Propagation of Native Forest Tree Species for Forest Restoration in Doi Suthep-Pui National Park

ABSTRACT: Restoring forest ecosystems by tree planting requires production of high quality planting stock, on a large scale, of a very wide range of indigenous forest tree species. Many of these species have proved difficult to propagate from seed, due to long dormancy or seed production too late for seedlings to grow large enough by planting time. The aim of this study was, therefore, to investigate how to improve propagation of native trees for forest restoration, based on an understanding of their reproductive ecology. The study included development of new techniques to germinate the seeds of 30 indigenous tree species, of potential value to forest restoration, but which had not previously been grown in nurseries. Seed germination in the nursery was compared with that of seeds sown in a forest gap, where the effects of seed predation were also investigated. For 10 species, with limited seed germination, vegetative propagation was investigated, using a novel non-mist system, to propagate leafy stem cuttings, testing various chemical treatments to induce rooting. In addition, the seasonality of production of seeds and material for cuttings was investigated. The aim was to examine whether knowledge of the reproductive ecology of native tree species could be used to predict which horticultural practices are likely to result in successful propagation.

The phenology of 32 forest tree species was recorded monthly over 12 months. Leaf fall occurred in the dry season, in response to declining soil moisture, whilst flushing occurred in the dry season to early wet seasons. Most species (60%) flowered in April (hottest, driest time of year), when leafless or flushing with young leaves. Fruiting peaked in September (75% of species), whilst seed dispersal occurred over the late wet season to early dry season (August- January) (more than 50% of studied species). Consequently , most species required lengthy dormancy periods, to survive the dry season and germinate in the rainy season. Therefore, in order to accelerate seedling production in the nursery, treatments to break dormancy had to be developed.

Consequently, experiments to increase and accelerate seed germination were carried out on 30 indigenous forest tree species. Seven pre-treatments were tested, to promote seed germination. Scarification increased seed germination for Acrocarpus fraxinifolius, scarification + soaking for Afzelia xylocarpa, scarification alone and scarification + soaking for three species (Albizia chinensis, Elaeocarpus lanceifolius and Sindora siamensis), and scarification and/or acid treatment for 3 minutes for Cassia fistula. Accelerated and more synchronous germination was achieved for three species (Acrocarpus fraxinifolius, Albizia chinensis and Cassia fistula) because of the treatments. Despite low germination percentages (38-47%), ten other tree species might still qualify as potentially useful for forest restoration, due to other attributes, such as high growth rate in containers or good field performance. On the other hand, germination of Betula alnoides, Ficus hirta and Schleichera oleosa, were unacceptably low for all treatments (< 20%). Therefore, other seed pre-treatments or alternative propagation systems must, be considered for these species. The remaining three species had intermediate germination (Afzelia xylocarpa, Elaeocarpus lanceifolius and Sindora siamensis).

Shade dependence for germination and early seedling development would make a tree species unsuitable for forest restoration in open, degraded sites. Therefore, germination experiments were replicated in deep shade. Shade dependence was found only for one species, Elaeocarpus lanceifolius. Shade-tolerance was demonstrated for eighteen species. Only seven species were shade inhibited. However, four species produced mixed results. This indicates that very few tree species will be unable to grow in open degraded sites due to strong sunlight. It raises the possibility of planting pioneer and most climax tree species together in a single step for restoring forest to degraded sites.

To determine the influence of nursery conditions on germination and to investigate the possibility of direct seeding as an alternative to planting seedlings, seed experiments were also replicated in a forest gap. Fourteen species (47%) germinated better in the nursery than in the gap, five species germinated better in the gap than in the nursery and eleven species showed no difference in germination between nursery and gap. This indicates that nursery conditions generally enhance germination above natural levels.

The impact of seed predation on seed germination in the forest gap varied among species, with seed size and seed coat. The mean number of seeds removed was highest for Elaeocarpus prunifolius, Irvingia malayana, Reevesia pubescens and Terminalia chebula. Seven native tree species with high and rapid germination in the gap and no seed predation were identified as suitable for direct seeding. Except for the small seeds, burial did not seem to protect seeds from predators.

Some associations were found between ecological parameters and best treatments to break seed dormancy. Pre-treatments brought about significant improvement in germination of seeds with thick integuments (p=0.001), large and medium seed size (p=0.028) and seed dormancy (p=0.017). Prolonged dormancy was significantly associated with better seed germination under gap conditions (p=0.004) and with thick integuments (p=0.024). Better seed germination under nursery conditions (compared with the gap) was significantly associated with the small seed size (p=0.0024) and thin integuments (p=0.016). Heavy seed predation was strongly associated with large seed size (p=0.004) and thick integuments (endocarp) (p=0.040).

The effects of various hormone treatments on leafy stem cuttings varied among the species tested. Only five of ten tree species achieved a maximum of 60% or more of cuttings developing roots. Seradix #3 produced the best results with Debregeasia longifolia (68%) and Saurauia roxburghii (65%). IBA 3000 ppm produced the best results with Ficus superba (72%) and IBA 8000 ppm produced the best results with Colona flagrocarpa (63%). Also, Morus macroura cuttings (90%) grew roots most efficiently without any hormone treatment, with the non-treated control cuttings producing the highest success ranking scores. Unlike pre-treatments to promote seed germination, the chemical treatments to improve vegetative propagation showed no significant associations with ecological variables.

Relationships among ecological variables and best horticultural practices are clearly complex and will require further research, if useful, predictive models are to be developed.