Does teak have a future in tropical America?

A survey of Tectona grandis in the Caribbean, Central America' Venezuela and Colombia

Raymond M. Keogh

RAYMOND M. KEOGH is a member of the Forest and Wildlife Service of Ireland. He worked on teak plantations in the Caribbean with FAO.

Native to India and the Southeast Asian region (Burma-Thailand-Lao), teak has long been recognized for its excellent wood properties, making it one of the most valuable many-purpose timbers of the world. These properties include among others: strength with lightness; durability; dimensional stability; non-corroding properties; ease of working and seasoning; termite, fungus and weather resistance; attractiveness. It has been employed for a wide range of end uses such as bridge and wharf construction; for furniture, cabinet work, sleepers and general carpentry; it is highly prized as a shipbuilder's timber, especially for decking; it is suitable for carving and lasts well in contact with the ground.

Pioneer work took place in India in the 1840s to create plantations and propagate this desirable species in an artificial and controlled manner. Success promoted the extension of teak in and around its native areas and eventually it spread to other countries of the tropics and subtropics.

In the Americas, at least in Central America and the Caribbean, natural hardwood species have been and are being encroached upon but not replenished. Many of them present regeneration problems, either their silviculture is not fully understood or, as in the case of the Meliaceae, a natural barrier (Hypsipyla grandella) inhibits their development. There are of course exceptions, but depletion of the hardwood stocks is faster than their regeneration and this is sure to create, in the future, increased demands on the remaining species and on established hardwood plantations. Yet teak is apparently an ideal species for planting in suitable parts of tropical and subtropical America, with its good-quality timber, ease of regeneration, hardiness and aggressive silvicultural characteristics, relative ease of management in plantations and, in particular, the fact that there are similarities between sites in the Americas and its native areas.

Sometimes teak is encouraged on the basis of the exceptionally high prices it fetches on the international market. However, it is not without its silvicultural and management difficulties and there are social, economic and other considerations to take into account when proposing this species for a planting programme. The present report is a summary of work carried out exclusively in a region that includes Central America, the Caribbean, Venezuela and Colombia. It is thought that this area can be treated as a unit, at the present at least, for similarities in some growth parameters and in technical, social and economic problems that have been observed in groups of countries. Besides, the number of genetic types or strains introduced is low.

No overall analysis of teak in this area has taken place before and it is hoped that the information and some of the ideas and recommendations presented here may help interested planters to decide whether or not it is worth undertaking a teak-planting programme, and may also encourage those who have planted to bring their plantations into production if they have not done so already.

Plantation area

The first pure teak plantation was formed in Trinidad in 1913 (Brooks, 1941) and was probably the first plantation of the region. Little planting took place in the following years but a definite programme began in 1928 (Lamb, 1957). At the beginning the planting rate was some 80 ha/year and this was increased to 160 ha/year in 1940/41 and again to 240 ha/year by 1953. Bell (1973) reported an annual expansion of approximately 280 ha/ year. In all, it is estimated that the island possesses at least 9 700 hectares under this species, over 70 per cent the planted area of the region. (Table 1). Dardaine (1972) says that about 14 000 ha of forest land are suitable for establishing plantations in the country.

Outside Trinidad, programmes were not carried out on a comparable basis and in at least three countries - Honduras, Panama and Costa Rica - plantations formed by banana companies have been largely cut back. Belize has no planting programme for this species (Forest Service, Belize, 1978). 'The same is true for Nicaragua (Ortega, 1977). In spite of former interest it is not considered that Puerto Rico can offer suitable soils (Marrero, 1950) and for this reason an expansion of plantations in government forests is not expected. Costa Rica has no definite programme, though some teak is planted each year. Over 50 ha/year are planted in El Salvador, mostly on private holdings. Plantation increases in Colombia, Cuba, Venezuela, Panama and Honduras are uncertain, but may be in excess of 200 ha/year.

A global estimate of the area under teak plantations in the region comes to between 13 000 and 14 000 ha for the year 1978 (see Table 1). The author has calculated that this is growing at a rate of at least 300 ha/year and may be exceeding 450 ha/year. Comparison may be made to estimates for various parts of the world for 1965 (see Table 2).

Table 1. Approximate area in hectares planted to teak in Central America, the Caribbean, Venezuela and Colombia

Country	Area	Reference
Belize	30	Forest Service, 1978
Colombia	560	Sanin, 1975
Costa Rica	300	Keogh et al., 1978a
Cuba¹	200	Cozzo, 1976
El Salvador	230	Keogh, 1977a
Nicaragua	60	Ortega, 1977
Puerto Rico	800	Morales, 1977
Trinidad ²	9 700	1978
Venezuela	560	Mendoza, 1976
Panama Honduras³ and others	1 000	1978
Total 4	13 440

¹ Data for Cuba are published with the permission of Editorial Hemisferio Sur S.A., Buenos Aires, Argentina. - ²Trinidad had 8 500 ha (21000 acres) under teak in 1972 (Dardaine, 1972) and it has been estimated that at least 200 ha (500 acres) per year have been planted thereafter. - ³The area under teak in Panama and Honduras is difficult to establish. The United Fruit Company had 960 ha (2 370 acres) in Honduras in 1953 (Chable, 1967) and there were about 650 ha (1600 acres) under teak in Panama in 1969 (Kadambi, 1972). Since then much cutting has been done and the area has fallen. The author has calculated that there may be 1 000 hectares still standing, but this is largely guesswork. Nevertheless, if Panama and Honduras are combined with other countries not mentioned in the list (Jamaica, Haiti, Dominican Republic and several other islands of the Caribbean), the figure 1 000 hectares is probably conservative. - ⁴The total of 13 440 ha may be misleading: it does not represent the situation for one single year. However, it is safe to apply a figure of 13-14 000 ha as the area covered by teak in the region in 1978.

Teak provenances

A report on provenances, also by the author (Keogh, 1977b), includes, separate information for 12 countries of the region: Trinidad, Panama, Honduras, Cuba, Nicaragua, Puerto Rico, Venezuela, Costa Rica, Guatemala, Colombia, Belize and El Salvador. Other countries are treated very briefly. Here are some excerpts. Nineteen introductions of teak seed from the exterior into the region under discussion have been identified; two of these failed to germinate and a third was seemingly destroyed. Apparently, a minimum of two introductions were made in the 19th century, one from Burma and one from India; there: may have been others but this is difficult to determine. Two countries in the region participated in the Inter national Provenance Trials of the: Danish/FAO Forest Tree Seed Centre: Cuba and Venezuela. (The provenances of these two introductions, made for experimental purposes, are not listed in the paper.)

Teak has come from Burma, Thai land, India and Africa (Nigeria, Cam croon, ivory Coast and Gambia). The: most important introductions, apparently, were those made between 1913 and 1916 from Tenasserim, Burma, to Trinidad and in 1926 from Sri Lanka to Summit Gardens in Panama. The Tenasserim-Trinidad strain has been generally regarded as being a. "good" one. Seed of these two strains has been widely distributed in the region. Other strains have been planted mostly on a small scale, with the exception of the strain introduced into Honduras from Trinidad in 1927. This is probably of Burmese origin, although not necessarily of the Tenasserim-Trinidad strain.

Table 2. Teak area distribution in hectares, 1 965

West Africa	18 600
East Africa	8001
South Asia	219 300
East and Southeast Asia	40 800
Australia	900¹
Latin America	7 700¹
Near East	8 200¹

Source: FAO World Symposium on Man-made Forests. Unasylva, Vol. 21, Nos. 86-87, 1967.
¹Figures known to be incomplete.

Quite a lot of in-breeding is likely to have taken place between the relatively few individuals within strains in the region; in some cases small dots of seed have produced few offspring and the offspring have been used in turn as seed sources, and so on. On the other hand, cross-fertilization may be taking place in some areas between different provenances. However, this is speculation and new varieties may or may not have desirable qualities. In conclusion, it is difficult to recommend areas in this region, other than Trinidad, as seed sources. For countries willing to experiment in provenance trials, it is recommended that they become involved with an international group such as the Danish/FAO Forest Tree Seed Centre, rather than begin haphazard programmes on their own.

Growth and yield

Two other papers by the author deal with growth and yield in the region: one concentrates on the construction of a provisional site classification chart (Keogh, 1979a), the other discusses total volume production (Keogh, 1979b). Each is treated separately below.

A TEAK PLANTATION IN EL SALVADOR, high potentials for Latin America if the sites are good

Provisional site classification chart

The chart has been based on the best data available to the author: some 144 top, dominant and dominant-co-dominant observations, gathered from ten sources, or 13 countries throughout the region. No attempt was made to weigh the observations in any way. The main disadvantage with the data is that different definitions of dominance have been applied:

1. Top height (the mean height of the 100 largest diameter trees per hectare)
2. Dominant height (mean height of 100 tallest trees per hectare or average of dominant trees)
3. The height of the dominants and co-dominants.

However, it has been pointed out that teak, an intolerant species, does not develop readily the range of dominants, co-dominants, intermediates and suppressed trees; stands are normally composed of a large number of trees of more or less the same size. Thus the difference between the height definitions of the data employed may not be too great.

The 144 observations were divided according to height types into five groups and for each group the relationship between height and age was worked out by the least square method, utilizing the following formula:

Log h = a + b (x)

where h = height (dominant, top or height of the dominants and co-dominants)
a and b are regression coefficients
x= 1/age in years.

Thereafter, all 144 observations were brought together and the general relationship between height and age was calculated by the least square method, using equation 1. The slopes (b) of all equations were then compared, applying the method employed by Freese, 1967, for group regressions.

It was found that if data from Jamaica were excluded, it was possible to produce a common equation (D) from the remaining data, the slope of which was not significantly different from separate group equations. Besides, slopes of the separate equations were not significantly different from each other. The latter (equation D) was substituted for group equations. At this stage it is not possible to determine if top height data from Jamaica originate from a different population, but it is assumed that the difference produced is due to a chance distribution of points. All points were plotted on graph paper: height being represented on the "Y" axis (abscissa), and a line representing the common equation was drawn through the data. The year 25 was chosen as the "key year." The original points were segregated into height or site classes with the aid of the following formula:

where h = height (dominant, top or the mean of dominants and co-dominants)
log S.I. = logarithm of site index
x = age in years
xk = key year (25 years in this case)
b = regression coefficient calculated in equation (1)

Site classification chart for Tectona grandis (THIS GRAPH MAY BE USED AS A PRACTICAL GUIDE TO THE CLASSIFICATION AND PREDICTION OF TEAK GROWTH THROUGHOUT THE REGION UNDER DISCUSSION)

Eleven site indices were chosen above the key year in such a way that the, resulting equations would incorporate the majority of plotted points and at the same time cross the key year equidistant one from the other. Curves representing the 11 equations were, drawn on the graph between years 4 and 30. This resulted in 5 height or site classes. In the Figure the resulting site classes are illustrated bud: original data have not been incorporated.

The Figure is in effect a site classification chart applicable to the three definitions of dominance on which it: was based. However, it is strongly recommended that the concept "top height" (mean height of the 100 largest diameter trees per hectare) be applied. It is reliable and relatively easy to measure. Besides, if only one height classification is used, it will be feasible to combine data in the future and expect more reliable results. In the meantime, the site classification chart, although only tentatively valid, should be practically applicable throughout the region to compare, classify and predict growth in plantations. Among other uses it may also act as a guide in the identification of factors responsible for increment variation.

Total volume production

Basic data are drawn from Trinidad, El Salvador and Jamaica only. Therefore it remains to be seen whether or not the results obtained are applicable to the region as a wholes At any rate, it is suggested that the region be treated as a unit at the moment; if variations in volume growth patterns do show up later, they should be readily detected against the "standard" illustrated here. Judging by the data presented, the indications are that strong similarities do in fact exist, at least in the countries studied. One reason for the apparent similarities may be due to the fact that only two, or at most three, major provenances exist in the region.

Three volume-height formulae from different countries of the region and the regional site classification chart (see Figure) were used to estimate total volume production per unit area through time. The formulae are as follows:

(A) That calculated by Keogh (1977a) for El Salvador.

V = 3.394 (h) - 0.344 (h²) 62.78
where V = stem volume in cubic metres per hectare, under bark, between 0.3 m above ground level and to a minimum diameter of 8.0 cm below bark;
h = top height, in metres, defined as the mean height of the 100 largest diameter trees per hectare.

(B) That based on work by Fries (1972) for Jamaica.

V = 0.3889 (h²) - 0.8989 (h) +0.5031
where V --volume in cubic metres per hectare to 10 cm top diameter;
h = top height, in metres, corresponding to the 100 largest trees per ha (in diameter).

(C) Equation based on data from sources for Trinidad.

V = 72.2 - /4.22 (h) +0.86 (h²) where V = volume in cubic metres per hectare defined as "under bark" or "solid volume of stem wood;"
h = average or mean height of dominant trees in metres.

Heights taken from the mean curves of Classes I, II, III, IV and V for years 5, 10, 15, 20, 25 and 30 were employed in equations (A), (B) and (C) and total volume production in m³/ha was worked out in each case. The resulting estimate of total volume production was divided by the respective year to obtain mean annual increment. Results are shown in Table 3. It must be remembered that volume definitions are different for each equation and in the case of (A) and (B) height is defined as "top height," while in equation (C) height is "dominant."

Greater overall differences in mean annual increment (MAN) are found in site Class I and in the younger ages of all classes. The greatest overall difference was found in site Class I at 5 years, where equation (C) produced a MAI of 7 m³/ha/year, some 5 m³/ha/ year lower than that resulting from equation (B). Least differences were found in Classes II, III and IV for years 15, 25 and 30 respectively.

A general examination of Table 3 indicates that some similarities exist between countries in growth rates of MAT within classes. Based on the latter and taking into consideration precautions outlined in the study (Keogh, 1979b), a further comparison was made, but restricted to certain site index curves of Classes II and III and expanded to include a comparison of Miller's results from Trinidad (Miller, 1969). The site index curves were chosen from the regional site classification chart on the basis of their similarity in MAI production to the MAI production of height Classes I, II and III in Trinidad. It should be pointed out that 70 percent of the original 144 points lie between the outer curves of regional Classes II and III. Comparisons appear in Table 4. Miller's definition of volume is "under bark to a top diameter of 8 cm." Mean annual increment production in El Salvador and Jamaica shows similar trends to Trinidad, even though their plantations have not been managed under any formal silvicultural treatment.

Table 3. Mean annual increment (MAI) compared through time for different regional site classes and for three equations

Regional site class	I			II			III			IV			V
Equation	(A)	(B)	(C)	(A)	(B)	(C)	(A)	(B)	(C)	(A)	(B)	(C)	(A)	(B)	(C)
Years	MAI in m³/ha
5	9	12	7	4	8	4	-	5	3	-	3	3	-	2	-
10	17	16	16	12	12	10	7	8	6	3	5	3	-	3	2
15	16	17	11	11	11	7	7	6	4	5	3	1	3	1
20	14	13	16	10	10	11	7	7	6	4	4	3	1	2	1
25	13	12	15	9	9	10	6	6	6	4	4	3	1	2	1
30	11	10	13	8	8	9	6	5	5	3	3	3	1	2	1

Equations: (A) Keogh, 1977a - El Salvador (B) Fries, 1972 - Jamaica; (C) Group equation Trinidad.

Table 4. Comparison of mean annual increment (MAI) estimates, based on selected regional site index curves and two volume-height equations, to mean annual increment estimates based on three height classes for Trinidad

Class/site index curve	T¹ (I)	RSI ² (I/II)		T (III)	RSI (II/III)		T (III)	RSI (III)
Equation	-	(A)³	(B)4	-	(A)	(B)	-	(A)	(B)
Years	MAI in m³/ha
5	11	6	10	7	2	7	5	-	5
10	13	14	14	10	9	10	7	7	8
15	14	14	13	10	9	9	7	7	7
20	12	12	11	9	8	8	7	7	7
25	11	11	10	8	8	7	6	6	6
30	11	10	8	7	6	6	6	5

(¹T: Trinidad height class (Miller, 1969). -²RSI: Regional site index curve. - ³Equation (A): Keogh, 1977a - El Salvador. - ⁴ (B): Fries, 1972 - Jamaica.

Thinning practice

Thinning practice has also been discussed for the region by the author (Keogh, 1979b).

If teak responds in growth in a similar way to sites within the region, it may behave in a similar fashion to common thinning practice. There is no apparent reason, at the moment, why the silvicultural practices proposed by Miller (1969) for Trinidad should not be applied throughout the region as standard practice. Miller's suggestions are technically correct and simple, and are in line with thinning practice in India and Nigeria. Details are given in the following section. It is possible, too, that the yield tables compiled for Trinidad by Miller could have a wider application throughout the region as a guide to management (Miller, 1969).

If a common thinning practice :is not the optimum for teak in the region, very little harm can be done by its application. Consider the present situation: the literature shows that some unrelated work is taking place. Several institutions start out to find the optimum thinning regulations for the species in their own countries on the basis of very few plots, some of which are extremely small. The consequence is very slow headway in research, and management is offered little or no guidelines, resulting in the abandoned or unmanaged plantation. It is the author's view that simple, easy to apply thinning regimes should now be offered to forest managers in the region; there would appear to be no technical objections to this, and besides, teak will normally be planted for sawn timber, which eliminates urgent exploration of different regimes with different end products in mind. Further, common thinning practice would offer a standard against which other thinning regimes could be com.. pared and would show up parts of the region that behave differently fin growth or respond differently to thinnings. It might be possible, in the future, to combine data from the whole region or large sections of it and compile a regional yield table. The above should speed up advances in silvicultural research for the species in the region, decrease research costs to all countries concerned and bring many plantations into immediate production.

The thinning regime proposed is based on Miller's suggestions for Trinidad with some modifications (Miller, 1969). An initial spacing of 2 × 2.5 m is considered, giving 2 000 trees per ha. Heavy thinnings are recommended at first, during the years in which teak grows vigorously. The first two thinnings should be made in a semi-mechanical form: each one attempts to reduce the number of trees by 50 percent; thereafter thinning is based on basal area.

First thinning: when the stand reaches 8 m mean height, the first thinning is made; every second tree on each line is removed. It is of course possible to apply certain selection criteria to favour well-formed trees or remove very badly formed ones. Nevertheless, it is inevitable that in a semi-mechanical thinning some "good" trees will be felled and some badly formed or small ones will be left behind. The primary thinning reduces the density from 2 000 trees/ha to 1000 per ha. On good sites, stands may reach 8 metres in the third growing season; on poor sites, regional Class III, this may not happen before the sixth year.

Second thinning: when the stand reaches roughly 15 m mean height, the second thinning is made. This corresponds to height Class II Trinidad, apparently equivalent to site index curve II/III for the regional chart. For better classes (Class I Trinidad/regional site index I/II), the second thinning should be made when the stand reaches 16 m mean height, for lower classes (Class III Trinidad/' regional site index III) the second thinning is made when the stand reaches 13 m mean height. Better stands will reach the second thinning; stage at about 7 years, poorer ones at about 12 years. As in the first thinning, 50 percent of the standing trees are removed; this results in a reduction from 1000 trees per ha to 500 per ha. Every second tree on each line is removed. Again it is advisable to apply certain selection criteria to favour very well-formed trees and to remove badly formed ones.

Subsequent thinnings: for subsequent thinnings, the standing basal area should be allowed to build up to 20 or 21 m²/ha, when a thinning which removes 6 m²/ha should be made. Standing basal area may be estimated using the simple relascope technique.

Rotation length will be based on economic considerations, a subject not dealt with in the present paper.

The above recommendations are given on the assumption that teak will continue to be employed as a potentially useful species for the region, but how many countries will continue using it in their reforestation programmes or what proportion of the region will be or should be planted is open to debate.

Has teak a potential in the region?

Is it beneficial for countries in the region to undertake or continue planting programmes with this species? As noted previously, one or two are not planting or have discounted teak as a promising species. It is by no means simple to give a direct answer to this question, but some points should be considered.

To achieve the best returns it has been suggested that a site index of not less than 24 m top height at 50 years age is required (Kemp, personal communication, 1977). On the regional site classification chart a site index of 21.8 m should grow through a top height of 24 m at 50 years. It. would seem, therefore, that all sites below this should not be planted. The critical site index (21.8 m) cuts the key year between site-index curves II/III. This indicates that most: of site Class III would be rejected.

Miller (1969), considers that it is very difficult to justify economically the establishment of teak on sites of Class III Trinidad. Class III Trinidad compares closely to the mean values of the regional Class III, at least in regard to volume production (Table 4).

This suggests that lower sites (part of Class III and all of IV and V) should not be planted. To avoid these, the following should be aimed for: sites that contain deep (1-2 m), well-drained, flat or slightly sloping alluvial loam soils which tend to have a homogeneous profile. Good soil is the number one requisite for teak. Secondly, the area should be influenced by a 3- to 5- or 6-month dry season ("dry month" being defined as that in which 50 mm or less of precipitation are accumulated). Thirdly, the site should have an annual precipitation of over 1 500 mm. As precipitation increases, somewhat higher growth rates are to be expected.

These sites are few and far between; even somewhat poorer sites are relatively scarce, maybe not in relation to area, but competition from agriculture for them is too great, especially in areas of high population densities, and several countries in the region suffer in this respect.

Contrary to many opinions, teak is not a fast-growing species, although initial growth can be rather rapid. Of course, when considering the quality of wood produced, this may be acceptable. On the other hand, there is a long wait to obtain the larger dimension trees which are sought after when planting this species; a rotation of 70 years or more has been suggested for Trinidad. Long-term investments of this type will normally be considered only by governments. At the same time, it must be remembered that some returns will be coming in from the thinnings. Alternatively, shorter rotations have also been considered, but smaller tree dimensions and thus lower prices per cubic metre must be expected. Quick return, fast-growing conifers might be better, and these accept poorer sites. On the other hand, it is difficult to predict what needs will be 70 or 80 years from now. Based on this doubt, it is advisable for all countries to have certain hard wood reserves and not, so to speak, put all their eggs in one basket.

If it is decided to plant this species, it should be done with a view to supplying the local market and thus local prices will have to be accepted. The international market (Europe, United States) should not be a primary aim, where, although prices are good, competition is stiff, standards are high, larger dimensions are sought and a constant supply is required. Of course the local market may include the within-region market.

To sum up, individual countries must examine their own situations very thoroughly before commencing or continuing planting programmes with teak. Those with lower population densities and low land pressures have the greatest opportunities and may be able to establish small teak forests. Forests lend themselves to economies of scale. Those with higher population densities and greater land pressures will find it more difficult. Results will be patchy and teak will be forced onto the poorer sites. However, if the better site cannot be given to this species, it is advisable to consider alternatives.

References

BELL, T.I.W. 1973. Erosión en las plantaciones de teca en Trinidad. Instituto Forestal Latinoamericano de Investigación y Capacitación. Mérida, Venezuela. Boletín No. 44-45, p. 314.

BROOKS, R.L. 1941. Notes on pure teak plantations in Trinidad. Caribbean Forester, Vol. 3(1): 25-28.

CHABLE, A.C. 1967. Reforestation in the Republic of Honduras, Central America. Reprint in Ceiba (Honduras), Vol. 13(2): 1-56.

DARDAINE, S. 1972. Annual report of the Forestry Division. Trinidad and Tobago, 92 p.

FOREST SERVICE, BELIZE. 1978. Reply to teak questionaire.

FREESE, F. 1967. Elementary statistical methods for foresters. Agricultural Handbook 317. US Dept of Agric. Forest Service, Washington, D.C., 87 p.

FRIES, J. 1972. The profitability of forest plantations in Jamaica. FAO Forestry Development and Watershed Management in the Upland Regions, Jamaica. FO:SF/JAM 5, Technical Report 6, 69 p.

KADAMBI, K. 1972. Silviculture and management of teak. School of Forestry, Stephen F. Austin State University, Nacogdoches, Texas. Bulletin 24.

KEMP, R.H. 1977. Personal communication.

KEOGH, R.M. 1977a. Elaboración de una tabla de volumen y un estudio de incremento para teca (Tectona grandis) en El Salvador. Proyecto Forestal FAO/ELS/72/004. Documento de Trabajo No. 14. Organización de las Naciones Unidas pare la Agricultura y la Alimentación, Roma.

KEOGH, R.M. with FALLAS, J. & MORA, F. 1978a. Teca (Tectona grandis) en Costa Rica. Documento de Trabajo No. 16, PNUD/FAO/COS/7. Universidad Nacional, Escuela de Ciencias Ambientales: Organización de las Naciones Unidas pare la Agricultura y la Alimentación. Dirección General Forestal. Ministerio de Agricultura y Ganadería. San Jose, Costa Rica.

KEOGH, R.M. 1978b. Teak (Tectona grandis) provenances of the Caribbean; Central America; Venezuela and Colombia. Unpublished. Original document at FAO headquarters, Rome. 16 p.

KEOGH, R. M. 1979a. Teak (Tectona grandis) provisional site classification chart for the Caribbean; Central America; Venezuela and Colombia. Unpublished. Original document at FAO headquarters, Rome. 15 p.

KEOGH, R.M. 1979b. Teak (Tectona grandis) volume growth and thinning practice in the Caribbean; Central America; Venezuela and Colombia. Unpublished. Original document at FAO headquarters, Rome.

LAMB, A.F.A. 1957. Teak (Trinidad). Tropical Silviculture, Vol. II. FAO, Forestry and Forest Products Studies, No. 13, Rome. p. 179-186.

MARRERO, J. 1950. Results of planting in the insular forests of Puerto Rico. Caribbean Forester, Vol. 11 (3): 107147.

MILLER, A.D. 1969. Provisional Yield Tables for teak in Trinidad, 21 p. Government printer, Trinidad.

ORTEGA B., V.M. 1977. Reply to teak questionnaire. El Servicio Forestal de Nicaragua.