Appropriate Tree Species and Techniques for Direct Seeding for Forest Restoration in Chiang Mai and Lamphun Provinces

ABSTRACT: Most forest restoration projects involve planting nursery-raised tree seedlings, but this is highly labour- and capital-intensive. Seed collection, raising seedlings in a nursery, planting and maintaining planted saplings until they can establish and become independent all require substantial labour inputs. So, direct seeding is an alternative that might reduce the cost of forest restoration. This research tested the hypothesis that appropriate species and techniques of direct seeding produce better results and are more cost-effective compared with planting nursery-raised seedlings, for forest restoration in northern Thailand.

Direct seeding was tested in both a highland and a lowland site. Tree species, selected for study in the highlands, were: Aquilaria crassna Pierre ex Lec. (Thymelaceae), Balakata baccata (Roxb.) Esser. (Euphorbiaceae), Carallia brachiata (Lour.) Merr. (Rhizophoraceae), Eugenia fruticosa DC. (Myrtaceae), Sarcosperma arboreum Bth. (Sapotaceae) and Spondias axillaris Roxb (Anacardiaceae). Those selected for testing at the lowland site were: Afzeliaxylocarpa (Kurz) Craib (Leguminosae, Caesalpinioideae), Artocarpus lakoocha Roxb. (Moraceae), Casearia grewiifolia Vent. var. grewiifolia (Flacourtiaceae), Eugenia cumini (L.) Druce (Myrtaceae), Schleichera oleosa (Lour.) Oken (Sapindaceae) and Trewia nudiflora L. (Euphorbiaceae). Germination tests were carried out both in the nursery and in the field. This study used two methods to break seed dormancy, accelerate seed germination and thus reduce the amount of time available for seed predation: i) soaking in water (48 hrs) and ii) scarification. In addition, weeding was carried out every 2 months in the field.

Seed pre-treatments (plus soil from mother tree) had little effect on some species, significantly increasing the survival percentage for none and significantly reducing it only for B. baccata in the field. This was probably due to seedling predation and desiccation. Weed control had different effects on different species, significantly increasing the survival per cent for only C. brachiata and significantly reducing it for A. crassna, B. baccata and S. axillaries. It had little effect on A. xylocarpa and S. oleosa. This was probably due to seedling desiccation, which affected each species differently, depending on their characteristics.

Comparison of seedling growth in the second year made use of an experiment that had been established in the previous year by direct seeding Gmelina arborea Roxb. (Verbenaceae), Melia toosendan Sieb. & Zucc. (Meliaceae), Oroxylum indicum (L.) Kurz (Bignoniaceae), Prunus cerasoides D. Don (Rosaceae), Sarcosperma arboreum Bth. (Sapotaceae) and Spondias axillaris Roxb (Anacardiaceae). Nursery-raised plants, from the same seed batches, were planted next to the direct seeded plants and monitored for a year. Direct seeded G. arborea, M. toosendan and P. cerasoides grew significantly better with higher mean RCD, height and crown width and had higher survival per cent, compared with raised-nursery seedlings during their second year of growth in the field (p<0.05).

Weed competition was not a serious problem in the first year after sowing. It had no effect on germination per cent, MLD and growth performance of most of the species in this study. Some species appeared to be nurtured and supported by surrounding vegetation, which might protect them from strong sunlight and high temperatures. Treatments used to increase germination per cent and reduce MLD had variable and inconsistent effects. Therefore, more research, to develop more reliable treatments to accelerate germination, are required.

Calculations of costs showed that establishment of direct seeded trees can reduce costs by about 40%, compared with planting raised-nursery seedlings. The calculation was based on a stocking density of 500 trees per rai. Direct seeding may be suitable for forest restoration projects with low budgets but systematic planning will be needed.