GMELINA ARBOREA - INTERNATIONAL PROVENANCE TRIALS STUDY TOUR AND SEED COLLECTION IN INDIA, 1976 ¹

by

E.B. Lauridsen
Danish/FAO Forest Tree Seed Centre

¹ This condensed version is based on the author's original mimeographed “Report on Gmelina arborea - international provenance trials study tour and seed collection in India, 1976”, October 1976.

INTRODUCTION

As follow-up to a study tour carried out in May–June 1975, a representative of the Central Silviculturist, Forest Research Institute and Colleges, Dehra Dun, and of the Danish/FAO Forest Tree Seed Centre, Denmark, visited several states in India on a joint mission during 1976 to promote and arrange for seed collection of Gmelina arborea Linn.

Seed collection in Gmelina arborea including the preparatory work is a difficult and time consuming task, the major burden of which has to be borne by staff of Indian Federal and State Forest Departments together with staff of research organizations. The Danish/FAO Forest Tree Seed Centre wishes to express its gratitude for the very positive way in which the scheme has been received and for the invaluable support it has had. Grateful acknowledgements are made to the staff of the DANIDA Mission in Delhi for their help in planning the tour and freighting the seeds collected to Denmark. Special thanks are extended to Mr. A.C. Gupta, ecologist, F.R.I. & Colleges, Dehra Dun, with whom I had the pleasure to travel.

BACKGROUND AND SCOPE

Several forestry and research officials we met during the tour felt that they have not been given enough information on the purpose and technical aspects of the provenance scheme and did not, therefore, fully understand the international interest in a species that has generally proved unsuccessful in Indian plantation programmes.

The virtues of Gmelina arborea are fairly well known among Indian foresters, and they are listed by A.F.A. Lamb, 1968 (fast growing timber trees of the lowland tropics, Gmelina arborea, C.F.I. Dept. For. Univ. Oxford) as follows: Easy and comparatively cheap establishment, quick growth, i.e. on good sites for short rotations, yields exceeding 35 m³ per hectare per year (500 cu. ft. per acre) may be expected, consistent wood density whether fast or slow grown, extremely promising pulpwood characteristics for a broadleaved species and, in summary, few species are found that can rival Gmelina arborea in respect of early economic returns.

Irrespective of these assets, it has not been considered feasible in India to begin large scale plantation establishment with this species because of poor initial results ranging from poor growth, bad form, to sheer abandonment of plots due to disease or pests.

In contrast to India, Gmelina has been successfully introduced in several countries outside the species' natural range, especially in Africa and Latin America.

Probably the plantations established within and outside India are based on a limited part of the large distribution range of Gmelina, viz. Burma, Assam and Bengali provenances, where fine specimens could be found and seeds were abundant at least in earlier days. Theoretically the vast distribution range of Gmelina, in India alone from 8° to 27° N and 72° to 96° E, should have given rise to a rich genetical diversity that could serve as a base for an improvement scheme, starting with the present provenance trials. Hopefully seed sources can be found that in combination with adequate silvicultural measures will be suitable for Indian conditions too.

The international interest in Gmelina arborea caused the FAO Panel on Forest Gene Resources to include it as a top priority species in the list of forest trees whose gene resources need exploration, utilization and conservation (see FGRI 4).

From a slow start in 1974, the Gmelina exploration/collection scheme seems now to have materialized with the collection of some 30 seed samples from widely scattered localities in India, Thailand and Africa.

The coordination of the international scheme is delegated to the Danish/FAO Forest Tree Seed Centre, Denmark and the work is done in close cooperation with national forest services and research organizations, in India specifically, the organization of the provenance collections and trials is done by the Central Silviculturist, F.R.I. & Colleges, Dehra Dun.

DISTRIBUTION - COLLECTION ZONES - SELECTION OF STANDS

Provenance collections of Gmelina arborea have not hitherto been carried out and it was consequently decided to launch a 1st stage collection scheme where the entire distribution range in India was to be sampled.

Detailed information on distribution and variation pattern is lacking, or under preparation (see below), but some guidelines for the sampling were available from “Forest Types of India” (Champion and Seth 1968), some other literature and information obtained during the 1975 study tour (Report on study tour in India in connection with international provenance trials of Gmelina arborea, E.B. Lauridsen 1975).

The distribution range was primarily divided into strata according to broad forest types as follows:

Northern semi-evergreen forests,
Northern moist mixed deciduous forests,
Southern moist mixed deciduous forests, and
Dry deciduous forests.

Three varieties or forms are described in the Indian botanical literature and, although the occurrence of these varieties is not definitely established, it was decided to arrange three strata according to the approximate distribution range of these varieties or forms:

G. arborea, Linn. (holotype)
G.a. glaucescens, Hook
G.a. canescens, Haines

The strata and their approximate positions are presented in figure 5.

Within the strata proposals of localities from where we would like to obtain samples were presented for the Central Silviculturist, Forest Research Institute and Colleges, Dehra Dun, who finally prepared a modified scheme according to which we travelled.

Figure 5 shows, in addition to the above types, localities visited 1975–76, localities where seeds have been collected, and localities from where seed samples are wanted in 1977.

Due to time limits, it has not been possible to visit more than one or a few stands within each area proposed for investigation and in consequence there was no chance of choosing among several stands the one best suited for our purpose. We merely arranged for seed collections in any stand containing Gmelina trees enough for satisfactorily large seed samples.

For purpose of supplementing the 1st stage collection and for preparing a second stage scheme, information regarding the occurrence of Gmelina arborea is being obtained from state- and region-wise resource surveys, from working plans, selling lists and from special enumerations done in connection with the provenance scheme.

A special mention is made of the resource surveys, which will generally cover large areas and, in the most systematic way, provide the necessary data. It would be useful if the efforts of the FAO Panel of Experts on Forest Gene Resources to explore potential useful species could be coordinated with the efforts to survey the presently dominating species so the strategy for carrying out the surveys could take into account the irregular and infrequently occurring species, which are often those under consideration by the panel.

Resource surveys have been carried out in Gujarat and Maharashtra on a statewise basis. Useful information on occurrence of Gmelina arborea has been obtained from these surveys, although the information may not always be actual because the surveys were carried out over a period of 10 or more years. A scheme centrally coordinated in Dehra Dun, Pre-investment Survey of Forest Resources, includes surveys carried out in Uttar Pradesh, Madhya Pradesh, West Bengal, Tripura and Kerala. Useful information for our purpose is obtainable from Uttar Pradesh, North Bengal and to some extent from Madhya Pradesh, whereas it is not possible to identify Gmelina arborea in the surveys carried out in Tripura and Kerala.

SEED COLLECTION PROCEDURES

As stated above, it is intended to carry out a 1st stage provenance collection with the purpose of obtaining a knowledge of genetical variation in Gmelina arborea and selecting accordingly the seed sources best suited for plantation establishment in different environments.

In the current collections, the procedures generally recommended and followed in IUFRO provenance collections were adhered to. The requirements for obtaining a sufficient representation of a given population are stated below taking into account the special circumstances for Gmelina arborea, e.g. lack of knowledge of the breeding system.

1. In natural forests a minimum of 25 trees - preferably more - should be collected from, not only to be sure to get enough seeds, but also to be sure to cover most of the genetical variation in a stand under the assumption that a high proportion of trees may be self-pollinating.

2. In natural forests, seed trees should be located at least 100 m apart to avoid sampling several closely related individuals.

3. Seed trees should be average trees, i.e. not specifically selected for superiority.

4. Approximately the same amount of seeds should be collected from each tree to avoid dominance of prolific seeders.

Note 1. Results from this year's collections show that 25 trees may not be enough. The requirements for comprehensive international samples are 30–40 kg of clean dry stones per provenance. With approximate yields of 0.3 kg stones per tree, we shall generally need around 100 trees (see Appendix C).

There are no investigations that can support the assumption of self-pollination, but it is observed that single trees, well isolated from other Gmelina trees, will generally flower and fruit as heavily as trees in pure stands or in groups and, apart from any physiological barriers, the flower-structure lends itself easily to self-pollination, as the pistil is positioned in between or very close to the two sets of anthers and even a slight movement of the flowers can cause transfer of pollen.

Note 2. This provision may not be very helpful because the seeds can be dispersed widely by game and cattle. The procedure will prevent more than one tree in a group of trees at close spacing being sampled; the chances that such trees are relatives are assumed high.

In one case seeds were collected in a stand covering approximately 2 ha and 25 trees, and with no other Gmelina in the vicinity. The variation pattern of progeny of these trees in experimental plots may yield interesting results.

Note 3. Usually the problem of finding enough trees in a locality precludes any choice for specific traits.

TECHNICAL COLLECTION PROCEDURES

Flowering and consequently fruit-ripening periods vary very much from locality to locality, from tree to tree and within trees and inflorescences. Information collected so far does not show any trend in seed ripening period with latitude or forest type.

The fruit falls off the trees soon after ripening and will often be eaten by game and cattle. In order to have a control of seed sources, it is necessary to collect the fruits very soon after they have fallen. This, added to the extended seed-fall period, makes seed collection in Gmelina expensive and difficult to supervise adequately. It is virtually impossible for one collection team to take part in all collections, meaning that local officers have to be requested to carry out this task.

In some cases in the present collection scheme, fruits have been picked directly from seed trees within a day or two. Information is being sought on details of these collections, e.g. proportion of seeds collected of all seeds on trees, state of ripening, time consumption and eventual other problems to find out if this procedure is feasible.

SEED COLLECTION IN 1976

The provenance collections will this year comprise some 25 – 30 provenances covering most of the distribution range of Gmelina arborea. Figure 5 shows localities from where seed samples were obtained this year.

It is unfortunate, however, that not all samples are big enough for the purpose of international provenance trials and repeated collections have to be carried out in 1977 and maybe later.

The goal was to collect around 50 – 70 kg of clean stones per provenance. As seeds come in, it becomes evident that in only a few cases has this amount been obtained. The attached seed list shows the amounts of seeds obtained per provenance, refer to Appendix A (note that the amounts stated represent ⅔ of the total collected).

In one provenance - Ari Range in Seoni Division - 4 trees only out of 119 trees produced fruits in sufficient quantities. The area will be kept under observation for flowering from January 1977 onwards for indications of periodicity in flowering and fruiting.

The seeds collected have been divided between the Forest Research Institute & Colleges, Dehra Dun (The Central Silviculturist) and The Danish/FAO Forest Tree Seed Centre, Denmark, with one third of the seeds to F.R.I. and two thirds to the Seed Centre. The F.R.I. part of the samples, together with additional samples from other countries, are intended for use in the Indian part of the international provenance trials. The Seed Centre's part of the seeds is to be distributed to other countries to take part in the international trials. Seeds of the latter have been airfreighted to Denmark where they are now cool-stored (2–4° C). Samples have been taken out for tests of germination and other possible tests.

Export of forest tree seeds from India is covered by O.G.L. No. 3 of 5 May 1976, and certificates showing that seeds have been legally collected and are in surplus of local requirements have to be issued by the states concerned.

COLLECTION FOR GENECOLOGICAL STUDIES

In several localities, it was planned to keep the seeds collected from individual trees separate for progeny testing. So far in three localities trees have been numbered, seeds have been collected and descriptions as well as maps of location of trees are to be prepared.

In one locality - Seoni South Ari Range c. 169 - seed production was insufficient this year.

Some more localities have to be selected for this purpose.

All seeds for this purpose have been dispatched to Denmark for storage and are later to be distributed to India and other interested countries ready to carry out this kind of investigation.

PROTECTION

Trees and stands outside and inside forests can in India be protected from felling by means of local tree felling acts and stipulation in working plans, etc.

The Central Silviculturist, Dehra Dun, in a circular of May 1976 to the Chief Conservators of Forests concerned, requested that Gmelina trees in areas so far selected for seed-collection are to be protected in order to secure eventual further seed samples.

BULK SUPPLIES OF SEED

Most of the areas selected for sampling can yield only limited quantities of seeds of value for commercial purposes (large scale plantations).

With the initiation of the Indo/Danish Seed Procurement Project comprising cool store facilities, it will be possible to build up stocks of seed from at least some of the provenances through repeated collections. Also, for some of the provenances it may be possible to include adjacent areas of similar ecological character.

The prospects for commercial collections are best where large plantations are established as in North Bengal and Tripura.

PLANS FOR FURTHER OPERATIONS

In connection with the present provenance trials, additional samples will have to be obtained from new localities as well as from localities from where insufficiently large seed samples have already been obtained.

Several areas have to be visted in late 1976 or early 1977. They are localities where seeds were collected in 1976, but not yet inspected, and new localities proposed for seed collection in 1977. A list of localities proposed for visits and seed collections in 1977 is given in Appendix B (see figure 5 also).

As the scheduled provenance collections and trials may have to be repeated over several years, certain provenances will have to serve as standard seed sources. These areas will have to be selected in 1977 among established plantations in either Tripura, N. Bengal or Bihar where satisfactorily large seed samples can be obtained in the most economic way.

Dr. S.K. Kedharnath, leader of The Genetics Branch, F.R.I. & Colleges, Dehra Dun, has kindly offered to start investigation of chromosome numbers in seeds from various provenances.

Figure 1: Gmelina veneer-log 0.64 m3. Supplied in quantities of minimum 5 m3, Gmelina logs fetch the same prices as teak logs of same quality.	Figure 2: Gmelina arborea attacked by Loranthus sp. in 1928 plantation plot, Southern Assam. The plot is given up due to widespread attack, and Malocana bambusoides is grown now as a source of pulp.

Figure 3: It is difficult to envisage success with an improvement scheme involving such trees, but

Figure 4: … its progeny seem to start well. 1972 plantation of Gmelina arborea based on seeds from adjacent local stands. Karnataka, Thithimathi Range, Areckeri State Forest.

FIGURE 5: Gmelina - bearing forest types of India, and tentative distribution of Gmelina arborea, incl. varieties. (Forest types after Champion & Seth, 1968)

Figure 6: Riverain Forest (Northern Moist Deciduous Forest) in Bengal Duars (background) with characteristic Salmalia malabarica in upper storey. Density of Gmelina arborea may here reach up to 900 trees per sq-km.

Figure 7: Dry deciduous forest in Seoni, M.P. Gmelina is among dominant trees and densities up to 40 trees per sq-km are found. (Gmelina arborea is the light tree in centre.)

APPENDIX A

DANISH/FAO Forest

Tree Seed Centre

Gmelina arborea - Seed Stock List

APPENDIX B

APPENDIX C

APPENDIX D

Seed collection dates GMELINA ARBOREA