Height, diameter and crown dimensions of Cordia alliodora associated with tree density1

Forest Ecology and Management 127 (2000) 31±40

Height, diameter and crown dimensions of Cordia alliodora associated with tree density1 S. Hummel* Department of Forest Resources, Oregon State University, Corvallis, OR 97331, USA Received 12 June 1998; received in revised form 26 January 1999; accepted 3 March 1999

Abstract To investigate the responses of Cordia alliodora associated with tree density, three permanent Nelder plots and 31 temporary 0.09 ha plots in northern Costa Rica, representing a range of densities (100±195 883 per hectare), ages (1±45 years), and elevation (30±430 m), were measured between 1993 and 1996. Tree height, age, stem diameter at breast height, and crown diameter were analyzed with regression techniques. The height of C. alliodora was signi®cantly associated with tree age and stem diameter, but not with density. Stem diameter decreased with increasing tree density. Results were consistent with competition-density patterns observed in temperate forest trees. The ratio between the crown diameter and stem diameter in C. alliodora was not signi®cantly related to density, although this ratio varied directly with tree age. These results suggest that the merchantable yield of C. alliodora in the low-elevation moist tropics of Atlantic Costa Rica may be increased via standdensity management. Study results also indicate that silvicultural research techniques developed in temperate forests can be used to study tropical species. Tropical trees that regenerate in even-aged stands and that permit an estimate of age are the best candidates for such techniques. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Nelder density plots; Competition density relations; La selva biological station; Costa Rica

1. Introduction Cordia alliodora, (Ruiz and Pav.) Oken, is a deciduous, secondary forest tree with a native range from latitude 258N (Mexico) to 258S (Argentina) in the American neotropics (Liegel and Stead, 1990). Observations by British colonial foresters stimulated experimentation with C. alliodora both within the neotropics *Present address: USDA Forest Service PNW Research Station P.O. Box, Portland OR 97208 USA Tel. (503)808-2084; fax (503)808-2020. E-mail address: shummel/[email protected] (S. Hummel). 1 Paper 3272. Forest Research Laboratory, Oregon State University, Corvallis, OR, USA.

and elsewhere in Africa and the South Paci®c (Marshall, 1939; Hudson, 1984; Neil, 1988; Salah, 1989). By the 1970s, C. alliodora was touted as a valuable, fast-growingtreeforagroforestryandreforestationprojects(Johnson and Morales, 1972; FAO, 1977). A 1988 annotated bibliography of C. alliodora (Greaves and McCarter, 1988) illustrates the prevailing emphasis on observations or studies of reproductive biology, geographic distribution, nursery practices, wood properties, and performance in agroforestry systems. In comparison, little quantitative information is available about tree growth associated with intraspeci®c competition, stand dynamics, or causes of mortality, and to date no experimental results have been published for tree longevity.

0378-1127/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 2 7 ( 9 9 ) 0 0 1 2 0 - 6

32

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

The wood properties and appearance of C. alliodora are comparable to those of mahogany Swietenia mahogani (Liegel and Stead, 1990), teak (Tectonis grandis) (Jane, 1952), or walnut (Juglans regia) (Johnson and Morales, 1972). Seasonal rings are discernible and provide an estimate of tree age (Tschinkel, 1966; Devall et al., 1995). The tree is prized for timber and has a long history of use in Costa Rica as overstory shade for cash crops, such as cacao (Teobroma cacao) and coffee (Coffea arabica) (Record and Mell, 1924; Somarriba and Beer, 1987; CATIE, 1994). Costa Rican farmers receive a premium for C. alliodora; between 1988 and 1992 the stumpage price increased 60.9% (Howard, 1995). For this reason, the tree was a popular species for plantations and reforestation projects in the late 1980s (Butter®eld, 1994). Initial spacing in the plantations was based on local experience with different species, rather than on the autecology of C. alliodora. Recently, C. alliodora has lost popularity for reforestation projects in northern Costa Rica, in part because the performance of monospeci®c plantations could not be evaluated (Hummel, pers. obs. 1995). Autecological data on C. alliodora are needed for the design of silvicultural systems and for the management of plantations and reforestation programs. This study, therefore, investigated growth patterns in C. alliodora under different levels of inter- and intra-speci®c competition I designed this study of C. alliodora to quantify the relationships between stand density, or trees per hectare (tph), and tree growth characteristics of C. alliodora. The relationships investigated were (1) trees per hectare (density) and total height/height growth, (2) density and diameter/diameter growth, (3) age and height, (4) age and diameter, (5) height and diameter, (6) crown diameter (K) and stem diameter (d) (K/d ratio), and (7) K/d ratio and density. These variables are all important elements of stand structure, and the relationships among them in¯uence stand development patterns and the response of C. alliodora to silvicultural practices. 2. Site description The study sites fall into two categories: (1) Nelder density plots or (2) forest stands. All of the study sites were located in the Atlantic lowlands of northern

Costa Rica (latitude 108150 N±118000 N and longitude 848000 E±858000 E). The sites ranged from 30 to 430 m in elevation and had soil pH that varied from 4.3 to 6.98. 2.1. Nelder density plots For the ®rst site category, I made use of Nelder density plots planted at La Selva Biological Station in 1991 as part of the long-term Huertos experiment (Haggar and Ewel, 1994). La Selva (108260 N, 848000 W) lies within the tropical moist forest life zone (Holdridge et al., 1971) and receives an average of 4000 mm of rainfall per year (Butter®eld, 1994; Sollins et al., 1994). The Nelder density plots are on an alluvial ¯oodplain where the soils have been classi®ed as mixed isohypothermic, possibly andic, ¯uventic dystropepts (Haggar and Ewel, 1994). An abandoned, 30-year-old cacao plantation with an overstory dominated by C. alliodora previously occupied the site. The trees were harvested and the residual slash burned prior to installation of the Nelder density plots. A description of the Huertos site is provided by Haggar and Ewel (1994, 1995). 2.2. Forest stands In the second site category were secondary forests or plantations dominated by C. alliodora. The stands were either single-species monocultures or polycultures of C. alliodora managed together with pasture, perennial crops, or other tree species (Tables 1 and 2). The stands were located on farms in the cantons of SarapiquõÂ and San Miguel, or within the annexes of La Selva. The history of agricultural and forestry land-use in the study area has been reviewed by Montagnini (1994). 3. Design 3.1. Nelder density plots From an aerial view, a full Nelder plot resembles a spoked wheel (Fig. 1(A); Nelder, 1962). The design is based on a series of concentric circles (or arcs), with spokes radiating out from the center. The Nelder design originated in agricultural research but has been

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

33

Table 1 General characteristics of the sample polycultural stands in San Carlos and SarapiquõÂ, Costa Rica (1996), in order by age Age (years)

Location

Trees (ha)

Soil pH

Soil texturea

Associated crop

3 4 5 5 5 5 5 8 15 15 17 25 34 35 45

Flamenia Zapote Huertos Huertos Huertos La Selva La Selva Horquetas 4 Rios Venezia Tirimbina Chilamate La Selva/ Las Vegas 2 La Selva/ Chanchera La Selva/ Holdridge

333 856 1988 2897 2898 3664 4008 522 144 165 100 556 200 267 433

5.84 6.01 6.92 6.6 6.58 6.98 6.79 5.1 4.94 4.9 4.95 4.9 6.85 4.9 5.04

CL CL CL CL CL SL SL SC CL L CL SL CL CL L

bananas ±b Euterpe,c Heleconia Euterpe, Heleconia Euterpe, Heleconia HYAL,d CEOD, Euterpe, Heleconia HYAL, CEOD, Euterpe, Heleconia roblee pasture coffeef cacao cacao cacao cacao cacao

a

Soil texture: C is for clay, L for loam, S for sand and combinations thereof (after Thien, 1979). Insufficient data. c Euterpe macrospadix and E. oleracea plus Heleconia imbricata. d Hyeronima alchorneoides, Cedrela odorata, planted with c. e Tabebuia rosea. f Coffea arabica. b

used in forestry to investigate formation of non-random height patterns in Pinus sylvestris (Galinski et al., 1994), diameter increment and mortality in Alnus rubra (Knowe and Hibbs, 1996), inter- and intraspeci®c competition in Pseudotsuga menziesii and A. rubra (Cole and Newton, 1987), competition-density relations in Pinus pinaster (Lemoine, 1980) and P.

silvestris (Spellman and Nagel, 1992), and herbaceous understory development associated with Eucalyptus grandis (Cameron et al., 1989). The La Selva Nelder plot design is characterized by equal angles between spokes, and a geometrically increasing distance between arcs. One tree occupies the intersection of each spoke and arc. C. alliodora

Table 2 General characteristics of the sample monocultural stands in San Carlos and SarapiquõÂ, Costa Rica (1996), in order by age Age (years)

Location

Trees (ha)

Soil pH

Soil texturea

3 4 4 4 5 5 5 8 8 11 13 13 15 35

La Selva/La Flamenia Tres Rosales 1 Tres Rosales 2 La UnioÂn Huertos Huertos Huertos Zapote 1 Zapote 2 Cabinas Laureles San Ramon 1 Isla Grande San Ramon 2 La Selva/Arboleda

246 566 666 1932 1797 2085 2243 143 145 367 122 356 133 587

5.25 6.11 4.30 5.53 6.9 6.6 6.6 6.61 6.29 6.51 4.84 6.01 4.95 6.4

L C SCL L CL CL CL SCL C CL SCL L SC L

a

Soil texture: C is for clay, L for loam, S for sand and combinations thereof (after Thien, 1979).

34

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

three-dimensional strati®ed sampling matrix. Some growth conditions were unavailable (for example, no monocultures over 21 years and 1000 per hectare). From the stands available for sampling in SarapiquõÂ and San Carlos, I selected 31 stands for this study. In these stands, tree densities ranged from 100 to 4011 tree/ha (Tables 1 and 2). In the plantations and secondary forest stands, a temporary 30 m  30 m square plot was located randomly via compass coordinates within each stratum. This plot size was based on pre-study sampling and estimates of variability within different size classes of C. alliodora. This size (0.09 ha) is consistent with guidelines discussed by Curtis (1983) and Van Deusen and Bayle (1991). 4. Measurements and analyses 4.1. Nelder density plots

Fig. 1. Schematic diagram of Nelder density plot. The distance between arcs increases exponentially from the center outward. Thus, if each dot represents a planted tree, the area per tree increases from Arc 1 to Arc 10. The spokes repeat the same planting density pattern over and over. Thus the tree on Spoke 4, Arc 3 has the same area as the tree on Spoke 2, Arc 3. Equivalent planting densities (tph) are given in parentheses next to each arc.

occupies a quarter of each La Selva Nelder plot, or ®ve spokes and eleven arcs. Other sections of the Nelder plot are planted with Cedrela odorata and Hyeronima alchorneoides. The quarter plot planted in C. alliodora was replicated three times. 3.2. Forest stands I looked for stands that represented a range of the conditions in which C. alliodora grows and strati®ed them according to age, density, and site type. All possible combinations of age (1±10 years, 11±20 years, and over 21 years), density categories (1± 300, 301±1000, and over 1000 tph), and site types (monocultures or polycultures) were sought to ®ll a

The response of C. alliodora to competition was investigated over a gradient of planting density (represented by trees per hectare). Growth data from the Nelder density plots (1993±1996) were used to evaluate the relationship between density and height and diameter growth of trees. The Nelder density plot study was concerned with intraspeci®c competition only. Measurements in the Nelder density plots were made annually beginning in 1992 and include height and stem diameter at breast height. To regress the response variables against density, the growing space available to each tree in an arc and the equivalent tree density per hectare represented by each arc were ®rst calculated. Thus, the area of each arc ( ‰R22 ÿ R21 Š, when R ˆ radius) divided by 40 (the number of trees planted in each arc) is the plot area per tree (AT) in m2, and tree density per hectare is 104/AT. The central and the outer arcs, as well as the outer four spokes, served as buffers and were, therefore, excluded from analysis. The 1992±1993 data were also excluded because the measurement protocol was inconsistent. The remaining nine arcs on the central spoke are equivalent to C. alliodora planting densities of 1118, 2086, 4448, 8929, 17716, 36233, 70364, 142857, and 195883 per hectare (Fig. 1(B)). These density values were regressed against height and diameter measurements

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

for 1994±1996 to evaluate the response of C. alliodora associated with intraspeci®c competition. 4.2. Forest stands Data from the temporary plots installed in C. alliodora stands in 1996 were used to evaluate the relationships between density and (1) total height, (2) diameter, and (3) the ratio of crown diameter to stem diameter, and between age and (1) total height, (2) stem diameter, and (3) the ratio of crown diameter to stem diameter. The forest stand study does not include any periodic growth data. Competitive effects may be intra- or inter-speci®c, depending on whether the site is a monoculture or a polyculture. In each plot, two 20 g soil samples were collected for pH and texture analysis. A ®eld test of soil texture was made using the technique described by Thien (1979). Laboratory soil pH analyses followed procedures of Black (1965). Stem diameters for all plot trees were measured at breast height (1.37 m) (n ˆ 1030 in the monocultures and n ˆ 1151 in the polycultures). Height measurements were taken for a sample of trees from each 3 cm-diameter class (n ˆ 180). Both dominant and co-dominant trees were used to evaluate the relationship between height and density. On a subset

35

of trees measured for height, crown dimensions were measured, including height-to-crown-base and crown diameter (n ˆ 121). The lowest whorl with >75% live foliage was selected as the crown base. Crown diameter is the average of four quadrant measurements obtained from compass directions and 90 (clinometer readings, multiplied by 2 (Philip, 1994). The stands included a range of ages, tree densities, and soil characteristics (Tables 1 and 2). All data were analyzed using SAS1 6.11 (SAS Institute, 1988) and Excel software. Standard preliminary statistical methods such as checking residuals, con®rming assumptions of normality, and testing for in¯uential points were used to verify the suitability of model techniques. 5. Results Regression results for the Nelder plot study and the stand study are provided in Tables 3 and 4, respectively. Taken together, these results suggest that the response of C. alliodora to density is consistent with silvicultural principles described for temperate trees. Speci®cally, diameter increment was related to tree density, but height was not. As observed in other tropical species, the ratio between the crown diameter

Table 3 Summary results of simple linear regressions of total height (m), height growth (m), total diameter (cm) and diameter growth (cm) in C. alliodora on density (trees per hectare) in Nelder plots in northern Costa Rica SEa

Response variable Intercept

(Intercept)

Slope

(slope)

p-valueb

R2

n

Total height 1995 1996

8.74 8.80

0.48 0.59

ÿ0.18 ÿ0.36

0.25 0.29

0.49 0.26

0.04 0.14

11 11

Height growth 1993±94 1994±95 1995±96

8.94 8.49 7.58

0.27 0.27 0.32

ÿ0.85 ÿ0.70 ÿ0.2

0.20 0.29 0.29

0.0007 0.038 0.52

0.55 0.39 0.09

17 11 7

Total diameter 1996

11.49

0.42

ÿ1.67

0.2

0.0001

0.9

9

7.99 7.03

0.30 0.35

ÿ0.72 ÿ0.46

0.0 0.24

0.04 0.19

0.39 0.64

11 4

Diameter growth 1994±95 1995±96 a b

Standard error. Significance of density.

36

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

Table 4 Summary results of simple linear regressions of total height (m), height growth (m), total diameter (cm) and diameter growth (cm) in C. alliodora on density (trees per hectare) in Nelder plots in northern Costa Rica Response variable

Independent variable

p-value

R2

n

Height

age DBH density age density CD age density density

0.001 0.0001 0.27 0.0001 0.0006 0.0001 0.01 0.26 0.0018

0.40 0.75 0.02 0.47 0.06 0.74 0.05 0.01 0.08

179 179 71 179 178 121 121 121 121

DBH K/d CD DBH: diameter at breast height.

and the stem diameter (K/d ratio) was not associated with tree density. Height, diameter, and the K/d ratio were related to the age of an individual tree. Plot values for soil texture and pH are listed in Tables 1 and 2. 5.1. Nelder density plots In the Nelder study, total height in young C. alliodora was not signi®cantly associated with density in 1995 and 1996 (p ˆ 0.49 and p ˆ 0.26, respectively). Height growth of C. alliodora was not associated with density in 1995±1996 (p ˆ 0.52). However, for the 1993±1994 and 1994±1995 growth periods, there was a statistical relationship between height increment and density (p ˆ 0.007 and p ˆ 0.038, respectively) that could be biologically signi®cant (Table 3). In addition, density was related to the diameter growth increment in 1994±1995 and total diameter in 1996 (p ˆ 0.04 and p ˆ 0.001, respectively) (Table 3). Finally, the log±log relationship between diameter and trees per hectare for C. alliodora grown in the Nelder plots is linear and negative, as illustrated in Fig. 2.

Fig. 3, (2) height and diameter (p ˆ 0.0001) (Fig. 4), (3) diameter and age (p ˆ 0.0001) (Fig. 5), (4) diameter and trees per hectare, and (5) crown diameter and stem diameter (p ˆ 0.0001) (Fig. 6). The ratio of the crown diameter to stem diameter was related to tree age, as well as to whether the tree was growing in a monoculture or growing with other species in a polyculture (p ˆ 0.0053, R2 ˆ 0.49). Older trees exhibit lower ratios of crown diameter to stem diameter than do younger trees. The equation

5.2. Forest stands In the forest stand study, no signi®cant relationship existed between height and trees per hectare (p ˆ 0.27), or between the ratio of the crown diameter to the stem diameter and density (trees per hectare, or TPH; p ˆ 0.26) (Table 4). Signi®cant relationships were found between (1) height and age (p ˆ 0.001)

Fig. 2. Log±log relationship between stem diameter at breast height (DBH) and density (trees per hectare, TPH) for Cordia alliodora in Nelder plots in northern Costa Rica.

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

Fig. 3. The relationship between total height (in meters) and tree age (in years) for C. alliodora in northern Costa Rica.

Fig. 4. The relationship of total height (in meters) to diameter (in centimeters) for C. alliodora in northern Costa Rica.

37

Fig. 5. The relationship between tree age and stem diameter (DBH) for C. alliodora in northern Costa Rica.

Fig. 6. Relationship between the crown diameter (in meteres) and stem diameter (in centimeters) for C. alliodora in northern Costa Rica.

38

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

to predict the ratio in a polyculture is K/d ˆ 0.53±0.29 (AGE1/2) ÿ0.39 (log TPH). In a monoculture of C. alliodora, the equation is K/d ˆ 0.53±0.29 (AGE1/2) ÿ0.052 (log TPH). The mean crown diameter to stem diameter ratio of C. alliodora was 0.245. 6. Discussion Size-density relationships in the tropical secondary tree C. alliodora were consistent with patterns frequently observed in trees of temperate forests. According to size-density theory, height growth is relatively unaffected by competition (Daniel et al., 1979), but diameter growth decreases with increasing density (Harper, 1977; Drew and Flewelling, 1979). The total height of C. alliodora in this study was not related to density, but diameter was inversely related to tree density. These relationships were observed over a range of tree ages, stand densities, and site conditions during a 3-year period. The systematic spacing design of the Nelder plots permitted analysis of a range of densities in a compact area. Although the stand study design was statistically less rigorous, it provided evidence in support of the size-density relations observed in the Nelder plots under conditions of more interest to local landowners. The relationship of height increment to tree density observed in the earliest measurement periods may be a result of the extremely high densities in the Nelder plots or of the small sample size. No other published data from Nelder density plots are available for the response of C. alliodora associated with intraspeci®c competition. In 1984, C. alliodora was included among 12 promising fuelwood species in Nelder plots established in Costa Rica. Unfortunately, data from these earlier Nelder plots were never analyzed and have since been lost (Butter®eld, 1994). Schlonvoigt (1993) used the Nelder design to investigate interspeci®c competition between the trees C. alliodora and Eucalyptus deglupta and crops of Zea mays and Manihot esculenta. The results of this study may aid in interpreting results from other experiments. For example, when Kapp and Beer (1995) observed that C. alliodora trees grown with crops were taller than trees grown in monocultures, they were uncertain whether the

observed difference in height was due to differences in C. alliodora spacing between the monocultures and the mixed-species stands, or whether some aspect or treatment of the associated crop was involved. If, as the present study indicates, height in C. alliodora is not signi®cantly associated with tree density, spacing can be dismissed as an explanation. To further resolve the questions, future studies of height growth in C. alliodora should include a designed experiment to test the response of the tree to fertilizer, as well as one to investigate patterns in height growth over time. The signi®cant relationships between tree diameter and both height and age identi®ed in this study are consistent with results reported by Somarriba and Beer (1987) for C. alliodora grown in agroforestry systems in the Atlantic lowlands of Costa Rica. The relationship between tree diameter and age is also consistent with results obtained by Perez Figueroa (1954). The ratio between crown diameter and stem diameter in C. alliodora did not vary signi®cantly with density in this study. This result is consistent with Dawkins (1963) study of 18 other tree species in the African, American, and Asian tropics. In Dawkins' study, the ratio was consistent within species, regardless of site, age, or density. Although Curtin (1964) found that other species exhibited more variation in the ratio, he con®rmed a strong correlation between the crown diameter and stem diameter for any given species. In my study, this ratio was associated with tree age. Foresters are interested in the relationship between the crown and stem diameters because a rigid ratio would limit the number of trees of a certain mean diameter in an area for any given species (Philip, 1994). Dawkins (1963) suggested that trees with a low ratio are more suitable for pure even-aged plantations than those with higher ratios, because those with the low ratio could support a greater basal area per unit land area. The mean ratio of C. alliodora in this study, 0.245, is relatively low, which suggests that it may be suited for monocultural stands. In the absence of volume tables for C. alliodora the size of those grown in the Nelder density plots, patterns of volume increment could not be analyzed directly. However, volume is a function of diameter and height, and these parameters do appear to vary with density according to well-established principles of tree growth.

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

The results of this study are applicable for C. alliodora growing in the low-elevation moist tropics of Atlantic Costa Rica. Since silvicultural prescriptions are based on species-speci®c knowledge about the physiological responses of trees to competition, results from this study may be useful for the future establishment and management of C. alliodora in the region. Study results also indicate that silvicultural research techniques developed in temperate forests are suitable for studying tropical species. Tropical trees that regenerate in even-aged stands and that permit an estimate of age are the best candidates for such techniques. Acknowledgements Financial support for ®eld research and data analyses came from the OrganizacioÂn para Estudiantes Tropicales and from Oregon State University. The Nelder density plot material is based on work supported by the National Science Foundation under Grant #DEB9318403 to J.J. Ewel. The Centro Agricola Cantonal de SarapiquõÂ in Puerto Viejo provided information on the Programa de ReforestacioÂn con Incentivos as well as introductions to local landowners. Manuscript preparation was funded by fellowship 085/97S from the International Tropical Timber Organization. References Black, C.A. (Ed.), 1965. Methods of Soil Analysis, Part 2. Chemical and Microbiological Properties. Agronomy 9. Am. Soc. Agron., Madison, WI. Butterfield, R., 1994. Forestry in Costa Rica: status, research priorities, and the role of La Selva Biological Station. In: McDade, L., Bawa, K., Hespenheide H., Hartshorn, G. (Eds.), La Selva: Ecology and Natural History of a Neotropical Rain Forest. University of Chicago Press, Chicago, pp. 317±328. Cameron, D.M., Rance, S.J., Jones, R.M., Charles-Edwards, D.A., 1989. Project STAG: an experimental study in agroforestry. Aust. J. Agric. Res. 40, 699±714. CATIE, 1994. Laurel Cordia alliodora (Ruiz y Pavon) Oken, especie de arbol de uso multiple en America Central. Serie tecnica no. 239. Centro Agronomico Tropical de Investigacion y Ensenanza, Turrialba, Costa Rica. Cole, E.C., Newton, M., 1987. Fifth-year responses of Douglas-fir to crowding and nonconiferous competition. Can. J. For. Res. 17, 181±186.

39

Curtin, R.A., 1964. Stand density and the relationship of crown width to diameter and height in Eucalyptus obliqua. Aust. For. 28(2), 91±105. Curtis, R.O., 1983. Procedures for establishing and maintaining permanent plots for silvicultural and yield research. Gen. Tech. Rep. PNW-155. USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, Portland, OR, 56 pp. Daniel, T., Helms, J., Baker, F.S., 1979. Principles of Silviculture, 2nd edn. McGraw-Hill, New York, 500 pp. Dawkins, H.C., 1963. Crown diameters: their relation to bole diameter in tropical forest trees. Commonw. For. Rev. 42(4), 318±333. Devall, M.S., Parresol, B.R., Wright, S.J., 1995. Dendroecological analysis of Cordia alliodora, Pseudobombax septenatum and Annona spraguei in central Panama. IAWA J. 16(4), 411±424. Drew, T.J., Flewelling, J.W., 1979. Stand density management: an alternative approach and its application to Douglas-fir plantations. For. Sci. 25, 518±532. FAO, 1977. Report of the fourth session of the FAO panel of experts on forest gene resources. FO: FGR4 rep. Food and Agricultural Organization of the United Nations, Rome. Galinski, W., Witowski, J., Zwieniecki, M., 1994. Non-random height pattern formation in even aged Scots pine (Pinus sylvestris L.) Nelder plots as affected by spacing and site quality. Forestry 67(1), 49±61. Greaves, A., McCarter, P.S., 1988. Cordia alliodora 1922±1987. Annotated Bibliography. CAB International in cooperation witht the Oxford Forestry Institute. No. F40. Information Printing, Eynsham, Oxford, 28 pp. Haggar, J.P., Ewel, J.J., 1994. Experiments on the ecological basis of sustainability: early findings on nitrogen, phosphorus and root systems. Interciencia 19(6), 347±351. Haggar, J.P., Ewel, J.J., 1995. Establishment, resource acquisition, and early productivity as determined by biomass allocation patterns of three tropical tree species. For. Sci. 41, 689±708. Harper, J.L., 1977. Population Biology of Plants. Academic Press, London, 892 pp. Holdridge, L.R., Grenke, W.C., Hatheway, W.H., Liang, T., Tosi, J.A., 1971. Forest Environments in Tropical Life Zones. Pergamon Press, Oxford, 747 pp. Howard, A.F., 1995. Price trends for stumpage and selected agricultural products in Costa Rica. For. Ecol. Manage. 75, 101±110. Hudson, J.M., 1984. A note on Cordia alliodora in Vanuatu Commonw. For. Rev. 63(3), 181±183. Jane, F.W., 1952. The structure of timbers of the world. Timber News 60, 183±185. Johnson, P., Morales, R., 1972. A review of Cordia alliodora (Ruiz and Pav.) Oken. Turrialba 22(2), 210±220. Kapp, G.B., Beer, J., 1995. A comparison of agrisilvicultural systems with plantation forestry in the Atlantic lowlands of Costa Rica. Agrofor. Sys. 32, 207±223. Knowe, S.A., Hibbs, D.E., 1996. Stand structure and dynamics of young red alder as affected by planting density. For. Ecol. Manage. 82, 69±85. Lemoine, B., 1980. Stand density, competition and cooperation in maritime pine II. Results at 5 and 10 years of a plantation with

40

S. Hummel / Forest Ecology and Management 127 (2000) 31±40

variable spacing. Annales des Sciences Forestieres 37(3), 217± 237. Liegel, L.H., Stead, J.W., 1990. Cordia alliodora (Ruiz and Pav.) Oken. In: Burns, R.M., Honkala, B.H. (Tech. Coord.), Silvics of North America, vol. 2, Hardwoods. Agriculture Handbook 654. USDA Forest Service, Washington DC, pp. 270±277. Marshall, R.C., 1939. Silviculture of the Trees of Trinidad and Tobago, British West Indies. Oxford University Press, Oxford, 247 pp. Montagnini, F., 1994. Agricultural systems in the La Selva region. In: McDade, L., Bawa, K., Hespenheide, H., Hartshorn, G. (Eds.), La Selva: Ecology and Natural History of a Neotropical Rain Forest. University of Chicago Press, Chicago, pp. 307± 317. Neil, P.E., 1988. Root disease (Phellinus noxius) of Cordia alliodora in Vanuatu. Commonw. For. Rev. 67(4), 363±372. Nelder, J.A., 1962. New kinds of systematic designs for spacing experiments. Biometrics 18, 283±307. Perez Figueroa, C.A., 1954. Estudio forestal del laurel, Cordia alliodora (R and P) Cham, en Costa Rica. Tesis, Mag. Sc., Instituto Interamericano de Ciencias Agricolas, Turrialba, Costa Rica, 209 pp. Philip, M.S., 1994. Measuring Trees and Forests, 2nd edn. CAB International, Wallingford, Oxon, UK, 310 pp. Record, S.J., Mell, C.D., 1924. Timbers of Tropical America. Yale University Press, New Haven, CT, 610 pp.

Salah, G.H., 1989. Provenance trials of Cordia alliodora in the Sudan. Commonw. For. Rev. 68(1), 73±75. SAS Institute, 1988. SAS/STAT Users Guide, release 6.03 edn. SAS Institute, Cary, NC. Schlonvoigt, A., 1993. Untersuchungen zur Konkurrenz zwischen Baumen und annuellen Feldfruchten im humiden tropischen Tiefland Costa Rica. Goettinger-Beitrage-zur-Land-und-Forstwirtschaft-in-den-Tropen-und-Subtropen No. 83. Sollins, P., Sancho, F.M., Mata, R.C., Sanford, R.L., Jr., 1994. Soils and soils process research. In: McDade, L., Bawa, K., Hespenheide, H., Hartshorn, G. (Eds.), La Selva: Ecology and Natural History of a Neotropical Rain Forest. University of Chicago Press, Chicago, pp. 34±53. Somarriba, E.J., Beer, J.W., 1987. Dimensions, volume and growth of Cordia alliodora in agroforestry systems. For. Ecol. Manage. 18, 113±126. Spellman, H., Nagel, J., 1992. Auswertung des Nelder-Pflanzenversuches mit Kiefer im Forstamt Walsrode. Allg. Forst-u.-J.Ztg. 163. Jg. 11/12. Thien, S.J., 1979. A flow diagram for teaching texture-by-feel analysis. J. Agron. Educ. 8, 54±56. Tschinkel, H.M., 1966. Annual growth rings in Cordia alliodora. Turrialba 16(1), 73±80. Van Deusen, P.C., Bayle, B., 1991. Evaluating plot designs for the tropics. Gen. Tech. Rep. SO-87. USDA Forest Service, Southern Forest Experiment Station, New Orleans, LA, 15 pp.