of further concern is that harvested polesize trees are sometimes immature canopy trees, but are most frequently unreproductive individuals from understorey species. There is the potential for this selective harvesting of understorey species to affect forest tree dynamics, either causing deviation from the normal successional pathway for trees or arresting succession (Chapman et al., 1999) and changing tree species composition (Burgess et al., 2000). To date the ecology of understorey tree species has received little attention because these small trees are perceived to be abundant and of limited commercial value (Newbery et al., 1999). Indeed, in large forests that were exploited for their timber, a common silvicultural practice was to remove ‘‘useless’’ understorey species (Nicholson, 1965). Pole-size trees of understorey species may be important in the maintenance of forest dynamics and tree diversity (Newbery et al., 1999; Lawes and Obiri, 2003). The ecology of understorey species should be receiving greater attention, given that these stems are most targeted by subsistence harvesters in developing countries where the harvesting pressure on forests is greatest. Clearly, suitable tree species selection and harvest strategies that are based on the ecology of the polesize class need to be developed (Lawes and Obiri, 2003). In this study at OFR we evaluate the level of use of pole-size trees (species, size-class, harvested stems ha1) by local harvesters, as well as resource availability (the size of the residual stand). We identify the ecological (stand density and species richness) and the social (household distance and density) correlates of pole-size tree harvesting, and in so doing, develop a foundation for the management of the harvest of understorey tree species in a sub-tropical forest. 2. Methods 2.1. Study site The Ongoye Forest Reserve (288500 S, 318420 E; 3900 ha), created in 1914, is located on the Ongoye range of hills (altitude: 305–490 m), which are 12 km from the coast. The reserve comprises a species-rich S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 151 Fig. 1. Map of the study area with the 22 transects. The number of harvested stems ha1 is given for each transect. coastal scarp forest (2611 ha) set in a mosaic of coastal grassland and rocky granitic outcrops (1277 ha, Fig. 1). Scarp forest physiognomy is characterized by a poorly developed herb layer, a well-developed seedling and sapling stratum, an open understorey and a medium to high canopy (15–25 m; von Maltitz et al., 2003). The dominant tree species are Drypetes gerrardii, Englerophytum natalense, Millettia sutherlandii, Rinorea angustifolia, Rothmania globosa, Harpephyllum caffrum and Garcinia gerrardii. The mean number of understorey and canopy trees ha1 is 718 and 246, respectively and the mean canopy tree species richness 0.0625 ha1 plot is 9 (Krüger and Lawes, 1997). The forest was logged for saw-timber from the 1890s to 1924 and approximately 82,500 m3 of timber were removed (Anon., 1983). Logging was at its most intense from 1909 to 1919 by the ‘Ngoye Forest Company’ formed by Johnson and Carmont. No further legal extraction of large timber trees occurred after 1924. Since then it has been the perception of the conservation management institutions responsible for the Ongoye Forest Reserve that timber has been exploited by the local community to ‘an alarming extent’ (Anon., 1983). 2.2. Estimating harvest intensity Harvest intensity of pole-size trees was estimated from vegetation surveys conducted along 22 transects. Transects were 300 m long, with the exception of transects #4 (230 m), #6 (280 m) and #11 (270 m), and 5 m wide. The start of each transect was randomly located on a 1:10,000 orthophoto and was >50 m in from the forest edge. The position of the start and the end of each transect were recorded using a GPS. Transect direction was randomly chosen in each instance from 458 either side of a line judged to be perpendicular to the edge. Pole-size trees (5 cm < DBH < 15 cm) and the stumps of harvested stems (regardless of diameter) were identified to species level, counted and their DBH measured, in thirty contiguous 10 m 5 m quadrats, while smaller trees (2 cm < DBH < 5 cm) were recorded from 10 m 2 m quadrats located along the centre of the transect. Age of the harvested stems was ranked on a scale of 1–4 according to the degree of decay of the stump (1: wood colour and bark fresh, wood very hard, no coppices, green leaves on dead branches lying on the forest floor; 2: wood has darkened, bark in good 152 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 condition, wood hard but first signs of decay present, young coppice stems may be present; 3: wood dark and obvious signs of decay, bark still present but decaying, wood soft but is difficult to break, large coppice stems may be present; 4: wood rotten and stump easily broken, bark missing, coppice stems, if present, are dead). Truncated size-class frequency distributions (SCDs) were used to describe and evaluate the effect of harvesting on the recruitment of the most-commonly used species. We agreewith Condit et al. (1998) that such static and short-term data are not sufficient for predicting long-term forest dynamics. However, in the absence of longitudinal studies of the population dynamics of forest tree species, SCDs provide at least some insight into population stability and regeneration potential over time (Cawe and McKenzie, 1989; Lykke, 1998; Obiri et al., 2002). SCDs were analysed using the methods of Condit et al. (1998) and Lykke (1998). Thirteen DBH sizeclasses between 2 and 15 cm (i.e. >2–3, >3–4 >14– 15 cm), each with an interval of 1 cm, were used. The slope of the SCD was calculated for each of the seven most commonly used species (see below) using leastsquares linear regression, with size-class as the independent variable and the number of individuals in a class (Ni) as the dependent variable. To derive straightline plots of the size-class distribution, the number of individuals in each size-class was transformed by ln (Ni + 1) because some classes had zero individuals (Condit et al., 1998). The slopes of these regressions are referred to as SCD slopes in this study. The ecological interpretation of the truncated SCD was based on the shape of the regression slope, yielding four functional types of SCD. The difficulty of interpreting SCD slopes arises from the fact that any of the four functional types can lead to population persistence in a forest. Negative slopes of type I SCDs are a result of a higher number of individuals in the smaller size-classes than in the larger ones. An inverse J-shaped distribution is characteristic of the untransformed SCD for this type and is assumed to indicate normal recruitment and growth. A flat slope (not significantly different from zero) of type II SCDs indicates an equal number of regenerating trees and mature individuals. This SCD type can be the result of low recruitment or simply differs from type I because of fast growth in the smallest size-classes (Condit et al., 1998). SCDs of types III and IV have positive slopes because of many canopy individuals but few (type III) or no (type IV) regenerating individuals. Although these latter types can indicate regeneration processes that rely on the recurrence of standreplacing disturbances, they are considered, at least in South African forest (Everard et al., 1994), to be indicative of forest where recruitment and regeneration has been severely affected by causes, such as harvesting or other disturbances. Household density and distance from the forest were determined from the 1996 series of orthophotos (1:10,000 scale). The Ongoye Forest and the households within 3 km of the forest boundary were digitized from the orthophotos using ArcView (ESRI, 1996). Households were defined as obvious clusters of huts forming a homestead. 2.3. Statistical analysis Differences in the species composition of pole-size trees across the forest were tested using ANOSIM (analyis of similarity; BioDiversity Professional v.2.0, McAleece et al., 1997) based on species abundance data from the transects. Transects were grouped in pairs based on their proximity to each other (11 groups of 2 transects each). ANOSIM is analogous to standard univariate ANOVA and tests a priori defined groups against random groups in ordinate space. A ‘‘0’’ value indicates that there is no difference among group while a ‘‘1’’ value indicates that all samples within groups are more similar to one another than any samples from different groups. Two separate analyses were performed, one using all the species encountered along the transects, the other using only the seven species preferred by harvesters (see below). Differences in the harvest intensity between transects were examined using ANOVA. Stem size-class preference by harvesters was tested using the chi-square goodness-of-fit test (Siegel and Castellan, 1988). For this test, the observed values were the number of harvested stems in each of the size-classes while the expected values were obtained by multiplying the total number of harvested stems by the relative abundance of stems prior to harvesting (living and harvested stems) in each size-class. The effect of harvesting intensity on residual stem density and species richness was examined using linear regressions. For this analysis, residual density and species richness were estimated S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 153 from the number of live stems and species, respectively, recorded in 5 m 10 m quadrats along the 22 transects, while harvest intensity was indexed by the number of harvested stems per quadrat. Because of the potential autocorrelation in harvesting intensity between adjacent quadrats along the transects, we tested for serial correlation by computing the correlation coefficient of the proportion of cut stems in adjacent quadrats along the transects. In 20 out of 22 transects, there was no serial correlation (P > 0.05), corroborating field observations that harvesting intensity and pole-size tree density were very variable from one quadrat to the other. household distance and topography, and was ranked on a scale from 1 (easy access to the transect) to 4 (difficult access to the transect). Two interactions were also considered. First, species diameter class was added following a graphical exploration of the data that revealed no difference in the use of diameter classes across species, except for Chrysophyllum viridifolium in the larger size-class. This single interaction was coded as a zero for all other species and as one for C. viridifolium and the largest diameter class. Second, species availability was added because we suspected that some species might be preferentially selected even if their availability was low. The GLM was of the form: 2.4. Modelling the harvest intensity ln ðSi j Þ ¼ a0 þ a1 x1i j þ a2 x2i j þ We use a generalised linear model (GLM) based on a Poisson distribution with a logarithmic link function (McCullagh and Nelder, 1989) to model the harvest intensity. Using the model we identified which ecological and/or social variables influenced the harvest intensity. All species (n = 7) each comprising >3% of the total harvested stems were included in the model. The response variable was the number of harvested stems of a given species from a specified diameter class (2–5, 5–10, and 10–15 cm). Based on our knowledge of how harvesters select pole-size trees for extraction, the following variables were included in the a priori models that were examined: species, diameter class, household distance (distance of the nearest household from the start of the transect), household density (number of households within a 1 km radius of the nearest household to the start of the transect), availability of the resource (total number of stems before harvesting = stems + stumps of a given diameter for each species), and transect accessibility (Table 1). Accessibility is a composite index including where ln is the logarithm to base e, (Sij) is the number of harvested stems of species i and diameter class j; a0, a1, a2, are estimated coefficients for the variables x1ij, x2ij, for species i and diameter class j. All models were fitted using GENSTAT version 6.1 (Genstat 6 Committee, 2002). (1) 2.5. Model selection Ten candidate models were formulated a priori to avoid data dredging (Burnham and Anderson, 1998). To objectively select the most parsimonious model we used Akaike’s Information Criterion (AIC), which balances the fit of the model against the number of parameters used in the model (Anderson and Burnham, 2001). The AIC is calculated as residual deviance + 2 K (McCullagh and Nelder, 1989), where K is the number of parameters in the model. The model with the lowest AIC value (and a difference of at least two AIC units from other models) is accepted as the model best fitting the data. Table 1 Variables included in the generalised linear model of harvest intensity Variable Levels Species 7 Diameter class Household distance Household density Accessibility Availability 3 Continuous Continuous 4 Continuous Chrysophyllum viridifolium, Drypetes gerrardii, Englerophytum natalense, Garcinia gerrardii, Oxyanthus speciosus, Rinorea angustifolia, Tabernaemontana ventricosa 2–5 cm, 5–10 cm, 10–15 cm Varying from 534 to 2337 m Varying from 0 to 58 1–4 Varying from 1 to 25 154 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Table 2 The most abundant understorey species recorded at the Ongoye Forest Reserve Species Stems ha1 Percentage of all stems Harvested stems ha1 Percentage of harvested stems Englerophytum natalense Tabernaemontana ventricosa Garcinia gerrardii Rinorea angustifolia Alchornea hirtella Oxyanthus speciosus Drypetes gerrardii Tricalysia sonderiana Psychotria capensis Elaeodendron croceum Tricalysia capensis Cola natalensis Oricia bachmannii Chrysophyllum viridifoliuma 475 423 325 292 237 164 129 121 115 111 108 108 94 34 13.4 11.9 9.2 8.2 6.7 4.6 3.6 3.4 3.2 3.1 3.0 3.0 2.6 1.0 70 18 40 9 4 8 18 5 1 4 3 5 0 8 33.3 8.5 19.2 4.4 1.8 3.8 8.7 2.3 0.6 1.8 1.5 2.3 0.1 4.0 a Only trees that account for at least 2% of the pole-size tree population are shown except for C. viridifolium, which is shown because of its high use by local people. 3. Results windthrow and understorey gaps, where it formed dense stands in association with various vine species. 3.1. Species composition of pole-size trees 3.2. Harvesting: main effects We measured 6526 pole-size trees along 22 transects. We identified 68 species from these polesize trees (4.4% of the stems could not be identified). The species composition of the pole-size trees was dominated by seven species, which together accounted for 61% of the pole-size trees: Englerophytum natalense, Tabernaemontana ventricosa, Rinorea angustifolia, Garcinia gerrardii, Alchornea hirtella, Oxyanthus speciosus and Drypetes gerrardii (Table 2). Few stems from large canopy species (n = 537, 8.2%) were recorded posing a puzzle about the regeneration of canopy species. Among these seldom encountered canopy species, C. viridifolium appears to be the only species of importance to local harvesters (see below). Although G. gerrardii and D. gerrardii can sometimes reach substantial size, these species were considered part of the understorey/subcanopy stratum. Species composition of pole-size trees did not vary significantly among transects and thus across the forest (all species: RANOSIM = 0.60, P = 0.1; preferred species: RANOSIM = 0.52, P = 0.2). Except for A. hirtella, all the most abundant species mentioned above were present in all transects. A. hirtella was abundant in only 6 of the 22 transects and was found in disturbed areas, mainly canopy gaps caused by We recorded 853 harvested stems from 33 species representing 11.6% of the pole-size trees and a harvesting intensity of 264 30 stems ha1 (mean 1 S.E.; Table 3). No instance of canopy tree logging was recorded from the 22 transects. Harvesting intensity varied significantly among transects and across the forest (range = 87–587 stumps ha1; F 21,647 = 5.12, P < 0.001). The western (transects #13, #14, #16 and #17, mean = 477 stumps ha1) and the south-eastern sectors of the forest (transect #22, 400 stumps ha1) were most heavily used, while the northern half of the forest was less used (transects #1–#11, mean = 192 stumps ha1). Assuming that most of the harvested stems were cut in the last 5 years, local woodcutters remove on average 978 m3 ha1 year1. This represents pole consumption per capita of 19 poles year1 or 0.133 m3 ha1 year1 (616 households within 1.5 km of the forest, 11.9 persons/household, mean pole length = 2.5 m). We identified 683 (80%) of the 853 harvested stems to species level. The 170 unidentified harvested stems were rotten (class of decay 3 and 4), without bark or had no coppice stems. Seven species accounted for most (82%) of the identified harvested stems: E. natalense (33%), G. gerrardii (19%), D. gerrardii 155 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Table 3 Harvest intensity of pole-size trees recorded from east and southern Africa Forest Harvest intensity (%) Poles per capita (m3 year1) Forest area (ha) Availabilitya (stem ha1) Sustainability (yes/no) Source Ongoye For. Res. (RSA) Community forestsb (RSA) Gendagenda For. (Tan.) Litipo For. Res. (Tan.) Kimboza For. Res. (Tan.) Kimboza For. Res. (Tan.) Pande For. Res. (Tan.) Pugu For. Res. (Tan.) Kitulanghalo For. Res. (Tan.) Kitulanghalo Pub. land (Tan). Kitulanghalo For. Res. (Tan.) Manga For. Res. (Tan.) New Dab./Ulam. For. Res. (Tan.) West Kilombero For. Res. (Tan.) Bombo East 1 For. Res. (Tan.) Bombo East 2 For. Res. (Tan.) Amani Nature Res. (Tan.) Mgambo For. Res. (Tan.) Mlinga For. Res. (Tan.) Mpanga Village For. Res. (Tan.) Mtai For. Res. (Tan.) Nilo For. Res. (Tan.) Segoma For. Res. (Tan.) Semdoe For. Res. (Tan.) Kambai For. Res. (Tan.) Uganda Shimba Hills Forest (Ken.) Ndare Forest (Ken.) Mount Kenya Forest (Ken.) Kakamega Forest (Ken.) 11.6 10.0 10.6 8.1 13.1 44.0c 72.0c 58.0c 14.0 33.3 n.a. 2.7 d 1.6 d 0.3 d 7.0 d 4.5 d 4.1 d 4.9 d 3.2 d 22.5d 11.7d 7.1 d 0.8 d 4.4 d 31.0 n.a. 4.7 e 4.0 e 3.9 e 4.7 e 0.13 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a 0.14 n.a. n.a. n.a. n.a n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 0.04 n.a. n.a. n.a. n.a. 2611 4–56 2800 999 400 400 1450 2650 2452 n.a. 2638 1616 3728 104296 448 404 8380 1346 890 24 3107 6025 1168 980 1046 n.a. 9500 2500 51000 10100 746 1174 2288 3912 4656 4656 n.a. n.a. 174 286 n.a. 498d 1027d 773d 349d 268d 550d 506d 352d 545d 467d 411d 522d 472d 247 n.a 338 454 330 235 Yes No No No No No No No ? No Yes Yes Yes Yes ? ? ? ? ? ? ? ? ? ? ? ? Not Not Not Not This study Obiri et al. (2002) Burgess et al. (2000) Burgess et al. (2000) Burgess et al. (2000) Hall and Rodgers (1986) Hall and Rodgers (1986) Hall and Rodgers (1986) Luoga et al. (2002) Luoga et al. (2002) Luoga et al. (2000) Doggart et al. (1999a) Topp-Jørgensen et al. (2001a) Topp-Jørgensen et al. (2001b) Salter et al. (2002a) Salter et al. (2002b) Doody et al. (2001a) Oliver et al. (2002) Hall et al. (2002) Doody et al. (2001b) Doggart et al. (1999b) Beharrell et al. (2002) Doody et al. (2001c) Doggart et al. (2001) Cunneyworth (1996) Cunningham (1993) Wass (1995) Wass (1995) Wass (1995) Wass (1995) app. app. app. app. a Based on the number of stem with 5 cm > DBH < 10 cm except for the Kitulanghalo Forest Reserve (4 cm > DBH < 10 cm). Community forests of the Umzimvubu district of the Eastern Cape Province (South Africa): Nomyezo (56 ha), Mkolwane (44 ha), Kobemnyango (21 ha), Telelo (15 ha), Umgazi (10 ha), Magobiyani (6 ha), Ludume (4 ha). c Maximal harvest intensity recorded in the forest reserve. d Based on the number of saplings (5 cm > DBH < 15 cm) and a minimum of 2 m long relatively straight trunk. e Estimated pole-size stem yield. b (9%), T. ventricosa (9%), R. angustifolia (4%), O. speciosus (4%) and C. viridifolium (4%). The remaining harvested stems (18%) were distributed across 26 species. Cutting height was between 0 and 30 cm above the ground for most species except G. gerrardii for which stumps were often up to 1 m high. Poles were collected from one broad size class and most of the harvested stems (68%) had a diameter <6 cm (Fig. 2). A preference for poles in the 3–5 cm and the 7–10 cm diameter size-classes was observed as more stems were harvested than expected based on the availability of stems in these size-classes. Only the 2–3 cm size-class was less used than its availability would suggest (x212 ¼ 227:843, d.f. 12, P < 0.001). Regression analysis showed that pole-size tree density after harvesting was independent of the harvest intensity (density after harvesting = 0.096 [harvest intensity] + 9.945, F 1,646 = 1.17, P = 0.28). The latter suggests that harvesters reduce local pole-size tree density in a stand to a threshold level (mean 1 S.E.: 10 0.2 stems 0.005 ha1 or 2014 31 stems ha1) before moving on to harvest other stands. Below this critical threshold density it may not be profitable for a harvester to continue 156 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Table 4 Regresssion intercept and slope values for the size-class distributions of the seven most-commonly harvested species in Ongoye forest reserve Fig. 2. Size-class distribution of the harvested stems recorded in the 22 transects at the Ongoye Forest Reserve. Species Intercept Slope F0.05(1,11) P value Englerophytum natalense Garcinia gerrardii Drypetes gerrardii Rinorea angustifolia Tabernaemontana ventricosa Oxyanthus speciosus Chrysophyllum viridifolium 6.4971 0.2674 99.44 <0.001 6.1272 4.7079 6.1679 6.0032 0.3274 0.2063 0.2910 0.2022 49.48 21.54 215.75 79.76 <0.001 <0.001 <0.001 <0.001 5.2036 3.5647 0.2106 0.1981 127.01 29.10 <0.001 <0.001 harvesting (Fig. 3). In addition, there was a significant but small increase (not biologically significant) in species richness of residual stands with increasing harvest intensity (species richness after harvesting = 0.100 [harvest intensity] + 5.270, F 1,646 = 4.71, P = 0.03). per size-class are harvested, the only exception being the larger size-classes of C. viridifolium (13–14 cm; 14–15 cm) where more than 50% of the available stems were harvested (Fig. 4). 3.3. Size-class distributions—potential dynamics of the important harvested species The best fit GLM was Model 7 (Table 5). All the relationships described below are derived from this model. Harvest intensity increased exponentially with the availability of the resource or, in other words, the proportion of harvested stems from a given species for a given size-class increased with increasing availability (Fig. 5). As expected from the abundance of stems in the smaller size classes, the 2–5 cm size-class supported the highest harvest intensity, followed by the 5–10 cm and the 10–15 cm size-classes respectively, for all species except C. viridifolium (see above). Harvest intensity decreased with increasing distance from the nearest household (Fig. 6). We expected this relationship because harvesters tend to cut poles from those available closest to their household. There was no relationship between harvest intensity and the density of households closest to a transect (Fig. 6; harvested stumps ha1 = 0.064 [households] + 281, F 1,20 = 0.09, P = 0.77). The seven most commonly used species all displayed inverse J-shaped distributions with truncated SCD slopes varying between 0.195 and 0.341 (all significantly different from 0, Table 4) and are characterized by abundant seedling and sapling recruitment and a fine-grained spatial scale of regeneration (Lawes and Obiri, 2003). No missing size-classes were observed in the SCDs of the seven species. In fact, less than 25% of the available stems 3.4. Modelling harvesting intensity 4. Discussion Fig. 3. Relationship between harvest intensity and residual stem density of pole-size trees at the Ongoye Forest Reserve (mean 1 S.E.). Number in brackets represents the number of quadrat in each harvest intensity class. Quadrats with six stumps or more were pooled to facilitate graphic presentation. Pole-size trees in the range 3–10 cm DBH are mostly collected from the Ongoye Forest Reserve and evidence of harvesting activity was recorded S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 157 Fig. 4. Size-class distribution of the pole-size trees of the seven most-commonly used species at the Ongoye Forest Reserve. Harvested stems size-classes (scx) were adjusted according to age of the harvested stems: no correction for a fresh stump; from scx to scx+1 for a mid-aged stump; from scx to scx+2 for an old stump; and from scx to scx+3 for a very old stump. throughout the forest. Harvesters focussed almost exclusively on pole-size trees from understorey species and no canopy trees were harvested from the sample transects. Harvesting pressure was concentrated on, and proportional to, the availability of a few fine-grained understorey species (sensu Lawes and Obiri, 2003), whose density and rate of regeneration appears to be sufficient to support current harvesting rates. Nevertheless, the selective nature of harvesting will require careful management in the future to avoid the risk of depleting preferred species below sustainable levels. The uncontrolled harvesting of the most abundant and potentially most resilient species, has resulted in resource depletion in African forests (Struhsaker, 1997; Burgess et al., 2000) with significant changes to consumer preferences and forest dynamics (Hall and Rodgers, 1986; Obiri et al., 2002). It is our conviction that the ecology of understorey tree species and their role in the maintenance of tree biodiversity should be urgently researched to better understand and manage the long-term impacts of pole harvesting. 158 1064 1050 1018.6 1030 1209 1215 1215 1460 Minimum AIC 1018.6 1598 AIC 1392 1028 1315 18 1014 1315 18 980.6 1314 19 1004 1320 13 1179 1318 15 1191 1321 12 1193 1322 11 1454 1330 3 1378 1326 7 1596 1332 1 + 1.1730 + 0.9452 7 3 2 Continuous Continuous 4 Continuous Constant Species Diameter class Species diameter class Household distance Household density Access Availability Availability species Residual deviance Degree of freedom Parameters + + + + + + + 0.010450 1.990 + + 2.246 1.499 + + 2.181 0.000467 S0.745 + + 2.188 S0.000734 S0.022720 0.042 + + 1.985 0.000659 0.023700 0.011 + + 2.326 0.001004 0.01286 + 0.204 + + 2.320 0.000692 0.005230 0.4420 + + 2.317 0.000621 0.4960 6 5 4 3 2 1 0 = null Levels Model 4.1. Pole-size tree harvesting Factor/variable Table 5 Summary of the 10 candidate models built to infer the relationship between harvesting intensity and ecological and social correlates 7 8 9 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 The harvest intensity at the Ongoye Forest Reserve was evaluated at 11.6% of the standing crop and appears to be low to moderate (Table 3) compared to other studies in east and southern Africa (7–72%). Such harvest intensity is most likely due the large size of the forest (2611 ha) and the relatively low human population density living near the reserve (616 households within 1.5 km of the reserve, 12 persons/household, 149 persons km2, 2.8 persons ha1 of forest), and this may explain the lack of a relationship between household density and harvesting intensity. When expressed as the volume of wood used by local communities the average per capita pole consumption at the OFR was 0.13 m3 year1, assuming constancy in harvest intensity over the last 5 years. This result is similar to the value recorded in Tanzania (0.14 m3 year1; Luoga et al., 2000) but is much higher than the per capita consumption recorded in Uganda (0.04 m3 year1; Cunningham, 1993). Our static survey describes the harvesting that has occurred during the last few years and most of the harvested stems are believed to be <5-year-old. However, statistical and sampling issues need to be addressed before one can infer annual trends in the harvesting intensity from our static assessment of the situation at the Ongoye Forest Reserve. Inference of annual trends in harvest intensity necessitates absolute number of harvested stems per year over several years. Such data can only be obtained when strong cooperation exists between users and management institutions. At worst, our estimate of harvesting intensity at OFR represents the maximum rate of pole extraction per year. Harvesting is focussed on, and proportional to the availability of relatively few species, suggesting that the abundance of the standing stock at OFR is sufficiently high for consumers to choose among species and furthermore that current harvesting levels are probably sustainable. Declining availability of pole-size trees may force local communities to harvest inferior species. For example, a shift in species choice was recorded from the Pugu Forest Reserve (Tanzania) where consumers took to harvesting cashew-nut trees (Evers, 1994). At OFR the preferred species are still sufficiently abundant that a shift to less preferred species has yet to occur. In fact, all the highly preferred species harvested from Ongoye have a S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 159 Fig. 5. Relationship between harvest intensity and the availability of the pole-size trees for the seven most-commonly used species. Predicted relationships are from a generalized linear model (log link function). 160 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Fig. 6. Relationship between harvest intensity and the distance from the nearest household and the number of households as inferred from the harvest intensity model. fine-grained spatial scale of regeneration and consequently the density of pole-size trees is locally high. These species can sustain relatively high levels of harvesting (Lawes and Obiri, 2003). However, sustainability of subsistence harvesting is also function of the number of available stems per hectare and the size of the forest. For example, similar harvesting intensity (10%) resulted in the local extinction of fine-grained species in small South African forest patches (<55 ha) where harvesters had open access to the resource, even though the density of pole-size trees was higher than at OFR (Obiri et al., 2002). Based on the number of poles within the 5 cm < DBH < 10 cm size-class, the availability of pole-size trees ha1 at the OFR (754 stems ha1) appears to be in the lower range of East and South African forests (174– 4656 stems ha1; Table 3). Only forests in Kenya and the Kitulanghalo forest in Tanzania have lower pole-size tree density. In most studies where harvesting intensity was >10% of the available stems, authors S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 expressed at least some concerns about the sustainability of subsistence harvesting, even in very large forests with a high density of pole-size trees ha1 (Hall and Rodgers, 1986; Burgess et al., 2000; Luoga et al., 2002). In Kenya the pole-size stem yield for sustainable harvesting was estimated at a maximum of 5% of the available stems regardless of the forest size (Wass, 1995). At OFR, the high species-preference of subsistence harvesters, the fine-grained spatial scale of regeneration of the harvested species and the high numbers of small regenerative stems (DBH < 2 cm) all suggest that current harvesting levels are sustainable. As occurs elsewhere in Africa (Abbot and Homewood, 1999; Kirubi et al., 2000) tree selection was strongly influenced first by tree species then by size. Although we do not have data for wood density, most of the preferred species in this study are highly durable and termite proof, especially E. natalense, D. gerrardii, G. gerrardii, while other species are probably selected because they grow straight and are abundant. C. viridifolium poses a puzzle as this is a canopy species that occurs at low-density in the understorey and whose wood is not especially durable or strong. Small pole-size trees (DBH < 10 cm) were harvested from OFR and were used as building material for huts and fences, as fuelwood and for the curio trade, as is the case in most southern African countries (Table 6). Similar trends in size-class selectivity were recorded in three Tanzanian forests where trees with a DBH < 10 cm constitute the main fraction of the harvested stems (Gendagenda forest, 97%; Litipo forest, 82%; and Kimboza forest, 97%; Burgess et al., 2000). The absence of canopy tree logging by local harvesters is likely related to the lack of proper equipment to carry and process large trees, but also because the abundance of poles in the Table 6 Uses of pole-size trees harvested from the Ongoye Forest Reserve Use Building Fencing Curio Fuelwood Total Percentage (%) 62.2 15.3 14.3 8.2 100 Results were gathered from interviews of 63 household. 161 understorey makes the cutting and processing of large trees for laths and supports unnecessary. Harvesting intensity at Ongoye is correlated with household distance but not with household density, being higher in areas where households are nearest the forest. This result is consistent with the findings of several studies (Zimbabwe: Vermeulen, 1996; South Africa: Shackleton, 1993; Obiri et al., 2002; Kenya: Kirubi et al., 2000; Tanzania: Luoga et al., 2002). In East Africa, subsistence use of forest products is greatest by communities living <1.5 km from a forest and declines rapidly at greater distances (>5 km; Wass, 1995). The consistency of this finding throughout eastern and southern Africa suggests that by assessing household distance from the forest from recent orthophotos, one can crudely estimate the potential harvesting intensity and management requirements for indigenous forests. 4.2. Subsistence harvesting: implications for forest management and ecology Forest management strategies based on selective logging of large canopy trees for timber markets purport to reconcile the economic interests of producers with the needs of conservation (Rice et al., 1997). Even though much has been written about sustainable harvesting practises, there are few truly sustainable harvesting models for commercial timber extraction from indigenous forests (Hartshorn, 1995). Sustainable harvesting models for subsistence harvesting practices are more difficult to devise and implement than those based on commercial ventures. Subsistence harvesters operate within a short-term decision making horizon; they discount future benefits and frequently attach zero values to future returns gained from forgoing existing use (Luckert and Campbell, 2002). Thus, users cannot easily be persuaded to alter harvest rates or their species preferences until it is too late and the resource base has been irreparably damaged (Hardin, 1968; Hall and Rodgers, 1986; Sutherland and Reynolds, 1998). Particularly in developing countries, successful management of harvest activities is dependent on harmonising two time-horizons—the unavoidable short-term needs of harvesters and the long-term response of the forest resource base to harvesting. The latter is generally lacking from most forest 162 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 management plans. This study emphasises the need for research on the ecology of pole-size understorey species to avoid unintentional over-harvesting. Subsistence harvesting pressure in most African countries focuses on the small and mostly unreproductive trees found in the understorey stratum (Cawe and McKenzie, 1989; Muir, 1990; Nomtshongwana, 1999; Burgess et al., 2000; Obiri et al., 2002; Luoga et al., 2002). High harvesting intensities could result in missing size-classes for highly preferred species, decreasing the likelihood of conspecific replacement and increasing the risk of collapse of the natural successional pathway (Hall and Rodgers, 1986; McKenzie, 1988). Moreover, the rapid colonisation of understorey gaps by lianas or other woody species may also arrest succession. Recent studies suggest that understorey gaps might play an important role in the regeneration of forest tree species (Connell et al., 1997). Growing evidence indicates that the selection that takes place beneath a closed canopy has a great influence on the pool of species available to take advantage of the short burst of resources in newly-created canopy gaps (Connell, 1989). For example, the growth of seedlings beneath a closed canopy was negatively influenced by the understorey layer of leaves overhanging them (Denslow et al., 1991; Brown and Parker, 1994). Therefore, understorey gaps may provide suitable conditions for the establishment of shade-tolerant species, contributing to the maintenance of forest diversity (Connell et al., 1997). Whether or not even low harvesting rates of the understorey provide conditions suitable for establishment of shade-tolerant species is unknown. Clearly, the ecological role of the understorey needs to be assessed to better understand the long-term impacts of subsistence harvesting. 5. Conclusion The Ongoye forest appears to be able to sustain the current rates of harvesting of pole-size trees. This conclusion is based in large part on the fine-grained scale of regeneration and abundance of the understorey species that are currently harvested. Although the harvesting intensity is lower than in some other African forests (e.g., Hall and Rodgers, 1986), it must be recognised that a substantial proportion of the harvested pole-sized stems are from just a few species. Of considerable concern is that the mid- to long-term impacts of this selective harvesting practice on the forest dynamics are largely unexplored and unknown (Lawes and Obiri, 2003). A sound understanding of the ecological role of understorey tree species is essential to develop optimal management policies that will ensure the continued sustainability of the polesize tree harvest and the maintenance of tree diversity at OFR. Acknowledgements We are grateful for financial support from the Sappi-WWF TreeRoutes Partnership, the Fonds Québécois de Recherche sur la Nature et les Technologies and the National Research Foundation of South Africa (Indigenous Knowledge Systems) of South Africa under Grant no. 2053633. Any opinion, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Research Foundation. We are grateful to Alastair Campbell for assistance in the field and to Sharon Louw and Graham Keat of Ezemvelo KwaZulu-Natal Wildlife for their support. This study would not have been possible without the logistical support provided by the Mazda Wildlife Fund. References Abbot, J.I.O., Homewood, K., 1999. A history of change: causes of miombo woodland decline in a protected area in Malawi. J. Appl. Ecol. 36, 422–433. Anderson, D.R., Burnham, K.P., 2001. Commentary on models in ecology. Bull. Ecol. Soc. Am. 82, 160–161. Anon., 1983. Selection and assessment of conservation worthy areas for inclusion in the Nakor national plan for nature conservation: Ngoye forest. Natal Provincial working group for the assessment of land for nature conservation. Beharrell, N.K., Hall, S.M., Ntemi, S.A., 2002. Vegetation. In: Beharrell, N.K., Fanning, E.F., Howell, K.M. (Eds.), Nilo Forest Reserve—A Biodiversity Survey. Technical Paper 52, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–74. Brown, M.J., Parker, G.G., 1994. Canopy light transmittance in a chronosequence of mixed-species deciduous forests. Can. J. For. Res. 24, 1691–1703. S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Burgess, N.D., Matthews, P., Evers, Y., Woodcock, K., 2000. Non-Timber uses, threats and local attitudes. In: Burgess, N.D., Clarke, G.P. (Eds.), Coastal Forests of Eastern Africa. IUCN, Cambridge, pp. 281–302. Burnham, K.P., Anderson, D.R., 1998. Model selection and inference: a practical information theoretic approach. Springer-Verlag, New York. Cannon, C.H., Peart, D.R., Leighton, M., Kartawinata, K., 1994. The structure of lowland rainforest after selective logging in west Kalimantan, Indonesia. For. Ecol. Manage. 67, 49–68. Cawe, S.G., McKenzie, B., 1989. The Afromontane forests of Transkei, southern Africa. IV. Aspects of their exploitation potential. S. Afr. J. Bot. 55, 45–55. Chapman, C.A., Chapman, L.J., Kaufman, L., Zanne, A.E., 1999. Potential causes of arrested succession in Kibale National Park, Uganda: growth and mortality of seedlings. Afr. J. Ecol. 37, 81– 92. Condit, R., Sukumar, R., Hubbell, S.P., Foster, R.B., 1998. Predicting population trends from size distributions: a direct test in a tropical tree community. Am. Nat. 152, 495–509. Connell, J.H., 1989. Some processes affecting the species composition in forest gaps. Ecology 70, 560–562. Connell, J.H., Lowman, M.D., Noble, I.R., 1997. Subcanopy gaps in temperate and tropical forests. Aust. J. Ecol. 22, 163–168. Cunneyworth, P., 1996. Kambai Forest Reserve: A Biodiversity Survey. Technical Paper 41, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania. Cunningham, A.B., 1993. Low cost housing needs, wood use and woodlands. In: Pierce, G.D., Gombo, D.J. (Eds.), The Ecology and Management of Indigenous Forests in Southern Africa. Proceedings of an International Symposium, Victoria Falls, Zimbabwe, 27–29 July 1992, ZFC & SAREC, Zimbabwe, pp. 283–291. Denslow, J.S., Newell, E., Ellison, A.M., 1991. The effect of understorey palms and cyclanths on the growth and survival of Inga seedlings. Biotropica 23, 225–234. Doggart, N.H., Joseph, L., Bayliss, J., Fanning, E.F., 1999a. Manga Forest Reserve—A Biodiversity Survey. Technical Paper 41, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania. Doggart, N.H., Dilger, M., Kilenga, R., Fanning, E.F., 1999b. Mtai Forest Reserve—A Biodiversity Survey. Technical Paper 39, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania. Doggart, N., Ntemi, A., Kilenga, R., Doody, K.Z., 2001. Vegetation. In: Doggart, N.H., Doody, K.Z., Howell, K.M., Fanning, E.F. (Eds.), Semdoe Forest Reserve—A Biodiversity Survey. Technical Paper 42, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 10–34. Doody, K.Z., Kilenga, R., Ntemi, A., Barrio-Froján, C., 2001a. Vegetation. In: Doody, K.Z., Howell, K.M., Fanning, E.F. (Eds.), Amani Nature Reserve—A Biodiversity Survey. Technical Paper 52, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–64. 163 Doody, K.Z., Ntemi, A., Killenga, R., Beharrell, N.K., 2001b. Vegetation. In: Doody, K.Z., Beharrell, N.K., Howell, K.M., Fanning, E.F. (Eds.), Mpanga Village Forest Reserve—A Biodiversity Survey. Technical Paper 51, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 9–39. Doody K.Z., Kilenga, R., Ntemi, A., 2001c. Vegetation. In: Doody, K.Z., Howell, K.M., Fanning, E.F. (Eds.), Segoma Forest Reserve—A Biodiversity Survey. Technical Paper 50, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 13–50. ESRI, 1996. ArcView GIS, version 3.0. Environmental Systems Research Institute Inc., Redlands, California. Everard, D.A., van Wyk, G.F., Midgley, J.J., 1994. Disturbance and the diversity of forests in Natal, South Africa: lessons for their utilization. In: Huntley, B.J. (Ed.), Strelitzia 1. Botanical Diversity in Southern Africa. National Botanical Institute, Pretoria, pp. 275–285. Evers, Y.D., 1994. Subsistence strategies and wild resource utilization: Pugu Forest Reserve, Tanzania. M.Sc. Dissertation. Department of Anthropology, University College, London. Genstat 6 Committee, 2002. Genstat 5 reference manual. Clarendon Press, Oxford. Hall, J.B., Rodgers, W.A., 1986. Pole cutting pressure in Tanzanian forest. For. Ecol. Manage. 14, 133–140. Hall, S.M., Ntemi, A.S., Oliver, S., Smith, J., Murphy, A., 2002. Vegetation. In: Hall, S.M., Fanning, E.F., Howell, K.M., Pohjonen, V. (Eds.), Mlinga Forest Reserve—A Biodiversity Survey. Technical Paper 56, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–56. Ham, C., Theron, J.M., 1999. Community forestry and woodlot development in South Africa: the past, present and future. S. Afr. For. J. 184, 71–79. Hardin, G., 1968. The tragedy of the commons. Science 162, 1243– 1248. Hartshorn, G.S., 1995. Ecological basis for sustainable development in tropical forests. Ann. Rev. Ecol. Evol. Syst. 26, 155–175. Kirubi, C., Wamicha, W.N., Laichena, J.K., 2000. The effects of fuelwood consumption in the ASAL areas of Kenya: the case study of Marsabit forest. Afr. J. Ecol. 38, 47–52. Krüger, S.C., Lawes, M.J., 1997. Edge effects at an induced forestgrassland boundary: forest birds in the Ongoye Forest Reserve, KwaZulu-Natal. S. Afr. J. Zool. 32, 82–91. Lawes, M.J., Obiri, J.A.F., 2003. Using the spatial grain of regeneration to select harvestable tree species in subtropical forest. For. Ecol. Manage. 184, 105–114. Lawes, M.J., Obiri, J.A.F., Eeley, H.A.C., 2004. The uses and value of indigenous forest resources. In: Lawes, M.J., Eeley, H.A.C., Schackleton, C., Geach, B. (Eds.), Indigenous Forests and Woodlands in South Africa: Policy, People and Practice. University of Natal Press, Pietermaritzburg, pp. 227–276. Luckert, M.K., Campbell, B.M., 2002. Expanding our conceptual and methodological understanding of the role of trees and forests in rural livelihoods. In: Campbell, B.M., Luckert, M.K. (Eds.), Uncovering the Hidden Harvest: Valuation of 164 S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Methods for Woodland and Forest Resources. Earthscan Publications Ltd., London, pp. 228–238. Luoga, E.J., Witkowski, E.T.F., Balkwill, K., 2000. Subsistence use of wood products and shifting cultivation within a miombo woodland of eastern Tanzania, with some notes on commercial uses. S. Afr. J. Bot. 66, 72–85. Luoga, E.J., Witkowski, E.T.F., Balkwill, K., 2002. Harvested and standing wood stocks in protected and communal miombo woodlands of eastern Tanzania. For. Ecol. Manage. 164, 15–30. Lykke, A.M., 1998. Assessment of species composition change in savanna vegetation by means of woody plants’ size class distributions and local information. Biodiv. Conserv. 7, 1261–1275. McAleece, N., Lambshead, P.J.D., Paterson, G.L.J., Cage, J.D., 1997. Biodiversity Pro (v. 2.0). Natural History Museum and Scottish Association for Marine Science. McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models. Chapman and Hall, London. McGregor, J., 1994. Woodland pattern and structure in an peasant farming area of Zimbabwe: ecological determinants, present and past use. For. Ecol. Manage. 63, 97–133. McKenzie, B., 1988. Guidelines for the sustained use of indigenous forests and forest products. Occasional Report no 35. Foundation for Research and Development, Pretoria. Muir, D., 1990. Forest utilisation in KwaZulu. A case study of Hlatikulu forest reserve, Maputaland. Ph.D. dissertation. Institute of Natural Resources, University of Natal, Pietermaritzburg. Newbery, D.M., Kennedy, D.N., Petol, G.H., Madani, L., Ridsdale, C.E., 1999. Primary forest dynamics in lowland dipterocarp forest at Danum Valley, Sabah, Malaysia, and the role of the understorey. Phil. Trans. R. Soc. London B 354, 1763–1782. Nicholson, D.I., 1965. A review of natural regeneration in the dipterocarp forests of Sabah. Malay. Forester 28, 4–25. Nomtshongwana, N., 1999. Indigenous plant use and traditional exploitation methods of high value plants at Gxalingenwa and KwaYili State Forest. M.Sc. dissertation. Center for Environment and Development, University of Natal, Pietermaritzburg. Oates, J.F., 1999. Myth and reality in the rain forest: how conservation strategies are failing in West Africa. University of California Press, Berkley. Obiri, J.A.F., Lawes, M.J., Mukolwe, M., 2002. The dynamics and sustainable use of high-value tree species of the coastal Pondoland forests of the Eastern Cape Province, South Africa. For. Ecol. Manage. 166, 131–148. Oliver, S.A., Ntemi, A.S., Bracebridge, C.E., 2002. Vegetation. In: Oliver, S.A., Bracebridge, C.E., Fanning, E.F., Howell, K.M. (Eds.), Mgambo Forest Reserve—A Biodiversity Survey. Technical Paper 59, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–42. Peden, D.G., Okorio, J., Wajja-Musukwe, N., 1996. Commercial pole production in linear agroforestry systems. Agrofor. Syst. 33, 177–186. Poorter, L., Bongers, F., van Rompaey, R.S.A.R., de Klerk, M., 1996. Regeneration of canopy tree species at five sites in West African moist forest. For. Ecol. Manage. 84, 61–69. Rice, R.E., Gullison, R.E., Reid, J.W., 1997. Can sustainable management save tropical forests? Sci. Am. 34–39 April. Salter, R.F., Ntemi, A.S., Svoboda, N.S., Staddon, S., Bracebridge, C.E., 2002a. Vegetation. In: Staddon, S., Fanning, E.F., Howell, K.M. (Eds.), Bombo East I Forest Reserve—A Biodiversity Survey. Technical Paper 41, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–43. Salter, R.F., Ntemi, A.S., Svoboda, N.S., Staddon, S., Bracebridge, C.E., 2002b. Vegetation. In: Staddon, S., Fanning, E.F., Howell, K.M. (Eds.), Bombo East II Forest Reserve—A Biodiversity Survey. Technical Paper 58, East Usambara Conservation Area Management Programme, Ministry of Natural Resources and Tourism, Tanzania, pp. 11–42. Semesi, A.K., 1998. Mangrove management and utilization in Eastern Africa. Ambio 27, 620–626. Seydack, A.H.W., 1995. An unconventional approach to timber yield regulation for multi-aged, multispecies forests. I. Fundamental considerations. For. Ecol. Manage. 77, 139–153. Seydack, A.H.W., 2000. Theory and practice of yield regulation systems for sustainable management of tropical and subtropical moist natural forests. In: von Gadow, K., Pukkala, T., Tomé, M. (Eds.), Sustainable Forest Management. Kluwer Academic Publishers, Netherlands, pp. 257–317. Seydack, A.H.W., Vermeulen, W.J., Heyns, H.E., Durrheim, G.P., Vermeulen, C., Willems, D., Ferguson, M.A., Huisamen, J., Roth, J., 1995. An unconventional approach to timber yield regulation for multi-aged, multispecies forests. II. Application to a South African forest. For. Ecol. Manage. 77, 155– 168. Shackleton, C.M., 1993. Demography and dynamics of the dominant woody species in a communal and protected area of the eastern Transval lowveld. S. Afr. J. Bot. 59, 569–574. Siegel, S., Castellan Jr., N.J., 1988. Nonparametric statistics for the behavioral sciences. McGraw-Hill Book Company, Singapore. Struhsaker, T.T., 1997. Ecology of an African rain forest: logging in Kibale and the conflict between conservation and exploitation. University Press of Florida, Gainesville. Sutherland, W.S., Reynolds, J.D., 1998. Sustainable and unsustainable exploitation. In: Sutherland, W.S. (Ed.), Conservation science in action. Blackwell Science, Oxford, pp. 90– 115. Topp-Jørgensen, J.E., Marshall, A.R., Brink, H., 2001a. Assessment of the impact of human forest use in New Dabaga/Ulangambi Forest Reserve. In: Doody, K.Z., Howell, K.M., Fanning, E.F. (Eds.), New Dabaga/Ulangambi Forest Reserve—Botanical and Forest Use Report. Report for the Udzungwa Mountains Forest Management and Biodiversity Conservation Project, MEMA, Iringa, Tanzania, pp. 80–93. Topp-Jørgensen, J.E., Marshall, A.R., Brink, H., 2001b. Assessment of the impact of human forest use in West Kilombero Scarp Forest Reserve. In: Doody, K.Z., Howell, K.M., Fanning, E.F. (Eds.), West Kilombero Scarp Forest Reserve—Botanical and Forest Use Report. Report for the Udzungwa Mountains Forest Management and Biodiversity Conservation Project, MEMA, Iringa, Tanzania, pp. 74–82. Vermeulen, S.J., 1996. Cutting of trees by local residents in a communal area and an adjacent state forest in Zimbabwe. For. Ecol. Manage. 81, 101–111. S. Boudreau et al. / Forest Ecology and Management 216 (2005) 149–165 Vermeulen, S.J., Campbell, B.M., Matzke, G.E., 1996. The consumption of wood by rural households in Gokwe communal area. Zimbabwe Hum. Ecol. 24, 479–491. von Maltitz, G., Mucina, L., Geldenhuys, C., Lawes, M.J., Eeley, H.A.C., Adie, H., Vink, D., Fleming, G., Bailey, C., 2003. 165 Classification System for South African Indigenous Forests: An objective classification for the Department of Water Affairs and Forestry, report no. ENV-P-C 2003-017. CSIR, Pretoria. Wass, P., 1995. Kenya’s indigenous forests: status, management and conservation. IUCN, Cambridge.
