Skip to Main content Skip to Navigation

Seed ecology and grassland resilience : the case of campo rupestre

Abstract : Our knowledge of tropical old-growth grassland resilience (i.e. resistance to degradation and capacity for recovery) remains limited relative to our knowledge of temperate grasslands. Although highly resilient to endogenous disturbances such fires, the vegetation dynamics in old-growth fire-prone tropical grasslands seems hampered by anthropogenic disturbances, especially topsoil degradation. After topsoil disturbance, the internal species pool (i.e. remaining vegetation and the seed bank) is often depleted or even absent, and natural regeneration depends mainly on seed dispersal from surrounding sites via the seed rain. However, plant communities on areas disturbed by soil removal can remain very different from preserved sites many years after degradation, with almost no recovery of the natural vegetation. Despite extremely relevant, seed dispersal dynamics and resilience of tropical old-growth grasslands after soil disturbance remains much overlooked. In this thesis, I carried out a review and three experimental studies about seed rain and diaspore removal dynamics in grasslands. In CHAPTER1, I did a systematic literature survey about seed rain studies in global grasslands. I (1) assessed where, how and why research on seed rain has been carried out; (2) examined how methodological design and results have been reported; and (3) provided guidelines for future research on seed rain in grasslands. I found a remarkable unbalance in the numbers of studies between grassland types, which becomes even more dissimilar across global climatic ranges when the area covered by each grassland type is take into consideration. I identified significant knowledge gaps in grassland seed rain research. I also found a great disparity of methods and data being reported across studies. Additionally, I found that only a few attempts have been made to assess the seed trap efficiency and no studies to date have proposed any protocol to evaluate the effectiveness of seed sorting methods. Then, in CHAPTER 2, I ran a field experiment to propose a simple standard protocol to evaluate the efficiency of two seed trap types (sticky and funnel traps) and of a seed sorting method to assess the efficient of the methods prior to seed rain studies. As a case of study, in CHAPTER 3 and 4, I studied seed rain and secondary diaspore removal dynamics in preserved areas and in areas degraded by gravel exploitation for road construction in campo rupestre vegetation, a megadiverse edaphic grassland in southeastern Brazil. The small number of seeds captured in the seed rain, indicates seed limitation and suggests a close causal relationship between seed dispersal limitation and the low resilience after soil disturbance. The identification of some relatively abundant species in the seed rain places these species as good targets for reintroduction in future restoration projects. Topsoil removal changed the identity of ants interacting with diaspores, resulting in contrasting outcomes from ant-diaspore interactions. The lack of diaspore removal towards degraded areas indicates that establishment limitation is a factor hampering natural regeneration. Our findings help to explain, at least partially, why natural regeneration is compromised after soil removal, which strongly influences diaspore fate and interactions with potential ground-dwelling dispersers, resulting in different ecological outcomes and strong influencing vegetation dynamics and regeneration. I expect that these results will guide future research on seed dispersal and resilience in grasslands, underpinning decisions on restoration and conservation practices on these threatened environments.
Document type :
Complete list of metadatas

Cited literature [283 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Tuesday, March 24, 2020 - 12:33:16 PM
Last modification on : Wednesday, May 6, 2020 - 1:16:43 AM
Long-term archiving on: : Thursday, June 25, 2020 - 1:53:57 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02517124, version 1


André Jardim Arruda. Seed ecology and grassland resilience : the case of campo rupestre. Earth Sciences. Université d'Avignon; Universidade federal de Minas Gerais. Facultade de educação (Brésil), 2019. English. ⟨NNT : 2019AVIG0353⟩. ⟨tel-02517124⟩



Record views


Files downloads