Insect herbivory plays a crucial role in shaping plant communities in many terrestrial ecosystems. However, in African savannas, insect herbivory has been relatively understudied compared to large mammalian herbivory. In this study, we examined the impact of insect herbivory, focusing on leaf chewers and miners, in a South African savanna‐forest mosaic (including patches of forest, thicket and savanna) in Hluhluwe iMfolozi Park, South Africa. Our investigation spanned gradients of rainfall, fire frequency and mammal density. We surveyed a total of 864 woody plants from 48 plant species in 38 plots. Insects consumed 6% of leaf biomass, which is comparable to their impact in temperate broadleaf forests, but the extent of herbivory damage varied between vegetation types. Overall, leaf loss was 70% higher in forests and savanna than that in thicket. Plants in the forests experienced greater damage from chewing insects, whereas miners caused relatively more damage in savannas. Rates of insect herbivory also varied among plant species, declining with carbon and dry matter content but increasing with specific leaf area. Although no significant trade‐off was detected between insect and mammal herbivory, plant species with limited physical defences against mammals tended to experience high levels of insect herbivory. Our findings highlight the intricate dynamics of insect herbivory in different vegetation types and suggest that insect leaf herbivory, alongside mammalian herbivory, could play a significant role in influencing plant community composition and overall savanna ecosystem functioning.
Birds, bats and ants are recognised as significant arthropod predators. However, empirical studies reveal inconsistent trends in their relative roles in top‐down control across strata. Here, we describe the differences between forest strata in the separate effects of birds, bats and ants on arthropod densities and their cascading effects on plant damage. We implemented a factorial design to exclude vertebrates and ants in both the canopy and understorey. Additionally, we separately excluded birds and bats from the understorey using diurnal and nocturnal exclosures. At the end of the experiments, we collected all arthropods and assessed herbivory damage. Arthropods responded similarly to predator exclusion across forest strata, with a density increase of 81% on trees without vertebrates and 53% without both vertebrates and ants. Additionally, bird exclusion alone led to an 89% increase in arthropod density, while bat exclusion resulted in a 63% increase. Herbivory increased by 42% when vertebrates were excluded and by 35% when both vertebrates and ants were excluded. Bird exclusion alone increased herbivory damage by 28%, while the exclusion of bats showed a detectable but non‐significant increase (by 22%). In contrast, ant exclusion had no significant effect on arthropod density or herbivory damage across strata. Our results reveal that the effects of birds and bats on arthropod density and herbivory damage are similar between the forest canopy and understorey in this temperate forest. In addition, ants were not found to be significant predators in our system. Furthermore, birds, bats and ants appeared to exhibit antagonistic relationships in influencing arthropod density. These findings highlight, unprecedentedly, the equal importance of birds and bats in maintaining ecological balance across different strata of a temperate forest.
Hemiepiphytes have captured the attention of biologists since they seemingly hold clues to the evolution of epiphytes themselves. Hemiepiphytes are known to occur sporadically in the leptosporangiate ferns, but our understanding of their evolution remains limited by the relatively small number of detailed observations. This study adds to our knowledge by documenting seven species previously assumed to be holoepiphytes. This finding was based on fieldwork conducted in the Baining Mountains of Papua New Guinea that resulted in 319 collections representing 206 species. Approximately 3% of these species were hemiepiphytes: Asplenium acrobryum, A. amboinense, A. scandens, A. scolpendropsis, Crepidomanes aphlebioides, Leptochilus macrophyllus, and Sphaerostephanos scandens. All started growth as low-trunk epiphytes, and later, as larger climbing plants, exhibited strongly dimorphic roots consisting of short clasping ones that affixed the rhizome to the trunks and long feeding roots that entered the soil. Most of the seven hemiepiphyte species that we found exhibited distichous phyllotaxy and dorsiventrally flattened rhizomes, suggesting morphological convergence associated with this habit in four families. These new records suggest that large hemiepiphytic clades occur in Asplenium and Leptochilus. Our observations expand the geographic and taxonomic breadth of hemiepiphytic ferns, provide a baseline estimate of their diversity within a tropical flora, and offer morphological and phylogenetic clues to uncover additional records.
Birds, bats, and ants are recognized as significant arthropod predators. However, empirical studies reveal inconsistent trends in their relative roles in top-down control across strata. Here, we describe the differences between forest strata in the separate effects of birds, bats, and ants on arthropod communities and their cascading effects on plant damage. We implemented a factorial design to exclude vertebrates and ants in both the canopy and understory. Additionally, we separately excluded birds and bats from the understory using diurnal and nocturnal exclosures. At the end of the experiments, we collected all arthropods and assessed herbivory damage. Arthropods responded similarly to predator exclusion across forest strata, with a density increase of 81% on trees without vertebrates and 53% without both vertebrates and ants. Additionally, bird exclusion alone led to an 89% increase in arthropod density, while bat exclusion resulted in a 63% increase. Herbivory increased by 42% when vertebrates were excluded and by 35% when both vertebrates and ants were excluded. Bird exclusion alone increased herbivory damage by 28%, while the exclusion of bats showed a detectable but non-significant increase (by 22%). In contrast, ant exclusion had no significant effect on arthropod density or herbivory damage across strata. Our results reveal that the effects of birds and bats on arthropod density and herbivory damage are similar between the forest canopy and understory in this temperate forest. In addition, ants were not found to be significant predators in our system. Furthermore, birds, bats, and ants appeared to exhibit antagonistic relationships in influencing arthropod density. These findings highlight, unprecedentedly, the equal importance of birds and bats in maintaining ecological balance across different strata of a temperate forest.
Determining factors that regulate insect–plant interactions is of great interest in tropical forest ecology. However, our understanding of these factors across vertical stratification in tropical rainforests remains limited. We examined the effects of seasonality, microclimate, and leaf traits on insect herbivory and arthropod dynamics across vertical forest stratification in a weakly seasonal tropical rainforest in Papua New Guinea. We surveyed insect herbivory and arthropods on seven dominant tree species at 5 m increments from 1 m to 30 m, three times during both dry and wet seasons. We assessed insect herbivory on 9,425 leaves and collected 3,445 arthropods from 407.07 m2 of foliage. Herbivory decreased non-linearly with forest height and was highest during the wet season. Herbivorous arthropod density mirrored this pattern, peaking at the onset of the rainy season and remaining low at the start of the dry season and the end of the wet season. Predatory arthropod densities peaked in the canopy at the beginning of the wet season. Temperature, leaf dry matter content, and leaf toughness increased with forest height, while specific leaf area decreased. We conclude that forest stratification and seasonality play vital roles in mediating the rate of insect herbivory and arthropod community dynamics in the tropical lowland rainforest of Papua New Guinea.