Nursery and Field Establishment Techniques to Improve Seedling

Nursery and ﬁeld establishment techniques to improve seedling growth of three Costa Rican hardwoods

KEVYN ELIZABETH WIGHTMAN1, TED SHEAR1, BARRY GOLDFARB1 and JEREMY HAGGAR2 1North Carolina State University, Department of Forestry, Raleigh, NC, USA (E-mail: [email protected]); 2Organization for Tropical Studies, San Pedro, Costa Rica

Accepted 11 August 2000

Key words: Calophyllum brasiliense, compost, Cordia alliodora, fertilization, herbicide, Hyeronima alchorneoides, native species, reforestation, root trainers, weed control

Abstract. Seedlings of three economically important and ecologically different native hardwoods, Cordia alliodora (Boraginaceae), Hyeronima alchorneoides (Euphorbiaceae), and Calophyllum brasiliense (Clusiaceae), were grown in Rootrainers (a book-type container), paper pots, and plastic bags filled with either soil, soil with fertilizer, or compost substrates. After transplanting in the field, treatments with and without fertilizer and herbicide were applied to all nursery stock types. In the nursery, species responded primarily to substrate type. Cordia grew better in bags of soil with NPK fertilizer and compost than in unamended soil, probably responding to higher nitrogen availability. Despite large treatment differences at planting, there were no significant differences in plant size after one year in the field between book containers and bags. The exception were stump plants that were shorter and had higher mortality. Hyeronima grew better in compost than in soil with or without fertilizer, probably responding to higher phosphorus availability and lower bulk density of the compost. Plants produced in compost were also bigger after one year’s field growth. Plants produced with soil or in paper pots had higher mortality. Calophyllum grew less in compost compared to soil and grew better when micronutrients were added to the compost and soil. In the field, seedling produced in soil or with micronutrients had higher survival or growth, respectively. In general, species grew better with herbicide and fertilizer application after transplanting. However, there were no interactions with nursery treatments. Responses to field treatments were independent and thus additive to the nursery treatments. Differences in species response can be related to biomass allocation patterns and ecology of the species.

Introduction

Reforestation in Latin America is an increasingly popular activity due to the abandonment of unproductive cattle pastures and government incentive programs that support tree planting (Schelhas et al. 1997). To promote use of native species in forestry plantations and increase profitability for farmers in the Atlantic Lowlands of Costa Rica, species screening and provenance trials 76 for reforestation were established (Butterfield 1993, 1995; Butterfield and Fisher 1994). Currently, farmers in this region are almost exclusively planting native species used for furniture and construction wood because of their demonstrated fast growth and good form. Rotation periods of 15 years or less with one commercial thinning can be achieved. In Costa Rica, small holder farmers are the primary commercial reforesters, but the seedlings they obtain from community nurseries are often of poor quality. While research efforts often focus on identification of appropriate genotypes for reforestation, more attention needs to be given to seedling quality. Despite the enormous diversity of tropical tree species, the majority of non-industrial seedlings are produced in the same way, using 500–1500 cm3 perforated plastic (poly) bags with soil. The soil often contains high propor- tions of clay, has poor structure, and is low in plant nutrients. Generally little or no organic matter is incorporated. Due to poor substrates, plant growth is slow, extending into two nursery seasons, raising costs for the nursery, and inevitably reducing plant growth in the field. Plastic bags are used because they are inexpensive and readily available. They can cause root coiling, the spiral growth of roots along the smooth sides and bottom of the bag; this root deformation can cause toppling or basal sweep several years after planting, thus greatly lowering the value of the plantation (Mason 1985; Liegel and Venator 1987; Sharma 1987; Josiah and Jones 1992). Stumps plants (psuedoestacas) are an additional stock type produced in bareroot beds. They are derived from trees usually over 1.5 m tall by trim- ming major portions of the stem and root system (ideally 10 cm shoot and 15 cm root remain) after lifting. They are used for several species including Tectona grandis, Gmelina arborea, Bombacopsis quinata (Lamprecht 1986), and Cordia alliodora in the Atlantic lowland region of Costa Rica (Maroto, pers. comm.). They are popular because they require little maintenance in the nursery and are easy for landowners to transport. Limitations to seedling growth on abandoned pastures or agriculture fields in Latin America include weed competition, soil compaction and low fertility soils. Weed growth is persistent throughout the year due to favorable environmental conditions. Manual weeding with a machete is the most common vegetation control in plantations and weeding frequency depends on the cost and availability of labor (Rheingans 1996). Herbicides and fertilizers are used by the few landowners who have sufficient economic resources and are more commonly used in agriculture than in forestry. Early establishment is important for smallholder landowners in order to reduce weeding costs and possibly reduce time to harvesst. Early plant growth is regulated by the conditions at the planting site, and by the degree to which a plant’s phenotypic characteristics are adapted to a planting site (Burdett et al. 77

1983). High quality seedlings show substantial height growth the first year of planting, thus expressing their full genetic potential (Rose et al. 1990). They capture the site quicker, therefore allowing fuller expression of site potential (Fry and Poole 1980). In contrast, use of poor planting stock can lower plantation survival and growth, increase site maintenance costs, and reduce confidence in reforestation. The use of the most appropriate planting stock can help overcome site limitations, while early and intensive site management can also accelerate seedling growth (Ladrach 1992). Increasing investment in nursery stock relative to investment in site preparation can increase financial returns on overall reforestation investments (South et al. 1993). In order to capitalize on advances made in reforestation with native species, seedling production techniques should be improved. Assessing outplanting performance must be an integral part of defining and adjusting target seedling characteristics. The objective of this research was to determine, for three widely planted tropical species with contrasting ecological characteristics, how different substrates, container types, and container volumes affect seedling growth in the nursery and early growth in the field. Weed control and fertilization were also tested to determine if nursery and field techniques could be integrated to improve seedling growth during establishment.

Materials and methods

The nursery was located at the La Selva Biological Research Station of the Organization for Tropical Studies (OTS) in the Atlantic Lowlands of Costa Rica (10◦26 N, 86◦59 W). The average annual rainfall and temperature are 3900 mm and 24 ◦C, respectively, and elevation averages 40 m amsl (McDade et al. 1994). Three rapidly growing native hardwood species that are found throughout the neotropics and are commonly planted in the Sarapiquí canton were studied. Cordia alliodora (R.P.) Cham. (Boraginaceae) is an early successional species occurring on fertile soils; Hyeronima alchorneoides Fr. Allemao (Euphorbiaceae) is a late successional, canopy emergent species found on old alluvial soils; Calophyllum brasiliense Cambess (Clusiaceae) is an old growth, mid-canopy species found on residual soils (Ultisols).

Nursery study

For each species, 11 treatments were replicated in four randomized complete blocks. Treatments were: large (170 cm3) and small (85 cm3) Rootrainers with two composts; 500 cm3 bags (10×15 cm) with soil and NPK fertilizer, 78

Table 1. Substrate properties for nursery media.

Bulk density CEC (meq/ P K Ca Mg Treatment pH (g/cm3) 100 cm3) (ppm) (ppm) (ppm) (ppm)

Soil 5.7 0.99 14 67 338 1,680 40 Soil with NPK fertilizer 5.2 1.02 15 559 934 1,520 44 ∗ Compost A 7.2 0.45 53 1,935 8,392 420 120 ∗ Compost B 6.5 0.64 75 1,229 12,887 420 243 50% soil with 50% compost A 6.9 0.79 43 973 4,715 446 103 50% soil with 50% compost B 6.3 0.88 36 338 5,247 252 113 CEC = Cation Exchange Capacity ∗ with sulfur, 1 g/l

50% compost with 50% soil, 100% compost, or unamended soil; paper pots with two composts; and 250 cm3 bags with compost. Seeds were collected from phenotypically superior trees previously identified by OTS and each block contained seed from a single mother tree. Seed was directly sown in the containers between April and June 1995, depending on species. Each replicate of each treatment contained 30 plants: 20 for outplanting, 5 for destructive harvesting, and 5 for culling. From the three species combined, 660 trees were destructively harvested and 2,640 trees were planted. An alluvial soil was used in this experiment. For the fertilizer treatment, 10 grams of 10-30-10 N-P2O5-K2O fertilizer were incorporated into each bag of soil before sowing the seed. Compost A, commercially sold for seedling and vegetable production, was a mixture of 50% coconut husk fiber, 25% chicken manure (containing lime to control flies), 15% sugar cane bagasse and 10% charcoal. Compost B was a mixture of two composted fodder grasses, Pennis- etum purpureum (2 cultivars) and Axonopus scoparius. The pH of compost A was 7.7 and of compost B was 7.4, which are considered high for producing hardwood seedlings (van den Driessche 1984a). Elemental sulfur powder was added at 1 g/l to reduce the pH to 7.2 for compost A and 6.5 for B. When composts were mixed with soil, no sulfur was added. Substrates differed in their chemical and physical properties (Table 1). Rootrainers are plastic containers with grooves that direct roots down- ward to an open end to allow air pruning of the roots. Large and small Rootrainers, are book-type containers arranged in two halves connected by a hinge. Small plastic bags were made by cutting the large plastic bags with a heat sealer. Paper pots were locally produced wax-lined, rectangular, open-ended biodegradable containers and were planted with the trees. They were used by an forestry company in southern Costa Rica for production of Gmelina arborea seedlings. To allow air circulation under plants grown in book containers and paper pots, and retain randomization of treatment 79 order, all containers were placed on chicken wire frames 30 cm above the ground. Plastic was placed beneath the bags to avoid air pruning which normally would not occur when plants are placed on the ground. Containers were spaced so that the density of all seedling was equal, approximately 80 plants/m2. For Calophyllum, the addition of micronutrients was tested with two additional treatments: slow-release micronutrients were incorporated into large book containers with compost and plastic bags with soil. The small plastic bag with compost was omitted for this species. For Hyeronima and Cordia, micronutrients were added to all substrates including unamended soil. Micronutrients (Micromax, Sierra Chem. Co., Milpitas, CA) were added at 1 g/l of soil. The fertilizer contained: 15% S, 12% Fe, 2.5% Mn, 1% Zn, 0.5% Cu, 0.1% Bo, and 0.005% Mo. Shoot height and root collar diameter (RCD) were measured on 25 plants from all treatments immediately before planting. Five of these plant were destructively harvested to measure leaf area of fresh leaves with a LiCor (Li-3000) area meter and root length using the line-intercept method of Tennant (1975). The ratio of these two parameters were chosen instead of the commonly reported shoot weight to root weight ratio because the authors believe that they may better reflect the absorptive and evaporative capacity of the plants. Leaves, stems, and roots from these plants were oven-dried separately at 65 ◦C for 3 days and then weighed. Each tissue was digested separately by the method of Parkinson and Allen (1975). Tissue N, P, K, Ca, and Mg were measured on the sulfuric acid/hydrogen peroxide Kjeldahl digests. The N and P concentrations were measured colorimetrically, and K, Ca, and Mg with an atomic absorption spectrophotometer.

Field study

For each species, blocks containing all treatments were planted on the same date: between July and October 1995 during the rainy season. Seedlings were age 90 days for Cordia and Hyeronima and 140 days for Calophyllum.The blocks were located across four sites: La Flaminia, a part of the OTS La Selva Biological Research Station; two sites on a farm adjacent to La Selva; and a farm 5 km from the station. The soils at all locations were classiﬁed as Inceptisols, but varied in fertility status due to past use. All sites were pastures either dominated by gramalota (Axonopus micay), a tall grass found on relatively fertile sites, or ratana (Ischaemum ciliare), a short grass that forms a heavy thatch, and is probably indicative of site compaction. The farm sites were grazed by cattle immediately prior to this study. The La Flaminia site had been grazed within the last ﬁve years and had been pasture for the last 40 years. 80

For Cordia, paper pots and bags of soil treatments were not planted due to poor growth in the nursery and a stump stock type was added. The stumps originated from one-year-old trees (average height 1.6 m) grown in a bareroot nursery bed. Shoots and roots were cut to 18 and 17 cm length, respectively, and were planted the same day that they were cut. A split-split plot field design was used with four blocked replications. The seedlings were planted in four 0.22 ha monospecific blocks at a spacing of 3×3 meters (1,111 trees/ha). Each block was divided into manual weeding and herbicide competition control main plot treatments. Ten trees from each nursery treatment were planted in each of these two competition treatments. Fertilizer was applied to 5 plants from each nursery treatment within each competition treatment. For the fertilizer treatment, forty grams of 10-30-10 N-P2O5-K2O fertilizer were applied once around the base of each tree at 30 days after planting. For the herbicide treatments, plots were kept as weed-free as possible to minimize competition. Herbicide (glyphosate, Roundup) was applied four times during the first year. A 3 m diameter area was totally cleared of vegetation around the trees to protect them before herbicide was applied. Between the four herbicide applications, vegetation on the entire plot was cut back to a 30 cm height with a machete. On the manually cleaned plots, weeding by machete was done approximately six times per year alternating between cleaning 2 m around the tree and then whole site clearing (this is the most common vegetation control in the region). More competition occurred on the manually weeded plots because grasses overgrew the seedlings between weedings. Percent survival, height, and RCD were measured after 1, 2, 3, 6, 9, and 12 months. Tree mortality was scored as unexplained, induced by grazing, or by herbicide. Tree mortality due to accidental cattle grazing or herbicide was treated as missing data.

Data analysis

Treatments were designed to compare a series of a priori established ortho- gonal contrasts. A contrast between Rootrainers filled with compost and bags of soil was made to compare an alternative technology to the conventional method, even though it confounded substrate and container effects (volume differed by 230 cm3). Contrasts were then used to separate specific substrate, container, and volume effects. Substrates were tested by comparing them within the same containers. The two composts (data not shown), were not significantly different, so treatment means were combined. The container type was tested by comparing three containers with similar volumes and containing the same compost substrate: large book containers, paper pots, 81 and small plastic bags. The volume effect was tested between large and small book containers with the same compost substrate. Analyses of variance (GLM procedure) were used to detect significant treatment differences using contrast statements within the treatment groups of interest (SAS Institute 1990). Means were calculated across fertilization and weeding effects because there were no significant interactions between these field treatments and nursery treatments. The PROC MIXED procedure was used for the field data because it best adjusts for missing data (Wolfinger and Chang 1995). Correlations between response variables were calculated across treatments using PROC CORR and are reported as Pearson correlation coefficients.

Results

Cordia alliodora

Nursery results Substrate effects. Plants grown in book containers were signiﬁcantly larger than plants grown in bags of unamended soil. Plants grown in soil with NPK fertilizer obtained the largest mean height, RCD, and total dry weight (Table 2). Leaf area and root length were the same for the fertilizer and compost treatments. Plants grown in unamended soil had only 4% and 7% of the biomass of plants grown with NPK fertilizer and compost, respectively. Leaf N, P and K concentrations were the same for the four treatments in bags. Plants grown in the unamended soil and the compost treatments had higher Ca and Mg leaf concentrations than plants grown with NPK fertilizer. Leaf concentrations of N and K did not correlate with any of the size variables. Leaf concentrations of P, Ca, and Mg were negatively correlated with plant size: P was negatively correlated with total dry weight (r = −0.32), leaf area (r = −0.32) and root length (r = −0.30); Ca was negatively correlated with diameter and height (r = −0.26), and Mg was negatively correlated with total dry weight (r = −0.29), height (r = −0.33), and diameter (r = −0.31).

Container type and volume. Plants grown in book containers obtained, on average, twice the total dry weight and had greater height, diameter, and root length as plants grown in paper pots (Table 2). There were no significant differences between plants grown in planter books and small plastic bags despite the fact that bags were 80 cm3 larger. Seedling size increased with the larger root trainer volume. Root coiling of tap roots in book containers 82 ∗ ∗∗ ∗ =0.06 after 90 days of p ns ns ns ∗∗∗ Cordia alliodora ns ns nsns ns nsns ns ns ns ns ns ns ns ns na na na na ∗∗ ns =0.08 ns ns ns ns ns ns p ) (cm) %N LA:RL %P %K %Ca %Mg 2 ns ns ns ns ns ns 0.05, ns = not significant, na = not analyzed. < p ∗ 0.001, < p ∗∗ ∗∗∗ ns ns ns ns ns ns ns ns ns ns ns 2515 3.3 2.5 1.2 0.7 119 52 240 190 0.5 0.5 2.3 0.27 2.3 3.9 0.29 3.8 3 2.7 1.1 1 ∗∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗ ∗∗ ∗∗∗ ∗∗∗∗∗∗ ∗∗∗ ∗∗∗∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗∗ ∗ ∗∗∗ ∗∗∗ ∗∗∗ (cm) (mm) (g) (cm 0.0001, < p ∗∗∗ + + with compost vs. bags with soil root trainer, compost A + B bag with compost Apaper pot with compost A 14 25 2.4 3.7 0.5 1.1 65 111 191 246 0.6 0.5 na 2.7 0.23 na 3.8 na 2.7 na 1 1 bags Seedling morphology, leaf nutrient concentrations and significance levels of treatment contrasts for root trainer, compost A + B 3 3 3 3 3 the mean of the two composts, Compost vs. soil with fertilizer Small bag vs. large Rootrainer Substrate effects Unamended soil vs. soil with fertilizer 85 cm 250 cm 140 cm Treatment contrasts New vs. conventional technology Rootrainers Treatments500 cm Height RCD TDWT Leaf area Root length50% compost vs. 100% compostVolume and container effects Leaf nutrientLarge vs. concentration small Rootrainers ns ns ns ns Unamended soil vs. compost Paper pot vs. large Rootrainer Unamended soilSoil and NPK fertilizer50% soil and 50% compost100% A compost AOther containers 170 cm 43 61 8 5.7 7.4 1.4 4.5 48 7.1 0.3 6 360 488 30 4.2 650 565 338 131 0.6 0.9 2.6 442 0.4 2.4 0.18 0.16 2.2 4.5 0.8 3.6 0.23 2.9 3.2 2.3 2.1 0.23 3.3 1 3.8 0.8 1.1 3 1.1 Table 2. Least squares means are+ listed. RCD = root collar diameter; TDWT = total dry weight; LA:RT = the ratio of the leaf area to root length. nursery growth. 83 was present in 2% of the plants whereas in both the bags and in paper pots 9% of the plants had coiled tap roots.

Field results Effects of nursery treatments on growth. The significant differences in plant size persisted through 6 months in the field but were not present (with one exception) after one year’s field growth (Table 5). Plant height averaged between 150–200 cm for all of the treatments, except for stumps. Stumps remained shorter (102 cm) than all other treatments and had lower diameter growth. They did not regain their initial height when cut (160 cm) and were only 3 mm larger in diameter than when planted.

Effects of weeding and fertilization. There were no interactions between nursery treatments and weeding or fertilizing. When herbicide was applied to the plots, tree diameter, but not height, increased signiﬁcantly (Table 6). There was a signiﬁcant interaction between weeding and fertilizing for tree size. In plots treated with herbicide, fertilizer increased height and diameter by 19% and 31%, respectively, but there was no response to fertilizer in manually weeded plots.

Mortality. Survival, not including mortality from grazing or herbicide, was equally high for all container grown nursery treatments at about 88%. Stump plants had a signiﬁcantly lower survival rate of 49%, with most mortality within the ﬁrst two months after planting (Table 5). Grazing and herbicide caused an additional 20% mortality. Mortality due to all causes with herbicide damage probably being the main causes was 30% higher in the herbicide treated plots than in the manually weeded plots.

Hyeronima alchorneoides

Nursery results Substrate effects. Plants grown in book containers were significantly larger than plants grown in bags of soil. Plants grown in compost grew best, although leaf area and root length (p = 0.08) were greater in the 50% soil:compost mixture than in pure compost (Table 3). Total dry weight of plants grown with compost was nearly 4 times greater than plants grown with NPK fertilizer. The addition of fertilizer to soil increased total dry weight by 80%, but did not affect height significantly. Leaf N concentrations were greater in compost than in unamended soil. Plants grown in compost or with fertilizer had greater P and K leaf concentrations than plants grown in soil. However, plants grown in compost had lower Ca and Mg concentrations than plants grown in soil with or 84 after ∗ ∗∗ ns ns ∗∗∗ ∗∗ Hyeronima alchorneoides ns ns ns ∗∗ ∗∗∗ ∗∗ ∗ ∗∗ ns ns ns ns ns na na na na na ns nsns ns ns ns ns ns ns ns ns ns ns ns =0.06 ns p ) (cm) LA:RL %N %P %K %Ca %Mg 2 ∗ ∗∗ ∗∗∗ ∗∗ 0.05, ns = not significant, na = not analyzed. =0.06 ns ns ns p < p ∗ ns 0.001, < ∗∗ ∗ p ∗∗ ∗∗ nsns ns ns ∗∗ ∗∗ ∗∗∗∗ ∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗ 1610 3.5 2.4 1.7 0.6 180 66 534 315 0.3 0.2 1.5 0.26 1.5 2.6 0.26 2.8 0.5 0.6 0.6 0.7 ∗∗∗∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗ ∗ ∗∗∗ ∗∗ ∗ ∗∗ (cm) (mm) (g) (cm 0.0001, < + p + ∗∗∗ with compost A + B with compost A + B w/compost vs. bags with soil Rootrainer bag with compost Bpaper pot with compost B 15 15 3 3.5 1.3 1 145 146 441 250 0.3 0.7 1.7 na 0.26 na 2.9 na 0.4 na 0.5 na bags Seedling morphology, leaf nutrient concentrations and significance levels of treatment contrasts for Rootrainer 3 3 3 3 3 the mean of the two composts, 50% compost vs. 100% compost ns ns ns 85 cm 250 cm Unamended soil vs. compost Paper pot vs. large Rootrainer Treatments500 cm Height RCD TDWT Leaf area Root lengthSubstrate effects Unamended soil vs. soil with fertilizerCompost vs. soil with fertilizer Volume and container effects Leaf nutrient ns Large concentration vs. small root trainer Small bag vs. large Rootrainer 140 cm Treatment contrasts New vs. conventional technology Rootrainers Unamended soilSoil with NPK fertilizer50% soil witth 50% compost100% B compost BOther containers 170 cm 24 11 4.8 7 2.9 3.4 2.1 24 0.9 0.5 414 4.8 95 3.2 49 714 423 344 293 c 0.7 0.2 639 0.2 1.4 b 1.3 0.25 1.1 0.28 2.9 0.6 0.14 2.4 0.5 1.4 1.6 0.7 0.26 0.7 0.4 2.8 0.6 0.7 0.4 0.5 Table 3. Least squares means are+ listed. RCD = root collar diameter; TDWT = total dry weight; LA:RT = the ratio of the leaf area to root length. 90 days of nursery growth. 85 without fertilizer. Leaf P concentrations were positively correlated with total dry weight (r = 0.36) while Ca concentrations were negatively correlated (r = −0.45).

Container type and volume. Plants grown in small bags and in book containers obtained a greater total dry weight, root length, and diameter than plants grown in paper pots. Plant height was the same for the three treatments. Plants grown in the large book containers obtained the same size as the plants in small bags although leaf area was signiﬁcantly greater, despite the fact that bags were 80 cm3 larger. Plants grown in large book containers were larger than those grown in small book containers.

Field results Effects of nursery treatments on growth. After one year, plants grown in compost retained their size advantage (Table 5). Plants grown in book containers grew better than plants grown in soil. Plants grown with fertilizer had greater absolute height growth than plants grown in unamended soil, but ﬁnal height and diameter were not signiﬁcantly different. Plants grown in large book containers remained larger in height and diameter than plants grown in small book containers or in paper pots. Plants grown in small bags of compost were as large as plants grown in large book containers.

Effects of weeding and fertilization. Plants grew best under intensive management. There were no interactions between nursery main effects and fertilizing or weeding, nor between fertilizer and herbicide. Weed control and fertilization had significant effects on final height and diameter (Table 6). On average, plants were 20% taller and 29% greater in diameter in plots that received herbicide than in manually weeded plots. When fertilized, average plant height and diameter increased by a smaller but significant amount.

Mortality. The average survival rate of plants grown in soil (85%) and in paper pots (80%) was signiﬁcantly lower than the survival rates (average 95%) of plants grown in soil with fertilizer, bags of compost or book containers. Grazing increased the total mortality to 25%. Survival was not affected by fertilization or herbicide.

Calophyllum brasiliense

Nursery results Substrate effects. No differences were found in seedling size between plants grown in 170 cm3 book containers and those in bags of soil. The addition of NPK fertilizer to the soil did not increase growth, except for leaf 86 area (Table 4). Plants grown in compost were shorter, had smaller total dry weight and leaf area than other treatments. Root lengths did not differ signiﬁ- cantly among the treatments. Plant total dry weight and diameter increased signiﬁcantly with the addition of micronutrients to soil. The addition of micronutrients to compost in book containers increased plant height and diameter, reduced total root length, but did not affect total dry weight. Plants grown with compost had higher leaf concentrations of N and K than the other substrate treatments. Calcium concentrations were greatest in plants grown with NPK fertilizer and with micronutrients. Plants grown with NPK fertilizer had higher leaf concentrations of N, P, K and Ca than plants grown in unamended soil. There was a negative correlation between total dry weight and leaf concentrations of nitrogen (r = −0.56), phosphorus (r = −0.28), potassium (r = −0.28), while calcium was positively correlated (r = 0.47).

Container volume and type. Plant growth was not affected by container volume, although leaf nutrient concentrations of N and P were smaller in the smaller containers (Table 4). There were no differences between the plants grown in book containers and in paper pots. Root coiling was most prevalent in plants grown in plastic bags. Root coiling of tap root was present in 8% of the plants grown in book containers, 15% of those grown in paper pots and 23% of those in bags.

Effects of nursery treatments on growth. After one year, plant heights and diameters were not signiﬁcantly different for most of the contrasts of interest, despite some initial differences among nursery treatments (Table 5). The only signiﬁcant difference was between the diameters of plants grown in paper pots (9 mm) and those grown in large book containers (11 mm). Plants grown in 100% compost grew less than those grown in soil-based media. Height growth rates of plants grown with micronutrients whether in compost or in soil were greater than those grown in substrates without the addition of micronutrients.

Effects of weeding and fertilization. During the ﬁrst six months, plant height and diameter increased with fertilization, but this effect did not persist after 9 months. Plant height was greater in manually weeded plots (Table 6). There was no interaction between weeding and fertilizing.

Mortality. Plants produced in bags of soil or soil amended with fertilizer or micronutrients had an average survival of 93% whereas plants grown in compost had a significantly lower survival rate of 71%. Mortality due to 87 ns ns ns ∗∗ =0.08 p after 140 days ∗ ∗ ns ns ns =0.06 ns p ns ∗∗ ∗∗ ∗∗ =0.07 p ns ns ns Calophyllum brasiliense ns ns ns ns ns ∗∗ ∗ ∗∗∗ ∗∗ ∗∗ ∗∗∗ ∗ ∗∗ ∗ ∗ ∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗∗∗ nsns ns ) (cm) %N LA:RL %P %K %Ca %Mg 2 ns ns =0.08 ns p 0.05, ns = not significant, na = not analyzed. < ∗∗ p ns ns ∗∗ ∗∗ ∗ ns ns 0.001, < p ∗ ∗ ∗∗∗ ∗∗∗ ∗ nsns ns ns ns ns ns ns ns ns ns ns na na na na na 2020 2.6 2.6 1.5 1.3 108 90 280 223 0.4 0.4 1.5 0.20 1.3 0.12 1.6 1.3 0.4 0.4 0.2 0.2 (cm) (mm) (g) (cm ∗∗ 0.0001, < + p + ∗∗∗ with compost A + B w/compost A + micronutrients 26 3.5 1.5 119 128 1.1 1.3 0.09 1.5 0.4 0.1 with compost A + B with compost vs. bags with soil ns ns ns ns ns ns bag with soil + micronutrientsRootrainer paper pot with compost ARootrainer 22 4.2 2.1 21 123 2.8 171 1.4 0.7 98 1.0 0.06 225 0.6 0.4 0.5 na 0.1 na na na na bags Seedling morphology, leaf nutrient concentrations and significance levels of treatment contrasts for Rootrainer 3 3 3 3 3 3 the mean of the two composts, Unamended soil vs. compost 50% compost vs. 100% compostCompost vs. soil with fertilizer Soil with micronutrients vs. soil without ns ns ns ns ns ns ns ns Volume and container effects Large vs. small Rootrainers Treatment contrasts New vs. conventional technology Rootrainers Other containers 170 cm 170 cm Compost with micronutrients vs. compost without Paper pot vs. large Rootrainer Unamended soilSoil with NPK fertilizer50% soil with 50% compost100% A compost A500 cm 18 20 20 2.6 2.8 2.8 1.3 1.8 18 1.6 81 2.6 130 103 1 260 240 273 60 0.3 0.5 0.4 1.5 1.1 177 0.12 1.0 0.12 0.06 1.4 1 0.3 0.6 0.4 1.5 0.5 0.16 0.4 0.1 2.1 0.1 0.1 0.4 0.2 Treatments500 cm Height RCD TDWT Leaf areaSubstrate effects Root lengthUnamended soil vs. soil with fertilizer Leaf nutrient concentration ns ns ns 85 cm 140 cm Table 4. of nursery growth. Least squares means are+ listed. RCD = root collar diameter; TDWT = total dry weight; LA:RT = the ratio of the leaf area to root length. 88 ∗ ∗ ∗ ns =0.07 p ∗∗ ∗ ns ns ns ns ns ns ns ns ns ∗ ∗ ∗ ns ns ns ns ns ns ns ns ns ns ns ns ns ns ∗ =0.07 ns ns ns ns ns ns p ∗ ns ns ns ns ns ns ns ns =0.09 p ∗ ns =0.07 p ∗ ∗∗∗∗∗∗ ∗ ∗ ∗∗ ∗∗ ∗ na na na na na na na na na na ∗∗ ∗∗ ∗∗ ∗∗∗ ∗ 0.05, ns = not significant, na = not analyzed. < p ∗ ∗∗ ∗∗ 0.001, < p Cordia alliodora Hyeronima alchorneoides Calophyllum brasiliense ∗∗ ns ns ∗ ns ns ns ns ns ns ns ns ns ns na na na na na ns ns ns ns ns ns ns ns ns ns < 0.0001, Final Grow/day Final Grow/day (%) Final Grow/day Final Grow/day (%) Final Grow/day Final Grow/day (%) 102 (15) 0.20 25 (2) 0.01 49 na na na na na na na na na na 165 (18) 0.31 34 (5) 0.07 81 119 (7) 0.31 26 (1) 0.065 96 83 (5) 0.17 11 (1) 0.021 83 161 (23) 0.33 33 (7) 0.06 83 92 (6) 0.23 22 (1) 0.053 90 83 (5) 0.17 10 (1) 0.021 76 p ∗∗∗ micronutrients na na na na na na na na na na 100 (7) 0.23 12 (1) 0.027 88 + + + vs. bags na na na na na Seedling size, growth increment, survival and treatment contrasts after 12 months of field growth for three hardwood species. bagpaper potsRootrainers 151 (18) na 0.28 na 29 (4) 0.06 na 89 na 116 (6) na 0.28 84 (5) 25 (1) 0.2 0.062 20 (1) 94 0.049 na 80 na 76 (6) na 0.15 9 (1) na 0.017 na 72 Rootrainers bags Rootrainers 3 3 3 3 3 3 mean of two composts, Small bag vs. large Rootrainers Stumps vs. all treatments Volume and container effects Large vs. small Rootrainers Paper pot vs. large Rootrainers Compost w/micronurients vs. compost w/o na na na na na na na na na na ns Soil with micronutrients vs. soil without na na na na na na na na na na ns Unamended soil vs. compost ns ns ns ns ns Compost vs. soil with fertilizer ns ns ns ns ns 50% compost vs. 100% compost ns ns ns ns ns ns ns ns ns ns ns Substrate effects Unamended soil vs. soil w/fertilizer ns ns ns ns ns ns Stumps Treatment contrasts New vs. conventional technology Rootrainers 250 cm 140 cm 170 cm 170 cm SoilSoil with NPK fertilizer50% soil, 50% compost100% compostSoil with micronutrientsOther containers with compost 85 cm 196 (21) 196 (12) 0.25 0.38 39 (5) na 203 (24) 40 (4) na 0.07 0.37 0.09 na 41 90 (4) na 90 na 0.08 99 (4) na 147 (8) na 0.25 90 0.36 na 137 (10) 23 (1) na 32 (1) 0.33 na 0.055 0.078 na 31 (2) 79 (8) 94 97 0.074 na 89 0.2 (6) 89 (6) 98 0.19 18 na (1) 0.19 72 12 (6) (1) 0.042 11 (1) na 0.025 0.15 0.024 85 89 9 (1) na 77 87 (6) 0.018 92 (7) 0.19 65 0.22 11 (1) 10 0.022 (1) 0.022 95 90 Treatments500 cm Height (cm) Diameter (mm) Survival Height (cm) Diameter (mm) Survival Height (cm) Diameter (mm) Survival Table 5. Least squares means are+ listed. Std. errors in ( ). 89

Table 6. Effects of weeding and fertilization on height and diameter after 12 months of ﬁeld growth.

Cordia alliodora Hyeronima alchorneoides Calophyllum brasiliense Weed control Weed control Weed control Fertilizer Herbicide Manual Mean Herbicide Manual Mean Herbicide Manual Mean

Height (cm) With 191 (17) x 154 (14) j 172 (11) a 123 (8) 105 (7) 114 (5) a 80 (4) 87 (2) 83 (2) a Without 158 (16) y 164 (14) j 161 (11) a 115 (7) 94 (5) 105 (5) b 81 (4) 88 (2) 85 (2) a Mean 175 (14) a 159 (12) a 119 (5) a 99 (4) b 80 (3) b 88 (2) a

Diameter (mm) With 44 (4) x 28 (3) j 36 (3) a 29 (1) 23 (1) 26 (1) a 10 (0.4) 10 (0.3) 10 (0.2) a ∗ Without 35 (4) y 29 (3) j 32 (2) b 26 (1) 21 (1) 24 (1) b 11 (1) 10 (0.3) 10 (0.3) a Mean 39 (3) a 29 (3) b 27 (1) a 22 (1) b 11 (0.3) a 10 (0.1) a Treatment main effects are separated using a, b. Treatment interactions when present are separated using x, y, j. ∗ indicates a p value of 0.07. Least squares means are listed. Standard errors are in parenthesis. unexplained causes averaged 20%. This was equally partitioned between the plots that were manually weeded and those that received herbicide. Mortality due to herbicide or grazing was only 1%, with no difference between the vegetation management regimes.

Discussion

Cordia alliodora

Plants grown in both the fertilized substrate and in compost grew quickly and, based upon height and root development, were ready for outplanting after two months. In contrast, plants grown in unamended soil were less than 10 cm tall after five months in the nursery, and thus were not planted in the field. The negative correlations between P, Ca, and Mg concentrations in the leaves and total dry weight but positive response to fertilizer indicate that N, K, or both were limiting growth. This may be an example of the carbo- hydrate dilution effect where the increase of one limiting nutrient causes increased plant growth that results in the decrease in tissue levels of other non-limiting nutrients. Nitrogen supply was not determined in the substrates, but P and K supply were higher in the compost than in soil with fertilizer. In the compost, N supply may have been lower or N was bound in organic complexes rendering it unavailable to plants. This is a common problem with organic substrates with a high carbon:nitrogen ratio (Rose et al. 1995). Cordia 90 was probably N limited in this experiment, and thus the growth responses were primarily due to the different N concentrations in the substrates. Leaf concentrations were lower than typically found in leaves of trees growing in plantations in the Atlantic Lowlands of Costa Rica where Cordia is considered a N demanding species (Bergmann et al. 1994). Although plants produced in book containers were smaller at planting, they grew as well as the larger plants produced with NPK fertilizer and in bags of compost. They also had a lower leaf area to root length ratio than the larger plants grown in bags with fertilizer or compost; but during the first year’s growth their relative growth rate was significantly higher than these plants. The relative growth, that is growth proportional to original size, correlated negatively to the ratio of leaf area to root length (r = −0.59), indicating an advantage for plants with more roots than leaf area. Plants grown in book containers also grew better and had higher survival than stump plants. Even though it was raining almost daily when the study was planted, stumps died probably to due to insufficient nutrient reserves to recover and compete with aggressive weeds. The common recommendations to use stump plants (Calvo and Melendez 1999) should be viewed cautiously. Plants benefited from competition control. The interaction between weeding and fertilization clearly demonstrates that there was little benefit to fertilizing plants in weedy plots. When weed competition was high, trees did not respond to the addition of fertilizer. Reducing weed competition probably influenced nutrient availability to this species. The leaves of the trees in the plots that received herbicide were dark green, whereas trees were less green in manually-weeded plots. The lack of interaction between nursery treatments and weeding and fertilizing indicates that the small plants produced in book containers did not require more intensive field management than larger planting stock types, a commonly held belief. They also had fewer root deformities than plastic bags, a potential long-term advantage (Burdett et al. 1986).

Hyeronima alchorneoides

In the nursery, total dry weight, leaf area, and leaf nutrient concentration of P and K were greater in plants grown with fertilizer or compost than those grown in unamended soil. Availability of P and K were greater in the soil with fertilizer and in compost than in the unamended soil (Table 1). The positive correlation between leaf P concentration and plant size and the high total P content of plants grown in compost suggest that Hyeronima responded to the P availability in the substrates. Nevertheless, plants produced in compost grew considerably larger than those produced in soil with fertilizer. This may be due to the higher P availability in the compost than in the soil with 91 fertilizer or due to the better biological and physical properties of the compost substrate. In the field, final plant size was largely determined by initial plant size. There were significant positive correlations between the final diameter and initial plant size (total dry weight, r = 0.37, leaf area, r = 0.51, and root length, r = 0.30). The plants that grew best were those which had been produced in book containers or in plastic bags with organic substrates. Root mass was not significantly correlated to final plant size. The shoot to root ratio was positively correlated to final height and diameter in the field (r = 0.55). The plants in plastic bags retained their initial size advantage possibly due to their large leaf area or greater shoot mass. Hyeronima responded to the fertilizer on both herbicide treated and hand weeded plots. The high fine root density of this species may enable it to effec- tively compete with the vegetation for the applied nutrients even in weedy plots.

Calophyllum brasiliense

Calophyllum did not increase growth with the addition of NPK fertilizer to the soil and grew significantly worse with compost. Although nutrient availability was high in the compost, and the bulk density was lower than in the soil substrates, the pH may have been too high, causing a micronutrient deficiency. This is suggested by the addition of micronutrients to the soil (pH 5.7) that resulted in increased total dry weight while the addition to compost did not (pH 6.5). Organic matter can complex metallic cations such as Fe, Mn, and Zn, rendering them unavailable to plants, or can produce molecules that chelate micronutrients (van den Dreissche 1984a). The new leaves of many plants grown without micronutrients were chlorotic, possibly indicating an Fe deficiency. There were weak positive correlations between final height and both diameter and leaf area (r = 0.28, 0.34, p = 0.07, respectively). Plants with micronutrients had significantly greater leaf area:root length ratios. The effect of relieving micronutrient deficiencies seemed to cause a change in allocation between roots and leaves and, in turn, may have increased growth.

Species comparisons

The three species exhibited a range of responses underlining the importance of unique nursery management for each. Substrate quality and container type greatly inﬂuenced seedling growth. The same substrates are not recom- mended for the three species because nutrient demands and biomass allocation appear to be different among them. 92

Cordia displayed the largest differences in seedling development. The responses were primarily attributed to nutrient availability because fertilizer dramatically improved growth in soil. Cordia grew best in N rich substrates (compost and soil with NPK fertilizer) and seemed unaffected by the physical characteristics of the substrate. Hyeronima grew best in compost and only slightly better in soil with fertilizer than in soil without fertilizer. Nutrient availability possibly played a secondary role to the better physical and biological characteristics of the compost substrate which enhanced root growth. Both species grew significantly better in nutrient rich substrates than in soil alone. The responses of Calophyllum to the treatments were totally different than those of Cordia and Hyeronima. Plant size did not significantly increase with the addition of fertilizer to soil. In the compost and soil with fertilizer substrates, tissue concentrations of nutrients increased, but plant growth did not increase. The plants may have been affected by the high pH of the compost. Plants had greater total dry weight and N concentrations in the leaf tissues when micronutrients were added to the soil. In the nursery, container type influenced plant development. Through air pruning, plants in book containers had higher root densities and fewer root deformities than plants in bags. Lateral root development has been demonstrated to be related to seedling growth (Kormanik 1986). For Cordia,a tap-rooted species, root density (as root length per volume (cm/cm3)) was 1.4 in large book containers and 1 in small bags. For Hyeronima, a branchy rooted species, root density was 3.3 in large book containers and 1.8 in small bags. For Calophyllum, also a tap-rooted species, root density was 1.5 in book containers and 0.3 in large bags of compost, and 0.5 in bags of soil. A clear effect of container design on final plant size is difficult to discern. The fast relative growth of Cordia in book containers may be due to the air pruning effect. Hyeronima in book containers attained the same final height as plants in small plastic bags. Air pruning may not be as important as for branchy root species like Hyeronima as for the strongly tap-rooted Cordia or Calophyllum. For both Hyeronima and Cordia, paper pots did not promote good plant development. Despite the inner wax lining, the inner walls had deteriorated after one month. Roots of adjacent seedlings grew together and had to be cut apart for measuring and planting. Disintegration may have been accel- erated by the compost substrate. When growing Gmelina arborea seedlings in paper pots, the forestry company Ston Forestal in Costa Rica used a soil substrate and kept the plants in the nursery for only 30 days. Field management did not interact with nursery treatments. Field and nursery treatments were independent and their effects were additive. Different management was 93 not necessary for small plants. Cordia and Hyeronima responded positively to increasing intensity of field management. For Calophyllum, fertilizer and herbicide effects only lasted six months, but after 12 months, plants had grown better in manually weeded plots. Planting stock characteristics as well as weeding regimes and fertilization influenced plant size and plant growth of the three species during the 12-month period. The unique responses may reflect ecological differences among species. Cordia is a pioneer species with fast initial growth that requires sites with high nutrient availability and good drainage (Bergmann et al. 1994). Cordia is used as a shade overstory in intensively managed cacao and coffee plantations in Costa Rica where it probably benefits from fertilizer application and is reported to be damaged by herbicide (Beer 1979). Low specific leaf area allows rapid canopy growth, but low investment in roots make it sensitive to below-ground competition (Haggar and Ewel 1995, 1997). Smaller seedlings with lower leaf area:root length ratios (i.e. more roots) may have been able to obtain the same size as larger seedlings with higher leaf area:root length ratios (i.e. more investment in leaves) due to being less susceptible to competition. Hyeronima is a late secondary, canopy emergent species (Clark and Clark 1992). Further, this species is characterized by its low specific leaf area (fresh leaf area/oven dry mass of leaves) and high fine root density (Haggar and Ewel 1995). Canopy expansion is slower than for Cordia, but it is not competition sensitive. Large seedlings with a high leaf area when planted can capture more light without having to make additional leaves, and thus compared to small plants, they may have an advantage in early establishment. As fine root densities were inherently higher than for Cordia,theremay have been little disadvantage to the higher above- to below-ground biomass allocation found in large plants. This may explain why above-ground biomass allocation is more important for growth and final size than below-ground biomass in contrast to Cordia. Calophyllum has a large geographical distri- bution in the neotropics and occurs in primary forests. Although it generally has slow initial growth in plantations (Butterfield and Espinoza 1995), it has good form and is valued for its wood properties which are similar to those of Cedrela odorata (ACEN 1992). Calophyllum is a slow growing, shade tolerant, mid-canopy species commonly found on highly weathered soils. It is used for restoration of degraded sites in Puerto Rico (Little and Wadsworth 1964; Jordon and Farnworth 1982) and may be competitively viable on poor soils. The lack of seedling response to NPK fertilizer in the nursery as well as in the field may indicate that the species is well-adapted to nutrient-poor soils. Micronutrients added to soil improved growth in the nursery and subsequent growth through increasing leaf area relative to root length. It appears that 94 for this slow growing, low-nutrient tolerant species, growth rates were correlated to above ground biomass allocation and leaf area similar to Hyeronima. Calophyllum, in contrast to Cordia or Hyeronima, did not respond positively to the addition of compost, indicating organic substrates should not be used alone. The influence of container type may have diminished because of the poor response to the compost substrate. Although it is adapted to poor site conditions, a stronger response than what was seen to nutrient availability in the nursery was expected because many plants use resources when available.

Conclusions

Substrate quality was the dominant factor for plant growth for all species. Although responses were species specific, they can be interpreted with respect to the ecology of the species. Moreover, some general conclusions can be made. When plant growth was good in compost, book containers increased root mass and length. Plants grown in book containers had smaller leaf area:root length ratios than plants grown in bags of compost or soil with fertilizer; in some species this was an advantage (e.g., Cordia) in others not (e.g., Hyeronima). Book containers may control root spiraling problems for tap-rooted species (Cordia and Calophyllum). With a high quality substrate, small volume containers can be used. The lack of interactions between nursery treatments and field management indicates that field management does not need to be tailored to nursery stock type. Species of early to mid succession responded positively to better weed control and fertilization.

Acknowledgments

This research was supported by the Organization for Tropical Studies (OTS), Foundation for the Development of the Central Volcanic Mountain Range (FUNDECOR), and North Carolina State University (NCSU), Department of Forestry. Additional support of materials was provided by Ston Forestal, Costa Rica, and Spencer-Lemaire Industries, Edmonton-Alberta, Canada. Advice and encouragement from Norman Jones of The World Bank, Froylan Castañeda of FUNDECOR, and Cavell Brownie of the NCSU Statistics Department were greatly appreciated. Ricardo Murillo and Pedro Rojas kindly allowed us to conduct this research on their farms. 95

References

ACEN. 1994. Memoria [Proceedings]. Segundo encuentro regional sobre especies fore- stales nativas de la zona norte y atlantica de Costa Rica. Estacion biologica La Selva, Organizacion para Estudios Tropicales, Costa Rica. 24–25 Sept. 1992. 87 p. Amorini, E. and Fabbio, G. 1992. The ISSA system for production of container tree seedlings. Tree Planters’ Notes 43(4): 146–149. Beer, J. 1979. The UNU-CATIE “La Suiza” agroforestry case study, pp. 188–192. In: Salas, G. (Ed.) Workshop. Agro-forestry systems in Latin America. Turrialba, Costa Rica, March 26–30. Proceedings. Bergmann, C., Stuhrmann, M. and Zech, W. 1994. Site factors, foliar nutrient levels, and growth of Cordia alliodora plantations in the humid lowlands of Northern Costa Rica. Plant and Soil 166: 193–202. Burdett, A., Coates, H., Eremka and Martin, P. 1986. Toppling in British Columbia’s lodgepole pine plantations: Significance, cause and prevention. Forestry Chronicle (Oct): 433–439. Burdett, A., Simpson, D.G. and Thompson, C.F. 1983. Root development and plantation establishment. Plant and Soil 71: 103–110. Butterfield, R. 1993. Tropical Timber Species Growth in the Atlantic Lowlands of Costa Rica and Wood Variation of Two Native Species. Ph.D. dissertation, North Carolina State Univ. Raleigh, 76 p. Butterfield, R. and Fisher, R. 1994. Untapped potential. Native species for reforestation. J. of Forestry 92(6): 37–40. Butterfield, R. 1995. Promoting biodiversity: Advances in evaluating native species for reforestation. Forest Ecology and Management 75: 111–121. Butterfield, R. and Espinoza, M. 1995. Screening trial of 14 tropical hardwoods with an emphasis on species native to Costa Rica: Fourth year results. New Forests 9: 135–145. Calvo, G. and Melendez, L. 1999. Pseudoestacas de laurel para el enriquecimiento de cacaotales. Agroforesteria en las Américas 6(22): 25–27. Clark, D.A. and Clark, D.B. 1992. Life history diversity of canopy and emergent trees in a neotropical rain forest. Ecological Monographs 62(3): 315–344. Fry, G. and Poole, B. 1980. Evaluation of planting stock quality several years after planting (a discussion). New Zealand J. of Forestry Science 10: 299–300. Haggar, J. and Ewel, J. 1995. Establishment, resource acquisition, and early biomass allocation patterns of three tropical tree species. Forest Science 41: 689–708. Haggar, J. and Ewel, J. 1997. Primary productivity and resource partitioning in model tropical ecosystems. Ecology 78(4): 1211–1221. Jordon, C.F. and Farnworth, E.G. 1982. Natural vs. plantation forests: a case study of land reclamation strategies for the humid tropics. Environmental-Management 6(6): 485–492. Josiah, S. and Jones, N., 1992. Root Trainers in Seedling Production Systems for Tropical Forestry and Agroforestry. Land Resources Series No. 4. The World Bank Asia Technical Department. 40 pp. Kormanik, P. 1986. Lateral root morphology as an expression of sweetgum seedling quality. Forest Science 32(3): 95–604. Ladrach, W., 1992. Plantation establishment techniques in tropical America. Tree Planter’s Notes 43(4): 125–132. Lamprecht, H., 1986. Silviculture in the Tropics. GTZ, Eschborn. 296 pp. Liegel, L. and Venator, C., 1987. A Technical Guide for Forest Nursery Management in the Caribbean and Latin America. USDA, Puerto Rico. 156 pp. 96

Little, E. and Wadsworth, F. 1964. Common Trees of Puerto Rico and Virgin Islands. USDA. Agriculture Handbook No. 249. Morato, Oscar, technical advisor for Sarapiqui Agriculture Center. Personal communication, 1996. Mason, E.T. 1985. Causes of juvenile instability of Pinus radiata in New Zealand. New Zealand J. of Forestry Science 15(3): 263–280. McDade, L.A., Bawa, K.S., Hespenheide, H. and Hartshorn, G. (Eds.). 1994. La Selva. Ecology and Natural History of a Neotropical Rain Forest. Univ. of Chicago Press. 486 pp. Parkinson, J. and Allen, S. 1975. A wet oxidation procedure suitable for the determination of nitrogen and mineral nutrients in biological material. Commun. Soil Sci. Plant Anal. 6(1): 1–11. Rheingans, R. 1996. The Ecological, Economic and Political Effects of Tree Plantation Inter- cropping on Small Farm Forestry in the Atlantic Lowlands of Costa Rica. Ph.D. diss., Yale Univ. 195 pp. Rose, R., Carlson, W. and Morgan, P., 1990. The target seedling concept, pp. 1–8. In: Rose, R., Campbell, S. and Landis, T. (Eds.) Target Seedling Symposium: Proceedings, Combined Meeting of the Western Forest Nursery Associations. Aug. 13–17. Rose, R., Haase, D. and Boyer, D. 1995. Organic Matter Management in Forest Nurs- eries: Theory and Practice. Nursery Technology Cooperative, Oregon State University, Corvallis, OR, 65 p. SAS Institute. 1990. SAS Procedures Guide, Version 6. SAS Institute, Cary, NC. 724 pp. Shelhas, J., Jantzi T., Kleppner, C., O’Connor, K. and Thatcher, T. 1997. Meeting farmers’ need through forest stewardship. J. of Forestry (2): 33–38. Sharma, R.D. 1987. Some observations on coiling of roots in nursery raised plants. J. of Tropical Forestry 3: 207–212. South, D., Mitchell, R., Zutter, B., Balneaves, J., Barber, B., Nelson, D. and Zwolinski, J. 1993. Integration of nursery practices and vegetation management: Economic and biological potential for improving regeneration. Can. J. For. Res. 23: 2038–2092. Tennant, D. 1975. A test of a modiﬁed line intersect method of estimating root length. J. of Ecology 63: 995–1001. van den Driessche, R. 1984a. Soil fertility in forest nurseries, pp. 63–74. In: Duryea, M. and Landis, T. (Eds.) Forest Nursery Manual: Production of Bareroot Seedlings. Martinus Nijhoff/Dr. W. Junk Publishers. van den Driessche, R. 1984b. Relationship between spacing and nitrogen fertilization of seedlings in the nursery, seedling mineral nutrition and outplanting performance. Can. J. For. Research 14: 431–436. Wolﬁnger, R. and Chang, M. 1995. Comparing SAS GLM and MIXED Procedures for Repeated Measures. SAS Institute, Cary, NC.