Persistence and genetic vulnerability of tropical tree species in response to anthropogenic disturbances

Tropical forests face numerous threats, including overexploitation, selective logging, and
fragmentation arising from landscape changes. These disturbances reduce tree population
density and increase distances between individuals, altering their reproductive behaviour.
Decreased numbers of mature trees and disrupted gene flow reduce reproductive success and
genetic diversity, potentially leading to inbreeding, genetic drift, or even extinction.
Consequently, landscape restoration techniques became a global priority. Tropical trees are
long-lived organisms, and there is a significant knowledge gap regarding their long-term
survival in the context of anthropogenic disturbances and restoration. This thesis aims to
contribute to the conservation and restoration of tropical tree species using a mixed-methods
approach, encompassing meta-analysis, simulation modelling, and survival analysis. All
chapters focus on species from the ecologically and economically important tree family
Dipterocarpaceae. Many species possess winged fruits, which disperse by wind and gyration,
and they generally reproduce during supra-annual masting flowering events. It has been
hypothesised that trees respond differently to disturbances based on traits like flower size and
wood density.
The first chapter provides a general introduction to the topics addressed in this thesis and
outlines the existing knowledge gaps. It gives an overview of the impacts of anthropogenic
disturbances and restoration efforts on trees and highlights the importance of modelling in
understanding and predicting the effects of various landscape uses on trees.
In Chapter 2, I conducted a literature and data synthesis to investigate how fragmentation and
logging impacted 277 adult and offspring populations of 45 dipterocarp species. I extracted
data on three different genetic metrics to examine how genetic vulnerability was affected by
anthropogenic disturbance and test whether genetic vulnerability was differentially explained
by the traits of flower size, fruit size and wood density. Through a meta-analysis using
phylogenetic linear mixed models, I found that genetic diversity of adult populations was
significantly lower in fragmented and in forests with combined disturbances, while juvenile
populations were generally unaffected. Additionally, I found a negative relationships between
fruit size and genetic diversity as well as wood density on inbreeding in adult populations.
While flower size and fruit size positively correlate to both genetic diversity and inbreeding in
offspring populations. These results suggest that these traits could serve as useful surrogates
for assessing genetic vulnerability despite the limited data on offspring populations.

VII

Chapter 3 employs RangeShifter, an individual-based spatially explicit simulation modelling
platform, to test how the amount and spatial configuration of habitat fragmentation, as well as
dispersal kernel, affect the persistence of Rubroshorea leprosula. Using survival analyses,
these simulations suggested that the most important factor for maintaining species persistence
in fragmented landscapes is the increase in proportion of suitable habitat, while the effect of
aggregation was minimal. Long-distance seed dispersal of dipterocarps becomes evident only
with the increased proportion of suitable habitat. These results highlight the importance of the
retention of large remaining habitat fragments to ensure long-term survival of species.
Chapter 4 extends the modelling using new features of the RangeShifter platform to explore
logging and restoration scenarios across landscapes with varying amounts and configurations
of fragmentation. I model the impacts of different logging regimes and active planting as a
restoration technique on species persistence. Results indicate that logging in an already
fragmented landscape accelerates species extinction, independent of the logging treatment.
Both the proportion of suitable habitat and its spatial arrangement play a crucial role in
mitigating extinction risk in restoration scenarios. Survival is higher in landscapes with
increased percentage of suitable habitat and when restored areas are in proximity to existing
forests. Active planting also decreases time to extinction, independent of the number of
seedlings planted.
Taken together the research in this thesis highlights the significance of forest fragmentation as
a threat to the persistence of tropical tree populations, as a risk factor for loss of genetic
diversity and increasing extinction debt. Simulation approaches were essential for identifying
both the demographic trends and identifying potential solutions through restoration approaches
under varying fragmentation scenarios. These approaches hold promise for designing strategies
to enhance the sustainability of tropical forest ecosystems and long-term survival of tropical
tree species.