Studies were conducted to assess the possibility of commercial exploitation of fruits and seeds of forest species to meet the shortfall in industrial and edible vegetable oils in the country. Seeds from 15 forest species were collected from various forests of Bangladesh. In cases where the seed coat could be detached, the weight ratio of seed-coat to kernel was worked out. Oil content of kernel, pulp or whole seed was determined by solvent extraction process, depending on the ease or otherwise of separation of the coat. The free fatty acid content, saponification value and iodine value ofthe extracted oil samples were determined. High percentages of oils were obtained from most of the species under investigation. Slightly lower yield is likely to be obtained when oil is extracted by expression process.

Comparative lower yield of oil obtained from Shorea robusta, Ponganiea Pinnata, Aleurites montana and Garcinia cowa than that obtained in other countries is thought to be due to various degrees of deterioration set in the seeds collected from the forest bed. Acceptable grades of edible oils may be produced from several species.

Top-dying of Sundri (Heritiera fames Buch. -Ham.) has been occuring in an epidemic form in the Sunderbans forest for about a decade. It was suspected, in course of a multidisciplinary investigation made in 1973, that injury done by Chrysocroa sp. (Buprestidae : Coleoptera) into the twig croatches could be the primary cause, which subsequently facilitated the entrance of a probable pathogenic fungus causing the disorder.

A study was, therefore, made to ascertain the role of insects in top dying of Sundri. On investigation no correlation could be established between the number of larval tunnels and the number of branches found dead (r=0.147). The results of the investigation suggest that these beetles do not have any role in the initiation of topdying of Sundri.

Soil samples were collected from the genetic horizons of three soil profiles representing three different landtypes. Trace elements in soil and in leaf samples were determined spectrometrically.

It was observed that the vertical distribution of trace elements was related to the amount of clay rather than to the genetic soil horizons. The clay content in these profiles was the attribute ofthe alluvial parent material. Therefore, the redistribution of trace elements was not due to pedogenetic effects.

The clay fraction contains higher amounts of trace elements in all the profiles. It was, therefore, assumed that the topsoil might be depleted of trace elements with the loss of clay by illuviation or by run-off. Trace element/iron oxide ratios of several soil samples show that the vertical distribution of trace elements in the profiles studied were not significant.

It was also observed that in the studied profiles Mn and Cd content was higher but that of other elements waslower than the FAO’s prescribed highest levels of total content of trace elements in soil.

Trace element contents of several leaf samples of tree species were examined and found not related to the trace element contents of soil.

Satellite Remote Sensing data arc being extensively used in change detection and vegetation cover studies. Visual interpretation of LANDSAT imagery, black and white aerial photographs, colour Infra-red aerial photographs, diazo false colour composites of LANDSAT imagery, computer enhanced image, digitized image and stereoscopic photo interpretation techniques were adopted for the study of change detection and vegetation cover of “Nijhumdwip”. The purpose of the study was to find out the rate of accretion and the succession of vegetation in the island. Groundtruth missions were performed for the collection of data. Information generated by the various methods of interpretation were correlated with known field data.

A study was made on the emerging culm mortality of eight different bamboo species of Bangladesh. Mortality percentage varied from species to species and even from clump to clump. Average mortality percentage varied from 28 to 69 in the case of thick-walled and large dimension bamboo species, and 9 to 37 in the case of thinwalled, small dimension species.

Eco-physiological conditions and genetic make-up of each species and clump seem to influence the rate of mortality of emerging culm in bamboo.