Forest restoration in abandoned agricultural land: a case study from East Africa

Citation
Ca. Chapman et Lj. Chapman, Forest restoration in abandoned agricultural land: a case study from East Africa, CONSER BIOL, 13(6), 1999, pp. 1301-1311
Citations number
68
Categorie Soggetti
Environment/Ecology
Journal title
CONSERVATION BIOLOGY
ISSN journal
08888892 → ACNP
Volume
13
Issue
6
Year of publication
1999
Pages
1301 - 1311
Database
ISI
SICI code
0888-8892(199912)13:6<1301:FRIAAL>2.0.ZU;2-4
Abstract
Millions of hectares of tropical forests have been converted to agricultura l land and abandoned, so it is important that we understand the process of forest recovery and comprehend how pathways are modified by different types of disturbances in different geographic regions. In a 4-year case study, w e quantified the pattern of forest recovery following clearing and 3 years of cultivation of a moist-evergreen forest in Uganda. Long-term observation (746 hours) of frugivore visitation to the regenerating area demonstrated that birds were frequent visitors (5.8 birds/hour), whereas large mammalian frugivores rarely used the area. Frugivore visitation rates facilitated se edling recruitment that averaged 0.51 seedlings/m(2) from 22 tree species b y the end of the study. Recruitment included species with large seeds, desp ite the fact that seed-eating rodents were almost twice as abundant in the regenerating are than in mature forest. By the end of the study, however, o nly 20 trees were greater than or equal to 0.5 m tall, and no trees were gr eater than or equal to 2 m tall. This slow recruitment reflected high seedl ing mortality and dominance of the area by elephant grass (Pennisetum purpu reum) and the herb Acanthus pubescens. After 4 years, trees greater than or equal to 0.5 m tall attained a biomass of only 8.92 kg/ha, whereas the bio mass of P. purpureum and A. pubescens had reached 35,500 kg/ha and 18,100 k g/ha respectively. We provide an initial assessment of two programs designe d to enhance restoration of abandoned agricultural lands: planting of cutti ngs to act as dispersal foci and sowing of seeds. Our results showed that d ensity of seedlings growing in the management plot where we sowed seeds (0. 35 seedlings/m(2)) and in the plot where we established cuttings (0.30 seed ling/m(2)) was lower than in the control plot (0.51 seedlings/m(2)). This E ast African site was only lightly disturbed, yet tree recovery was occurrin g slower than in heavily degraded sites described from South America. The r ate of recovery seemed to be strongly determined by interactions between tr ee seedlings and P.purpureum and A.pubescens.