Similarity in seed removal patterns of four tree species in active and passive restoration treatments of the tropical seasonal forest after 23 years

Abstract

Seed removal has been considered an important ecological filter in plant community assembly during the first years of restoration. However, little is known about the influence of the restored ecosystems attributes on seed consumption patterns by predators and, consequently, on the composition and structure of the regenerating community. We tested whether seed removal differs between natural and restored ecosystems and whether it depends on the seed consuming agent (insects, rodents and birds). We evaluated seed removal in three forest restoration treatments (two active and one passive) established over 20 years ago. We used a fragment of old-growth native seasonal semideciduous tropical forest close to the experimental area, as a reference ecosystem. We installed exclusion stations for different consumer groups (insects, rodents and birds) of seeds belonging to four tree species from different successional groups and dispersal syndromes (Tabernaemontana hystrix Steud, Schinus terebinthifolius Raddi, Pterogyne nitens Tul and Anadenanthera colubrina var. cebil (Griseb.) Altschul. Insects were the main agent removing seeds from the species tested as a whole, regardless of the restoration treatment. Seed removal proportions were generally low and differed between active and passive restoration treatments, depending on the tree species. However, they were similar to the native forest, regardless of the restoration treatment. We found an effect of the interaction of restoration treatments and consuming agents only for one of the zoochoric species tested (Tabernaemontana hystrix). Seed removal no longer seems to represent a biotic filter limiting plant recruitment for T. hystrix, S. terebinthifolius, P. nitens and A. colubrina recruitment in the restoration treatments, regardless of the restoration technique.