Genetic effects of selective logging and pollen gene flow in a low-density population of the dioecious tropical tree Bagassa guianensis in the Brazilian Amazon

• Published in: Forest ecology and management 2008-04, Vol.255 (5), p.1548-1558
• Main Authors: Silva, Marivana Borges, Kanashiro, Milton, Ciampi, Ana Yamaguishi, Thompson, Ian, Sebbenn, Alexandre Magno
• Format: Article
• Language: English
• Subjects: alleles; Animal and plant ecology; Animal, plant and microbial ecology; Biological and medical sciences; Correlated matings; Dioecious species; dioecy; Forest harvesting and working in forest; forest trees; Forestry; Fundamental and applied biological sciences. Psychology; gene frequency; genetic markers; genetic variation; heterosis; inbreeding; logging; microsatellite repeats; pollen flow; population density; population genetics; sequence analysis; spatial data; Synecology; Tatajuba; Terrestrial ecosystems; tropical forests; Tropical tree species; TwoGener analysis
• ISSN: 0378-1127; 1872-7042
• DOI: 10.1016/j.foreco.2007.11.012

Forest logging reduces population density and increases the distance between co-specifics and so can cause the loss of alleles, and affect the genetic diversity, spatial genetic structure (SGS), mating system, and pollen flow of the population. These factors were studied in the tropical tree species Bagassa guianensis Aubl. occurring in a low-density population in the Brazilian Amazon forest. Genetic diversity was compared among offspring, juveniles and adult trees, before and after selective logging. Non-significant differences were observed between these samples. The harvest of 61% of the adult trees caused the loss of three alleles in the reproductive population. However, these alleles were present in juveniles and offspring and so were not lost from the population. SGS was detected up to 300 m before logging in the overall and adult populations. After logging, significant SGS was not observed. Deviations from random matings were evident throughout biparental inbreeding ( 1 − t ˆ s = 0.067 , P < 0.05), correlated mating ( r ˆ p ( m ) = 0.193 , P < 0.05), and differentiation in pollen gene pool ( Φ ˆ ft = 0.081 , P < 0.05). The effective number of pollen donors was estimated as 5–7 trees. The distance of pollen gene dispersal was estimated as 308–961 m, depending on the dispersal model used (normal and exponential) and assumed population density. The estimated neighbourhood pollination area ( A ep) ranges from 81 to 812 ha, depending also on the assumed population density. Reproduction by obligatory outcrossing, pollen immigration from trees outside of the plot, the long-pollen dispersal distance, and large A ep suggested that the species can be resilient to the impacts of logging.