Growth potential of twelve Acacia species on acid soils in Hawaii

Pores;~;ology Management EISEVIER Forest Ecology and Management 80 (1996) 175-186 Growth potential of twelve Acacia species on acid soils in Hawaii Thomas G. Cole av* , Russell S. Yost b, Richard Kablan b, Thomas Olsen b a Institute of Pacific Islands Forestry, Pacific Southwest Research Station, Forest Service, US Department b Department of Agronomy 1151 Punchbowl and Soils, University St. Rm. 323, Hotwlulu, of Hawaii at Manoa, HI 96813, USA 1910 East-West Rd., Honolulu, of Agriculture, HI 96822, USA Accepted 27 June 1995 Abstract Reforestation of degraded tropical sites is often hampered by soils of high acidity, high aluminum saturation, and low fertility. To evaluate the possibility of cultivating Acacia species on such soils, a study was conducted at Waiawa, HI, to test growth under conditions of (1) high acidity (primarily aluminum) and nutrient stress, and (2) no acidity stress and high nutrient availability. Twelve Acacia species, including the important native Hawaiian species Acucin koa, were established on a Ustic Kanhaplohumult soil. The experimental design was a split plot with two fertility treatments as the main plots and the 12 Acacia species as subplots. The treatments were: low fertility (F,; 143 kg ha- ’ 14- 14-14 plus micronutrients) and high fertility (F,; 8 Mg ha-’ lime, 143 kg ha- ’ 14-14-14 plus micronutrients, 200 kg P ha- ‘, and 77 kg K ha- ‘). Acacia angustissima, Acacia aulacocarpa, Acacia auriculiformis, Acacia cincinnata, Acacia crassicarpa, Acacia implexa, Acacia koa, and Acacia mangium grew significantly faster under the high fertility treatment. Three species, A. cincinnata, A. crassicarpn, and A. mungium, are recommended for planting on infertile acid soils. The volume of A. koa was increased ten-fold by the high fertility treatment. Additional study on koa’s nutritional requirements is suggested in order to identify the nutrients contributing to this increased growth. Keywords: Silviculture; Reforestation; Plantation 1. Introduction Reforestation efforts on degraded grassland sites in the tropics are often hampered by soil of low fertility. Soils on degraded sites are often characterized by high acidity (pH in 1:l water less than 5.5), high aluminum saturation, and low fertility. They are commonly Oxisols or Ultisols. Acid soils are quite prevalent in the Pacific region. A survey of 15 major * Corresponding author. island countries in Oceania estimated that 15% of the total land area (85 000 km21is classified as acid soil (Morrison, 1988). Acid soils are very fragile and, if cleared of their protective vegetative cover, can quickly lose fertility owing to accelerated decomposition and erosion. Burning may further degrade the site (MuellerDombois, 1981). Once an area has degraded as a result of fire and erosion, natural reforestation is difficult. Even if fire is excluded from the area, the surrounding forest may not contain speciesthat are capable of recolonizing such sites. For successful 0378-l 127/96/$15.00 0 19% Elsevier Science B.V. All rights reserved SSDI 0378-l 127(95)03610-5 176 T.G. Cole et al./F‘orest Ecology and Management reforestation of degraded areas, it is therefore necessary to utilize trees having special adaptations and characteristics. Many legumes are ideal candidates for reforestation of degraded sites because of their ability to convert molecular nitrogen (N,) to ammonium. Besides supplying nitrogen for their own nutritional needs, legumes can enrich the soil through decomposition of their nitrogen-rich leaf litter. Legumes planted in soils low in nitrogen and organic matter therefore require little supplemental nitrogen fertilization beyond establishment and can aid in improving the soil. Lowland tropical trees in the genus Acacia are known for their excellent performance when used for reforestation on infertile acid soils (National Research Council, 1983). Eleven non-native Acacia species were selected for this trial on acid soils in Hawaii (Table 1). During the selection process a balance was sought between those acacias that have been relatively unstudied outside their natural habitat and those that have been successfully introduced outside their native range such as A. mangium and A. auriculiformis (Tumbull, 1987; Boland and Tumbull, 1989). Acacia aulacocarpa, A. cincinnata, A. crassicarpa, and A. polystachya were included in this trial because of reported good wood properties, fast growth, and adaptability to a wide range of site conditions (National Research Council, 1983). Other Table 1 Species and provenance information, Species A. A. A. A. A. A. A. A. A. A. A. A. angustissima (Mill.) 0. Kuntze. aulacocurpa A. Cunn. ex Benth. auriculiformis A. Cunn. ex Benth cincinnatu F. Muell crassicarpa A. Cunn. ex Benth implexa Benth koa Gray leiocalyx (Domin) Pedley leptocurpu A. Cunn. ex Beoth. mangium Willd. orites Pedley polystachya A. Cunn. ex Benth. Acacia species trial, Waiawa, CSIRO seedlot I6947 18246 15365 16977 14740 15.557 18003 17701 17266 13871 ‘I 80 (19961175-186 species in this experiment, selected for having similar characteristics, were Acacia angustissima, Acucia implexu, Acacia leioculyx, Acaciu leptacarpu, and Acacia orites. All of these species are considered pioneer species. A native Hawaiian species, A. kw. was alis:) included in this trial. Although koa is a climax speciesunlike the other 11 acacias, it is an extremely valuable local wood, and it is very important to the Hawaiian culture. Koa has been extensively planted in Hawaii with mixed results (Carbon and Bryan, 1959; Whitesell, 1964; Bums and Honkala, 1990; Scowcroft and Adee, 1991). Slow growth rates, susceptibility to diseaseand insect attack, use of inap propriate provenances, and a lack of management have contributed to the difficulty of establishingkoa in plantations. We felt that including koa in this experiment might provide some insight concerning koa’s nutritional requirementsand other management needs. The experiment was terminated after I! years and the trees harvested. Data from the harvest will be used to calculate biomassequations for each of the 12 species. The site was replanted with A. cincinnata, A. crassicarpa, and A. mangium. The early end to the experiment was necessarybecause we felt the goal of identifying suitable specieshad been met and that due to the small plot size, boundary effects were becoming significant, HI, 1993 Seed source h Lat. Long. Elevation (m) HI PNG 22”OO’ N 7”56’S 15”01’S 1 h”3S.S 8”49’S 27”05’S I 9”40’ N 26”24’S 12”19’S 12”45’S 28”14’S 16”58’S 158”OO’ W 142”3S’ E 143”40’ E 145”25’ E 142”48’ E 151”46’E 155”l I’W 151”23’E 133”19’E 143”17’E 153”17’E 147”37’ E 20 15 loo 410 4.5 600 630 380 60 37 550 37 _- QLD QLD PNG QLD ED NT QLD QLD QLD a Commonwealth Scientific and Industrial Research Organization, Australia. h HI, Hawaii; PNG, Papua-New Guinea; QLD, Queensland, Australia; NT, Northern ii Provenance from Kaumana, Big Island of Hawaii. Territory, Australia. -I-.-.- ,.... T.G. Cole et al./Forest 2. Materials Ecology and Management and methods 2.1. Study site and experimental design The experiment was established in June 1991 at the 80 ha Waiawa Correctional Facility (WCF), a minimum security prison in the Waiawa area of Oabu, HI (Fig. 1; 21”30’N; 157”5O’W). The site lies at 250-300 m elevation and is on the summit of a gently sloping hill having a westerly aspect. The soil at Waiawa is a well-drained clayey, oxidic, isohyperthermic Ustic Kanhaplohumult of the Leilehua se- Fig. 1. Location of the Waiawa 80 (1996) 175-186 177 ries, characterized by low pH and high aluminum saturation throughout the soil profile. Annual rainfall at Waiawa ranges from 1000 to 1500 mm year-’ (Giambelluca et al., 1986); incomplete rainfall data from the site indicate that this was a normal rainfall year. In general the site has wet winters and dry summers. Annual air temperatures at Waiawa average 22°C with a minimum of 16”C, and maximum of 29°C. The experimental layout was a split-plot design with three replications, two fertility levels, and 12 Acacia species. The main plots received either a low Correctional Facility, Waiawa, Oahu, HI. 178 T.G. Cole et al./ Forest Ecology (F,) or high (F,) fertility treatment (details given below). Each of the 12 subplots contained six individual trees of one Acacia species.Tree spacingwas 2.5 m by 2 m in two three-tree rows. Main plots were arranged in a randomized complete block design; subplots were randomly located. In October 1990 the site was cleared by bulldozer and plowed to a depth of 20 cm. After 2 weeks the site was rototilled to incorporate a lime application on the F, plots. The lime treatment consisted of 17 parts calcite (CaCO,) to 1 part dolomite (52:48 CaO:MgO). The site lay fallow for 7 months to allow the lime to take effect. 2.2. Seedlingpreparation Seeds were sown in dibble tubes (115 cm3per tube) on 9 October 1990. The potting soil was inoculated with a combination of rhizobia and VA mycorrhiza (Glomus aggregatum). The specific inoculum requirements of only two species(A. auriculiformis and A. mangium) were known, in the absenceof better guidelines all speciesreceived the sameinoculum. Seedlingswere fertilized in the nursery twice with liquid fertilizer (20-20-20) and once with 1 g of slow release fertilizer (14- 14- 14). The seedlings grew in the nursery for approximately 8 months and were outplanted on 30 May 1991. Seedlings were watered immediately after planting and again on 31 May 1991. Trees in replication 1 were watered again on 3 June 1991, becausethey showed signsof stress. 2.3. Weed and insect control Weed growth was variable at the site; the heaviest concentration of weeds were in replications 2 and 3. The two replications were sprayed with 2 1 of glyphosate ’ (8 ml I- ’ of water) on 21 May 1991. On 17 June 1991, the few weeds in replication 1 were removed by hand sickle. Replications 2 (F,,) and 3 were weeded by hand sickle on 5 November ’ No recommendation of pesticide use is intended by this publication, nor does it imply that the pesticide has been registered by the appropriate government agencies. and Management 80 (1996) 175-186 1991, and replications 1 and 2 (F, ) on 12 November 1991. The Chinese rose beetle (Adore&s sinicus Burmeister), a polyphagous leaf-eating beetle, caused extensive phyllode damage (greater than 90% on some seedlings)in the first week after planting. All seedlingswere immediately sprayed with Sevin’ on 5 June 1991. Insecticide was also applied on 11 June, 17 June and 5 August 1991. 2.4. Fertilization and soil analysis At planting, all trees in the experiment received an initial fertilization to aid establishment.On 3 June 1991, each tree was fertilized with 71.4 g (142.8 kg ha-’ ) of complete fertilizer (14-14-14 + Fe) and with 20.8 g (41.6 kg ha- ’ ) of a micronutrient fertilizer mix (S, 12%; B, 0.1%; Cu. 0.5%; Fe, 12%; Mn, 2.5%; MO, 0.05%; Zn, 1%). After 4 months, trees in the F, plots received additional fertilization. Triplesuperphosphate(TSP) and potassium chloride &Cl) were applied at a rate of 500 g TSP per tree (200 kg Pha-‘)and73gKClpertree(77kgKha-‘)on24 September 1991. Fertilizer was applied in two opposite holes, each approximately 15 cm from the seedling. Soil sampIes were collected before the site was cleared and again after the experiment was terminated. A single composite surface sample was collected in June 1990, prior to clearing. In March 1993, surface horizon sampleswere taken from the center of each six-tree plot and averaged to get main plot means (F, and F,). In addition, a composite sample was collected from each main plot for subsurface nutrient analysis at 22 months. Soils were analyzed at the University of Hawaii’s Agricultural Diagnostic Service Center, using standard techniques (extractants: modified Truog for phosphorusand ammonium acetate for bases). 2.5. Measurements Total tree height (HT) was measured1 week after planting and at 2, 4, 6, 9, 12, and 19 months. Height was measuredat the highest point of viable meristematic tissue (apical meristem if alive). Tree basal diameter (BD at 30 cm) and diameter at breastheight (DBH at 1.3 m) were measuredat 6, 9, 12, and 19 T.G. Cole et al. / Forest Ecology months. Because many of the species were multistemmed, the largest stem was chosen for DBH and BD measurements during early time periods. At 19 months the DBH and BD of each individual stem were measured and averaged by tree. Mean annual increment of height @IT MAI) and diameter (DBH MAI) were calculated by dividing I-IT or DBH by the number of years since planting. Phyllode damage was assessed at all time periods and was the visual estimation of the percent phyllode loss by insect feeding. The same technician evaluated damage at all time periods. Tree volume (VOL) was calculated by and Management 80 (1996) 175-l 179 86 using the mathematical formula for the frustum of a cone. Although this method probably underestimates tree volume, it was deemed adequate for comparative purposes. 2.6. Statistical analysis Analysis of variance was performed using the general linear models (GLM) procedure of the Statistical Analysis Systems @AS) Institute Inc. (1985). Significant differences between fertility levels for each species were identified using the Games and Table 2 ANOVA results for treatments and their interactions, according to the General Linear Model Procedure (SAS Institute. species trial, Waiawa, Hawaii, 1993 Growth Months after planting variable a 0 2 4 6 9 12 1985) in the Acacia 19 Total height (rn) Fertility (Fert) Species (Spec) Fert X Spec Phyllode akmage 0.54 0.00 0.85 0.12 0.00 0.19 - 0.42 0.46 0.46 0.10 0.00 0.79 0.09 0.00 0.04 0.08 0.00 0.02 0.01 0.00 0.51 0.00 0.66 0.23 0.00 0.00 0.42 0.46 0.45 - 0.11 0.00 0.01 0.11 0.00 0.00 0.01 0.21 0.00 0.02 0.59 0.00 0.30 0.42 0.46 0.46 0.14 0.02 0.64 0.74 0.12 0.97 0.76 0.00 0.95 0.07 0.00 0.18 0.04 0.00 0.05 0.06 0.00 0.01 0.10 0.00 0.04 0.00 0.00 0.00 0.07 0.00 0.05 0.08 0.00 0.00 0.12 0.00 f %) Fertility (Fett) Species &e.c) Fert x Spec Basal diameter fern) Fertility (Fert) Species (Spec) Fert x Spec Diameter 0.82 0.04 0.00 I%) Fertility (Fert) Species (Spec) Fert X Spec Survival 0.24 0.00 0.49 - at breast height (cm) Fertility (Fert) Species (Spec) Fert x Spec - 0.01 Tree volume cm3 per tree) Fertility (Fert) Species (Spec) Fert X Spec - - - 0.06 0.00 0.05 Stems per tree Fertility (Fert) Species (Spec) Fert X Spe.c - 0.05 0.00 0.10 P < 0.05 are in bold face. All significance levels < 0.005 are designated as 0.00. a Natural log tramformation applied to HT, BD, DBH, VOL. and stems per tree. Arcsine kansformation applied to survival and phyllode damage. T.G. Cole et al. /Forest 180 Ecology and Management Howell modification of Tukey’s honestly significant difference test (Milliken and Johnson, 1992). Tukey ‘s Studentized Range Test was used to test between soil 80 (1996) 175-186 fertility levels. Data from this experiment violated the assumption of homogeneity of variances, and therefore transformation was necessary. A natural Table 3 Growth variables a at 19 months after planting, for 12 Acacia species grown under low (F,,) and high (F,) fertility levels, Waiawa, HI, 1993 Species and fertility level DBH km) DBH MAI (cm year-’ 1 Stems per free HT MAI Cmyear- ’ 1 VUL Cm’ ha- :j A. angustissima h Fl 1.0 * 0.6 1.9 1.2 1.6 * 2.3 2.1 1.7 4.6 2.9 3.6 7.6 2.0 * * 15.6 3.3 2.1 2.7 3.6 4.4 2.3 2.8 4.7 4.3 6.6 I. 4' 2.5 16.4 ** * * A. aulacocarpa 51 F, 5.7 ” A. auriculiformis 2.1 1.3 3.3 3.7 2.3 4.3 2.1 2.7 4.1 5.5 1.4 1.9 1l.K A. cincinnata 6, F, 3.4 5.4 2.1 4.0 5.3 2.5 3.3 6.0 8.7 1.3 1.8 28.9 A. crassicarpa F,, F, 4.8 6.2 3.0 3.9 4.4 2.8 7.0 1.1 i2.1 * 5.5 3.5 8.6 1.6 27.6 A. implexa ‘7, F, 1.9 3.5 1.2 3.4 4.3 2.1 3.0 * 2.2 2.7 5.2 1.1 I.3 7.4 * 10.0 0.9 2.2 1.7 * 4.5 1.1 1.4 0.4 4.3 6.8 10.5 5, F, A. koa ‘7, F, 3.4 4.0 I!.2 2.4 1.5 1.4 3.5 3.2 3.3 2.0 2.1 4.2 4.5 2.6 2.8 4.4 5.2 1.5 1.9 2.7 2.6 1.7 1.6 3.4 3.6 2.1 2.3 5.0 5.2 1.1 4.4 6.3 2.8 4.0 4.0 2.5 5.9 5.2 3.3 8.4 1.8 2.5 1.7 1.1 1.5 3.4 4.6 2.1 2.9 2.1 3.4 1.9 2.7 3. : 7.7 0.3 0.7 1.9 2.4 1.2 1.5 2.2 2.8 2.5 3.7 3.0 0.5 l * 0.3 l * * * ’ 4 A. leiocalyx 5, Fl A. leptocarpa 5, F, A. mangium h F, 1.5 6. i 8.3 14.0 36.3 * A. orites 5, F, 2.4 A. polystachya 6, F, 0.4 1.1 1.4 a Natural log transformation applied to DBH, HT, BD, VOL, and stems per tree rest&s. Resented rest& are bias-cotxcted, re%ansformed means (Meyer, 1944). (I = 0.01: Tukey’s HSD between fertjlity treatments for the same species was applied to transfwrncd means @@iken and Johnson. 1992). * a = 0.05; l * T.G. Cole et al./ Forest Ecology and Management 80 (1996) 181 175-186 log transformation was performed on HT, DBH, BD, VOL, and stems per tree. An arcsine transformation was performed on percent survival and phyllode damage. Retransformed logarithmic means were corrected for bias (Meyer, 1944). 2.7. Speciesperformance 0 A species’ performance or site adaptability was determined by its volume Cm3 ha-‘) under each of the two fertility treatments. Volume was used to appraise performance because all growth measures (HT, DBH, BD, and number of stems per tree> were used in its calculation. Although expansion of tree volume in such small plots to a per hectare basis is questionable, it allows us also to factor species survival into performance appraisal. Growth of A. koa, assessed by HT MAI, was compared with growth rates for koa grown in three other experiments at Waiawa. Unpublished data were available for koa grown in phosphorus rate studies on Oxisols and Ultisols (Silva, 1994) and a seed source study (Ikawa, 1994). The growth of A. koa was also compared with that of the other Acacia species in our experiment. This comparison however, may be misleading because, as mentioned earlier, koa is not a fast-growing pioneer species. 3. Results 3.1. Growth In the nursery A. koa grew the fastest while A. orites and A. implexa grew the slowest. Results of the analysis of variance on initial tree HT (0 months) showed no significant difference by fertility level or in the species by fertility interaction (Table 2). The effect of liming the soil on early tree growth was evaluated during the first 4 months of the experiment. During this period, the F,, and F, treatments differed only in the application of lime in the F, plots. None of the 12 species significantly increased growth at 2 or 4 months in response to the increase in pH associated with the lime treatment (Table 2). At 6 months, HT and DBH growth showed a significant response in the interaction of species by fertility (Table 2). This response continued through 5 10 15 20 15 20 Months 0 5 10 Months Fig. 2. Diameter and height growth response of A. koa to liming and supplemental P and K fertilization at Waiawa, Oahu, HI, 1993. The asterisk indicates significance at a = 0.05. 19 months. Basal diameter growth was slower to respond to the fertilization at 4 months; differences became significant at 9 months and continued through 19 months. Survival was not affected by fertility level. Acacia koa had the lowest survival, 85% (F,) and 88% (F,). Acacia koa also clearly responded to the 4 month fertilization (Fig. 2). Eight species had a significant response to the F, treatment during this 19 month period: A. angustissima, A. aulacocarpa, A. auriculiformis, A. cincinnata, A. crassicarpa, A. implexa, A. koa, and A. mangium(Table 3). Acacia angustissimaand A. koa responded to higher fertility with increased HT, DBH, BD, and VOL, A. implexa with increased BD and VOL growth, and A. angustissimaand A. aulacocarpa with increased average number of stems per tree (A. angustissimafrom 2.3 (F,) to 7.6 (F,) and A. aulacocarpa from 1.4 (F,) to 2.5 (F,)). Acacia aulacocarpa, A. auriculiformis, A. cincinnata, A. crassicarpa, and A. mangium all had significantly greater volume in the F, treatment (Table 3). Each species doubled or tripled its volume 182 T.G. Cole et al. / Forest Table 4 Relative volume 1993 growth of 12 Acacia 1 2 3 4 5 6 7 8 9 10 11 12 species, ranked Ecology from best to worst grown 80 11996) 175-186 under low (F,,) and high (F,) Fo F, A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. mangium crassicarpa cincinnate leiocalyx ieptocarpa aulacocarpa auriculiformis orites implexa angustissima polystachya koa with increased nutrition. Based on these results, three species, A. cincinnata, A. crassicarpa, and A. mangium, were ranked the top three species in our trial under F, conditions (Table 4). The same three species, A. cincinnata, A. crassicarpa, and A. mangium, were also ranked the top three performers under F, conditions. 3.2. Soil analysis The lime treatment before planting had significant effect on several soil characteristics (Table 5). The Table 5 Soil characteristics before planting Waiawa. HI. 1993 and Management and after 22 months, fertility -_--. pH in the F, surface horizon CO- 15 cm) increased to 5.60, compared with the initial pH of 4.36 and the F0 plot’s pH of 4.74. Exchangeabie calcium and magnesium levels were significantly higher and exchangeable aluminum and aluminum saturation levels significantly lower in the high fertility treatment. Levels of potassium and phosphorus were not significantly greater in the F, plots compared with the F,, plots, despite the application of P and K at 4 months. Results from the sub-surface soil analysis (30-40 cm) showed no significant differences in pH between the two treatments (Table 5). Exchangeable calcium for two depths and two fertility levels, low (F,) and high (F,), FO O-15 cm pH (1: 1 H,O) P(M.Truog,mgPl-‘) K(cmol, kg-‘) Ca (cmol, kg- ‘1 Mg (cmol, kg- ‘1 Al (cm01 c kg - ’ ) Al saturation (percent) 4.34 2.64 0.24 0.72 0.80 1.91 52 4.74 7.72 0.14 0.54 0.31 2.76 74 a a a a a a a 5.60 9.19 0.16 6.61 0.57 0.29 5b b a a b b b 4.80 7.23 0.08 0.33 0.13 2.31 65 a a a a a a a 4.87 7.63 0.10 1.17 0.21 1.53 32 b a a a b b b are between columns, Acacia species trial, __I--_ Initial 30-40 cm pH (1: I H,O) P(M.Truog,mgPl-‘1 K (cmol, kg- ’ ) Ca (cmol, kg- ‘) Mg (cmol, kg- ’ ) Al (cmol, kg- ’ ) AI saturation (%) Tukey’s Studentized HI. mangium cincinnata crassicarpa aulacocarpa angustissima auriculiformis leiocalyx implexa leptocarpa odes koa polystachya Variable Comparisons levels, Waiawa, F, Range Test, (Y = 0.05; 22 month results are the mean of three replications. T.G. Cole et al./Forest Ecology and magnesium levels were, however, significantly higher in the F, plots. Extractable aluminum and aluminum saturation levels were significantly lower in the subsoil, which may be important in drought tolerance. 3.3. Insect damage Damage caused by the Chinese rose beetle was confined mainly to phyllodes of the affected species. In general, damage was high in the early months and declined after the insecticide treatment. Feeding, however, did continue on most species throughout the duration of the experiment. Several species seemed to be naturally resistant to beetle attack: A. angustissima, A. implexa, A. koa, and A. orites. Species heavily attacked by the beetle were A. aulacocarpa, A. auriculiformis, A. cincinnata, A. crassicarpa, A. leiocalyx, A. leptocarpa, and A. mangium. Patterns of herbivory varied by species and time period. No significant correlation was found between fertility level and severity of attack. A. Acada autaswca~ --a- 0 5 10 15 Atxcia 0 5 10 15 F, F. 20 kptocafpa - F, -+- F. 20 Months Fig. 3. Phyllode damage for A. auhcocarpa and A. leptocarpa by fertility level, Waiawa, Oahu, HI, 1993. The asterisk indicates significance at (Y= 0.05). and Management 80 (1996) 175-186 183 aulacocarpa, however, did consistently have more damage in the F, treatment, but the damage was only significant during three of the seven time periods (Fig. 3). By contrast, A. leptocarpa had greater damage occurring in the F,, treatment in five of the seven time periods, but the damage was only significantly different during one of those five periods. 4. Discussion Three species, A. cincinnata, A. crassicarpa, and A. mangium, are best adapted to conditions at Waiawa (Table 4). The three were top performers under both high and low fertility treatments. Acacia mangium has been planted throughout the tropics and is considered to be an excellent species for reclamation of denuded sites (National Research Council, 1983). Acacia mangium was also the best performer at Waiawa, where under minimum fertilization (F,) it grew at a rate of 2.5 m year-’ in HT and 2.8 cm year-’ in DBH (Table 3). Fertilized and limed (F,) A. mangium grew at a rate of 3.3 m year- ’ in HT and 4.0 cm year-’ in DBH. These growth rates are average for A. mangium, for which rates have been reported of 1.9-5.2 m year- ’ in I-IT and 2.0-4.5 cm year-’ in DBH depending on climatic conditions and site quality (National Research Council, 1983). Acacia cincinnuta and A. crassicarpa, while not as widely planted worldwide as A. mangium, are gaining recognition and have shown promise as reforestation species. In our trial, growth rates for A. cincinnata were higher than those reported for A. cincinnata grown in Australia (Ryan and Bell, 19891, the People’s Republic of China (Minquan et al., 19891, Thailand (Pinyopusarerk, 1989), and Zimbabwe (Gwaze, 1989). Average HT growth for A. cincinnata at Waiawa under F,, conditions was 2.5 m year- ’ and in F, plots, 3.3 m year-’ (Table 3). Average DBH growth was 3.4 cm year-] in the F, treatment and 2.1 cm year-’ in the F, treatment. Of the four countries listed above the best A. cincinnata growth (2.9 m year- ’ inHTand2.1 cmyear-‘in DBH) was reported from Thailand. Growth of A. crassicarpa at Waiawa was average compared with growth achieved elsewhere. Acacia crassicarpa’s HT growth was 3.5 m year- ’ in the F, 184 T.G. Cole et al./ Forest Ecology treatment and 2.8 m year ’ in the F, treatment. DBH growth rates were 3.9 cm year’ in the F, treatment and 3.0 cm year-’ in the F, treatment (Table 3). Growth rates reported for A. crussicurpa worldwide range from 0.6 to 5.3 m year’ in HT and 1.1-5.1 cm year-’ in DBH (Boland and Turnbull, 1989), depending on climate and site quality. Several of the remaining nine species responded to increased nutrition by doubling or tripling their total volume. The reaction of A. angustissimu to the F, treatment was a sevenfold increase in VOL and a doubling of I-IT, DBH, and BD growth. Much of the increase in VOL shown by A. ungustissimu can be attributed to the increase in the number of stems per tree associated with increased fertility (2.8 vs. 7.6). Acacia polystuchyu grew poorly in both treatments, and although it doubled in size with increased nutrition, it did not perform as well as the other species. Acacia kuu, growing in two phosphorusrate trials at Waiawa were out-performed by A. kou grown in our speciestrial (Table 6). The best HT growth rate for A. kou in these experiments was achieved by a Pacific Palisadesprovenance on Wahiawa soil fertilized at a rate of 600 kg P ha-‘, a total of 1.6 m year’ (Ikawa, 1994). Our best growth rate in the speciestrial for fertilized A. kou (F,) was substantially higher, 2.2 m year-‘. Koa in our F, treatment was planted on soil that was plowed, limed, roTable 6 Growth rates of various provenances of A. koa planted at Waiawa, Site Soil series a Provenance Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Waiawa Leilehua Leilehua Leilehua Leilehua Leilehua Leilehua Leilehua Wahiawa Wahiawa Wahiawa Wahiawa Wahiawa Wahiawa Kaumana. Hl Kaumana, HI Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Pacific Palisades, Kaumana, HI a b ’ d F, F, 1 1 1 1 I 2 2 2 2 2 3 and Management 175-186 totilled, and fertilized with -NPK Andymicronutrients together with rhizobia and VA mycorrhiza. The site used for the phosphorus trial was cleared with a bulldozer, planted, fertilized with phosphorus, and periodically mowed. The growth rate of A. koa in our F, treatment was almost identical to that achieved by koa that received no inputs in the phosphonls experiments (0.9 m year ’ vs. 0.8 andO. m year ‘i. The effect of management and nutrition on the growth of koa is even more apparent when comparing the growth of our Kaumana A. koa provenance with Kaumana koa established in a seed source experiment at Waiawa (Table 6, Waiawa 3). Kaumana was one of the worst growing provenances in the seed source trial. In this trial koa grew at a rate of only 0.1 m year-- ’ with minimal inputs. The provenance was heavily attacked by insects, which may have contributed to its poor growth. In comparison, our Kaumana koa grew at a rate of 0.9 m year- ’ with minimal fertilization (F,) on rhe Leilehua soil and was highly resistantto insects. The main difference between these two experiments was the plowing our site received before planting. The soil at Waiawa is extremely hard and compacted from years of pineapple cultivation. The importance of tillage to improve soil structure and consequently plant growth is well documented and appearsto be a major factor in improving koa growth at Waiawa. Differences in Hl in four studies Fertilization Oahu Oahu Oahu Oahu Oahu Oahu Oahu Oahu Oahu Oahu 80 (1996) (kg ha ’J N P K 20 c 20 0 0 0 0 0 0 0 0 0 0 d 20 220 0 150 300 600 1400 0 150 300 600 1400 20 93 0 0 0 0 0 0 0 0 0 0 Leilehua: claey, oxidic, isohyperthermic, Ustic Kanhaplohumuh. Wahiawa: fine, kaolin&, Data from: Waiawa 1 and Waiawa 2, J. Silva (unpublished data, 1993); Waiawa 3, Ikawa Fertilized with NPK, lime, and micronutrients. Fertilized with 168 g per tree of 14-14- 14 slow release fertilizer in two applications. isohyperthermic, (1994k Waiawa HT MAI -T-----(rn year ’ i --___ _---0.9 2.2 0.7 I .o 1.0 I.? 1.2 0.8 1.1 1.4 I .6 1.5 0.1 Rhodic Eutrustox. PO and Waiawa F,, Table 3. T.G. Cole et al./Forest Ecology the growth of koa attributed only to soil type appear to be minimal. Unfertilized and minimally fertilized Pacific Palisades koa grew identically on both the Leilehua and Wahiawa soil series in the duplicate phosphorus experiments. Even a poorly adapted provenance like the Kaumana seed source can be successfully grown and outperform the best adapted provenance by utilizing proper management techniques. Phyllode damage caused by the Chinese rose beetle occurred on the top three performing species in our trial, A. mangium, A. crassicarpa, and A. cincinnata. The effect of this loss of photosynthetic tissue on growth was not addressed in this study, but we feel our growth results are probably conservative. The beetle may have preferentially attacked A. aulacocarpa in the F,, treatment, but the data were extremely variable. None of the remaining species showed significant trends of beetle attack. Acacia mangium, A. crassicarpa, and A. cincinnata were the most productive species under acid, high aluminum soil conditions, as well as under well-fertilized conditions. Growth of these species should be evaluated on extremely degraded sites using the establishment procedures we have described in order to explore further their adaptability. These species seem promising as sources of wood and possibly as soil improvers due to their N-fixing characteristics. Careful study and research should be exercised, however, before planting exotics in order to prevent their becoming harmful pests. The fertility and management requirements of A. koa should be further studied to identify and quantify specific effects of site preparation, liming, phosphorus, and potassium on its growth. Our experiment was not designed to test for these effects, but our results indicate that dramatic growth increases can be achieved by increasing nutrition and management. Subsequent trials should be established on several soil types and at different elevations. If our results are confirmed, then the fact that fertility management can double the growth rate of koa could be an important finding. Acknowledgments We thank the Hawaii State Department of Corrections and staff of the Waiawa Correctional Facility and Management 80 (1996) 175-186 18.5 for supplying the site for this research project. The Australian Tree Seed Centre, CSIRO Division of Forestry and Forest Products, Canberra, Australia supplied the Acacia seeds from Australia and PapuaNew Guinea for this experiment. The Nitrogen-Fixing Tree Association (NFTA), Maui, Hawaii, supplied the A. angustissima seeds. Rhizobia used in inoculating the seedlings were supplied by H. Keyser, of the University of Hawaii’s Nitrogen-fixation by Tropical Agricultural Legumes Project (NiffAL), on Maui. The VA mycorrhiza was supplied by M. Habte, University of Hawaii at Manoa. We also thank Craig Whitesell, who helped design the experiment and choose the species to be tested. References Boland, D.J. and Turnbull, J.W., 1989. Australian tree species for fuelwood and agroforestry in China, Kenya, Thailand, and Zimbabwe. In: D.J. Boland (Editor), Trees for the TropicsGrowing Australian Multipurpose Trees and Shrubs in Developing Countries (Chap. 1). ACIAR Monogr. No. 10, Australian Centre for International Agricultural Research, Canberra, Australia, pp. 12-20. Bums, R.M. and Honkala, B.H. (Technical Coordinators), 1990. Silvics of North America: 2. Hardwoods. Agriculture Handbook 654. US Department of Agriculture, Forest Service, Washington, DC, Vol. 2, 887 pp. Carlson, N.K. and Bryan, L.W., 1959. Hawaiian timber for the coming generation. Trustees of Bishop Estate, Honolulu, Hawaii, 112 pp. Giambelluca, T.W., Nullet, M.A. and Schroeder, T.A., 1986. Rainfall atlas of Hawaii. Rep. R76, Water Resources Research Center, University of Hawaii at Manoa, Honolulu, HI, 267 pp. Gwaze, D.P., 1989. Growth and survival of Australian tree species in field trials in Zimbabwe. In: D.J. Boland (Editors), Trees for the Tropics-Growing Australian Multipurpose Trees and Shrubs in Developing Countries (Chap. 12). ACIAR Monograph No. 10, Australian Centre. for International Agricultural Research, Canberra, Australia, pp. 128- 138. Ikawa, H., 1994. Unpublished data. Department of Agronomy and Soils, University of Hawaii. Meyer, H.A., 1944. A correction for a systematic error occurring in the application of the logarithmic volume equation. Res. Pap. 7, Pennsylvania State Forestry School, 3 pp. Milliken, G.A. and Johnson, D.E., 1992. Analysis of Messy Data. Chapman and Hall, New York, 473 pp. Minquan, Y., Jiayu, B. and Yutian, 2.. 1989. Tropical Australian Acacia trials on Hainan Island, People’s Republic of China. In: D.J. Boland (Editor), Trees for the Tropics-Growing Australian Multipurpose Trees and Shrubs in Developing Countries (Chap. 8). ACIAR Monogr. No. 10, Australian 186 T.G. Cole et al./ Forest Ecology Centre for International Agricultural Research, Canberra, Australia, pp. 88-96. Morrison, R.J., 1988. The occurrence and distribution of acid soils in Oceania. In: Proc. of the 3rd Int. Soil Management Workshop, on the Management and Utilization of Acid Soils of Oceania, 2-7 February 1987, Republic of Belau. Agricultural Experiment Station, College of Agriculture and Life Sciences, University of Guam, pp. 10-17. Mueller-Dombois, D., 1981. Fire in tropical ecosystems. In: Fire Regimes and Ecosystem Properties, hoc. of the Conference. Gen. Tech. Rep. WO-26, US Department of Agriculture Forest Service, Washington, DC, pp. 137-176. National Research Council, 1983. Mangium and Other Acacias of the Humid Tropics. Innovations in Tropical Reforestation. National Academy Press, Washington, DC, 62 pp. Pinyopusarerk, K., 1989. Growth and survival of Australian tree species in field trials in Thailand. In: D.J. Boland (Editor), Trees for the Tropics-Growing Australian Multipurpose Trees and Shrubs in Developing Countries (Chap. I 1). ACIAR Monogr. No. 10. Australian Centre for International Agricultural Research, Canberra, Australia, pp. 108-127. Ryan, P.A. and Bell, R.E., 1989. Growth, coppicing and flowering and Management 80 (19961 175-186 of Australian tree species in trials in Southeast Queensland, Australia. In: D.J. Boland (Editor], Trees for the Tropics-Growing Australian Multipurpose Trees and Shrubs in Developing Countries (Chap. 5). ACIAR Monogr. No. 10. Australian Centre for International Agricultural Research, Carberra, Australia, pp. 49-65. Scowcroft, P.G. and Adee, K.T.. 1991. Site preparation affects survival, growth of koa on degraded montane forest land. Res. Pap. PSW-205, Pacific Southwest Research Station, US Department of Agriculture Forest Service, Berkeley, CA, 7 pp. Silva, J.A., 1994. Unpublished data, Department of Agronomy and Soils, University of Hawaii. Statistical Analysis Systems Institute Inc., 1985. SAS User‘s Guide: Statistics, Version 5 edn. SAS Institute Inc.. Cary. NC. 956 pp. Tumbull, J.W. (Editor), 1987. Australian acacias m developing countries. ACIAR Proc. No. 16, Australian Centre for Intemational Agricultural Research, Canberra, Australia. Whitesell, CD.. 1964. Silvical characteristics of koa (A. kou Gray). Res. Pap. PSW-16, Pacific Southwest Forest and Range Experiment Station, US Department of Agriculture Forest Service, Berkeley, CA, 12 pp.