Reduced dispersal of native plant species as a consequence of the reduced abundance of frugivore species in fragmented rainforest

Frugivorous animals disperse the seeds of the majority of rainforest plant species and hence play a key role in the trajectory of rainforest regeneration. This study investigated whether changes in the species composition of the frugivore community in fragmented rainforest in subtropical Australia is likely to impact the dispersal of native plant species. The potential of frugivorous bird and bat species to disperse the seeds of plant species in fragmented rainforest was assessed using published dietary information together with field surveys of frugivore abundance within intact forest, forest fragments and patches of regrowth. Frugivore species with reduced abundance in fragmented rainforest were the only known dispersers of 27 of the 221 native plant species in the data set (12% of species). These frugivore species were also major dispersers of plant species producing fruits wider than 10 mm and species from the families Rubiaceae, Lauraceae, Myrtaceae, Meliaceae, Lamiaceae and Vitaceae. Except for Rubiaceae, these plant taxa are also potentially dispersed by two of the frugivore species that were widespread in fragmented rainforest, Lopholaimus antarcticus and Ptilonorhynchus violaceus, although dispersal rates are likely to be lower in fragmented than in extensive rainforest. Consistent with other regions, large-seeded plants are susceptible to reduced dispersal in fragmented rainforest in subtropical Australia. However, we predict a smaller deficit in seed dispersal in fragmented forests than has been reported from other regions, due to factors such as functional overlap among frugivore species, the ability of many Australian rainforest vertebrates to persist in fragmented rainforest, and a lack of hunting in these forests. Nevertheless, rainforest fragmentation has reduced the abundance of a suite of frugivorous rainforest fauna, which in turn is likely to reduce the dispersal of a certain plant taxa and may alter patterns of plant regeneration in subtropical Australian rainforest fragments.     

The role of remnant trees in seed dispersal through the matrix: Being alone is not always so sad

Remnant trees within the non-forest matrix are common structures of forest landscapes which have high conservation value due to their supposed roles as biological legacies and stepping stones in fragmented forest scenarios. Fleshy-fruited remnant trees are dispersal foci for many forest plants, as seeds accumulate under their canopies after visitation by forest frugivores. Despite this recognized effect, little is known about the relative role of remnant trees in maintaining their seed dispersal function after forest fragmentation. In this work, conducted in the Cantabrian Range (Northern Spain), seed deposition by frugivorous birds and post-dispersal seed predation by rodents were compared beneath the canopies of hawthorn (Crataegus monogyna) trees scattered in the non-forested matrix, to those embedded in forest patches. We studied two years which had strong differences in community-wide fruit abundance (high in 2004, and low in 2005). Hawthorn and holly (Ilex aquifolium) seeds dominated the seed rain. The density of dispersed seeds differed between years and tree type, with higher values under patch trees relative to remnant ones. However, the effect of tree type depended on both the year and the seed species, as larger differences were found in the year of high fruit availability, and with holly seeds. Higher levels of post-dispersal seed predation on holly seeds also contributed to offset the differences between tree types. Our results suggest that remnant trees, by functioning as dispersal foci, facilitate the dispersal of the two most abundant plant species for forest succession through the matrix. More importantly, we demonstrated that in years of low fruit availability in which forest frugivores are forced to exploit scattered fruit resources, the role of remnant trees may even be equivalent to that played by forest trees.     

Abiotic and vertebrate seed dispersal in the Brazilian Atlantic forest: implications for forest regeneration

This study documents the proportion of woody plant species having different modes of seed dispersal, and classes of seed and fruit size at 16 Atlantic montane forest plots of southeastern Brazil (23-24° S). These plots represent six chronosequences (from 5-year-old to old-growth forest) of forest regeneration following small-scale shifting agriculture. Our results indicate that there is a gradual but predictable increase in the number of woody plant species relying on vertebrate-mediated seed dispersal in increasingly older plots. Moreover, the percentage of plant species bearing small seeds and fruits (< 0.6 cm in length) was more than halved from earlier to older forest plots, while the percentage of plant species producing middle-sized seeds and fruits at least doubled. Plant species belonging to the Melastomataceae, Myrsinaceae, Rubiaceae and Flacourtiaceae comprised a large proportion of the small-seeded and fruit species (33-53%), and largely occurred in early successional forest plots (5-30 years old). In contrast, species belonging to the Myrtaceae and Lauraceae represented most of the species (51-72%) bearing medium-sized seeds and fruits (0.6-1.5 cm in length), and occurred primarily in the old-growth forests. Shifts in the relative importance of dispersal strategies during the regeneration process of Atlantic montane forest, as well as in diaspore size appear to be related to the balance between early successional and shade-tolerant species associated with particular life forms and plant families. Finally, we discuss the relationships between the species richness of Myrtaceae and Lauraceae trees, seed dispersal by medium to large vertebrates, and possible regeneration scenarios for the Atlantic forest.     

Balanites wilsoniana: Regeneration with and without elephants

Tropical forest trees that produce large fruits and/or large seeds are believed to be dependant on a few and often rare large vertebrates for dispersal. However, little is known about the population dynamics of such trees when they lose their specialised dispersers. This study examines the juvenile spatial distribution of Balanites wilsoniana, Dawe & Spraque, which is believed to be dependant solely on elephants for dispersal, in forests with and without elephants. Using camera traps, elephants were confirmed to be the only frugivores feeding and thus dispersing Balanites wilsoniana fruits. There was a high density of seedlings beneath adult trees in one forest without elephants, which was attributed to low seed and seedling predation in this forest. Nevertheless, it was only in the forest with elephants that juveniles were established away from adult trees. Analysis of the spatial distribution of these juveniles by size revealed that saplings and poles are more likely to be found away from adult trees thus providing evidence that dispersal enhances juvenile survival. This study underlines the importance of seed dispersal for trees that cannot maintain their populations through seedlings germinating and surviving beneath adult conspecifics. The study also demonstrates that loss of vertebrates has ramifications for tree species dynamics above and beyond loss of seed dispersers.     

Biodiversity and abundance of terrestrial isopods along a gradient of disturbance in Sabah,East Malaysia

Connell's intermediate disturbance hypothesis predicts that the highest diversity is maintained at intermediate levels of disturbance. We have examined this hypothesis by observing differences in biodiversity of terrestrial isopods along a gradient of disturbance from two undisturbed primary tropical rainforest sites, to a logged site, a mixed native fruit orchard and a commercial oil palm plantation, in Sabah, East Malaysia. We describe a standardised protocol for the rapid assessment of isopod biodiversity on tropical forest floor sites and for measuring environmental variables to which we have related differences in species richness and relative abundance of the isopods. The results do not support Connell's hypothesis because there were no significant differences in diversity, species richness or equitability between disturbed sites and the nearest primary forests. The relative abundance of individual species was highest in the most disturbed environment. We suggest that this may be because particular species are well adapted to exploiting resources under the more ‘r’ selection conditions created by disturbance. Possible reasons for why the observations do not conform with predictions from the intermediate disturbance hypothesis are discussed. We conclude that Huston's dynamic equilibrium model is more appropriate than the intermediate disturbance hypothesis in predicting the effects of disturbance of tropical rainforests on these arthropod macro-decomposers.     

Physico-chemical properties of tamarind (Tamarindus indica L.) fruits from Mali: selection of elite trees for domestication

Tamarind (Tamarindus indica L.) is a tropical fruit tree highly valued for its fruit pulp. It has been identified as one of the priority species with great potential for domestication in the Sahelian countries of Sub-Saharan Africa. An important step in the domestication process is the characterization of the local natural variation of the species in order to select the most interesting phenotypes/genotypes which possess several desired traits for breeding purposes. Four provenances covering contrasting agro-ecological zones in Mali were selected for fruit morphological and nutritional traits (tartaric acid and sugars). Tamarind trees from the driest provenance contained smaller fruits with a smaller amount of pulp compared to provenances from wetter zones. Tamarind trees in Mali have a low real pulp value (maximum 9.5) and bear sour fruits (high tartaric acid content of 9–12 %). Some sweet-fruited trees could be identified, but the sweetness was low (maximum 8). Fruit traits seem to be influenced by climate and soil characteristics, which should be taken into account in the domestication process, when planting trees in a different region of origin. We selected a few elite trees within each provenance, containing the best combination for the most desirable fruit traits, using web diagrams. Frequency distributions of the different fruit traits showed some traits are probably the target of selection by farmers. All provenances, except the driest one, indicated a possible very first stage of domestication.     

Do frugivorous birds assist rainforest succession in weed dominated oldfield regrowth of subtropical Australia?

Exotic plants often form the first woody vegetation that grows on abandoned farmland. If this vegetation attracts vertebrate frugivores which disperse the seeds of native plants, then native plants may recruit to such oldfield sites. However, there is debate about the extent to which exotic vegetation assists or suppresses the regeneration of native plants, and about its effects on faunal biodiversity. These issues were investigated in subtropical eastern Australia, where rainforests were cleared for agriculture in the 19th century, and where regrowth dominated by camphor laurel (Cinnamomum camphora, an exotic, fleshy-fruited tree) has become common on former agricultural land. The study assessed the assemblages of frugivorous birds, and the recruitment of rainforest plants, at 24 patches of camphor laurel regrowth. The patches were used by nearly all frugivorous birds associated with subtropical rainforest. Many of these birds (16 of 34 species) are considered to have a medium to high potential to disperse the seeds of rainforest plants, and eight of these were abundant and widespread in regrowth patches. Of 208 recorded plant species, 181 were native to local rainforest. The ratio of native to exotic species was higher amongst tree recruits than adult trees, both for numbers of species and individuals. Among native tree recruits, 79% of 75 species, and 93% of 1928 individuals, were potentially dispersed by birds. These recruits included many late-successional species, and there were relatively more individuals of late-successional, bird-dispersed native species amongst recruits than adult trees. The species richness, but not the abundance, of both frugivorous birds and of bird-dispersed rainforest trees decreased with distance from major rainforest remnants. Camphor laurel regrowth provides habitat for rainforest birds and creates conditions suitable for the regeneration of native rainforest plants on abandoned farmland. Careful management of regrowth dominated by fleshy-fruited exotic invasive trees can provide an opportunity for broadscale reforestation in extensively-cleared landscapes.     

Are habitat-tracking metacommunities dispersal limited? Inferences from abundance-occupancy patterns of epiphylls in Amazonian forest fragments

In theory, habitat fragmentation alters plant community dynamics by influencing both local (within patch) and regional (among patch) processes. However, the lengthy generation times of plant taxa relative to the short duration of most experiments has precluded studies from assessing the impact of fragmentation at both local and regional scales. Due to their accelerated life cycles, high rates of local extinction, and naturally patchy substrates, epiphyllous bryophyte assemblages are an appropriate plant guild for empirically testing metacommunity-based predictions associated with habitat fragmentation. By examining the local abundance and regional distribution patterns of 67 epiphyllous (leaf-inhabiting) bryophyte species in an experimentally fragmented landscape in Amazonia, we demonstrate that changes in local abundance wrought by habitat fragmentation are best explained by fragment size rather than proximity to forest edge. Furthermore, evidence of a simultaneous inter-specific decline in epiphyll local abundance and regional distribution in small (1- and 10-ha) forest fragments corroborate with metapopulation-based predictions highlighting the importance of immigration in buffering from patch extinction risk (i.e., the rescue effect). Collectively, these results provide indirect evidence that dispersal limitation, rather than compromised habitat quality attributable to edge effects, likely account for species loss from small tropical forest fragments. Whether dispersal limitation is due to increased insularity from regional sources for epiphyll recolonization or rather to lowered within-fragment dispersal potential is unknown; nonetheless, the long-term persistence of these microscopic plant metacommunities depends on the preservation of rain forest reserves of at least 100-ha in size.     

Effect of forest fragmentation on fruit and seed predation of the tropical dry forest tree Ceiba aesculifolia

Forests fragmentation reduces the density of natural plant populations forming patches of the remaining individuals. One of the biotic interactions that can be affected by forest fragmentation and is poorly studied is seed predation. We determined the effects of forest fragmentation on seed and fruit predation in Ceiba aesculifolia by comparing trees in continuous forest with trees in fragmented forest. We compared the following variables: (a) frequency of fruit predation by Collie’s squirrel (Sciurus colliaei) in each habitat; (b) frequency of the cotton-staining bug seed predator (Dysdercus, Orden Hemiptera) in each habitat; (c) the effect of seed predation on germination frequency and time; and (d) the effect of different life stages of Dysdercus on seed viability. In continuous habitat, 100% of the trees presented fruits with squirrel predation while only 34% of trees in fragmented habitats presented fruit predation. In continuous forest 27% of the trees contained fruits with the seed predator Dysdercus, while only 2% of the trees in fragmented forest presented Dysdercus. The initial weight of damaged seeds was greater than seeds that were not damaged indicating that seed predators select heavier seeds to feed upon. Frequency of seed germination was affected by different life stages; pre-adults decreased germination significantly more than nymphs and adults. Seed predation significantly increased the time it took for germination to occur. Our study shows that forest fragmentation significantly affects predation patterns of squirrels and cotton-staining bugs. Reduction of natural seed predators in forest fragments may have long-term consequences on forest structure and diversity.     

The influence of forest management history on the integrity of the saproxylic beetle fauna in an Australian lowland tropical rainforest

Forest management in temperate and boreal regions is often based on a strong foundation of applied ecological research. Increasingly, this has allowed the needs of saproxylic (dead wood associated) insects to be addressed. However, there has been very little equivalent research in tropical forests, where saproxylic insect faunas are likely to be much richer and where forestry is usually subject to weaker environmental controls. This study compares the saproxylic beetle fauna of old-growth, selectively logged and regrowth rainforest in the Daintree lowlands of northeastern Queensland, Australia. Old-growth levels of abundance, species richness, assemblage composition and guild structure were not maintained in logged and regrowth forest, suggesting that intact assemblages may not survive in the long-term in managed tropical rainforest. However, retaining a continuous supply of commercially overmature trees in the managed stand may prevent a repeat of the widespread extinctions of saproxylic insects witnessed in temperate and boreal forest regions.     

Species compensation maintains abundance and raid rates of African swarm-raiding army ants in rainforest fragments

Habitat clearance and fragmentation is increasingly threatening the biodiversity of tropical rainforests, however, the response of many insect species, even of key organisms, is still little understood. Using an extensive data set spanning over four years, we analyzed the effects of clearance and fragmentation of a Congo-Guinean rainforest in western Kenya on the abundance and the raid rates of two species of swarm-raiding army ants, which are considered keystone organisms of tropical rainforests. The abundance of army ants was measured by transect monitoring and by short-term pitfall trapping while raid rates were measured by long-term pitfall trapping over a period of five months. Dorylus wilverthi was the most abundant army ant in undisturbed rainforest and its abundance and raid rate strongly declined in small forest fragments. In contrast, the abundance of Dorylus molestus increased with decreasing fragment size and compensated for the decline of D. wilverthi in terms of abundance and ecological functionality (i.e. raiding rates). D. molestus appears to have a higher ability of using the agricultural land surrounding the Kakamega Forest than D. wilverthi, which may explain the species-specific differences in the susceptibility to habitat fragmentation. Our study demonstrates that habitat fragmentation may have a differential effect on two ecologically highly similar keystone species. Moreover, it shows that species compensation might help in maintaining an important ecosystem function (i.e. raiding by swarm-raiding army ants) in fragmented tropical rainforests.     

Tree species-mediated spatial patchiness of the composition of microbial community and physicochemical properties in the topsoils of a tropical montane forest

To evaluate the importance of plant-soil feedbacks in forest ecosystems, it is fundamental to understand the spatial range within which plant species control soil physicochemical and microbial properties. We investigated the spatial pattern of soil properties associated with canopy trees in a tropical montane forest on Mt. Kianbalu, Borneo. We analyzed soil physicochemical properties and microbial communities (biomarker lipid abundance) as a function of soil depth and distance from the tree trunk of a conifer (Dacrydium gracilis) or a broadleaf tree (Lithocarpus clementianus). The concentration of condensed tannins and fungi-to-bacteria were higher beneath Dacrydium than beneath Lithocarpus. Furthermore, carbon-degrading enzyme activities were lower beneath Dacrydium. These effects of the tree species were more distinct on soil properties beneath the tree crown than on those outside the tree crown. These effects appeared to be largely due to differences in litter chemistry, and the distinct set of soil properties formed corresponding to the above canopy crown. In conclusion, the species-rich forest on the tropical mountain contains spatially distinct units of soil properties associated with canopy trees, and this spatial pattern can influence ecosystem dynamics in the forest through plant-soil feedback effects.     

Converting Australian tropical rainforest to native Araucariaceae plantations alters soil fungal communities

Native rainforest tree plantations are increasingly viewed as potentially important for high value timber production and provision of a range of ecological services in tropical and subtropical areas. In order to determine the extent to which conversion of rainforest to native Araucariaceae plantation influences soil fungi, we compared soil fungal communities under native rainforest and 73-74 year-old Araucaria bidwillii, Araucaria cunninghamii and Agathis robusta plantations at Gadgarra State Forest, Queensland, Australia. Following direct extraction of DNA from soil, terminal restriction fragment length polymorphism (T-RFLP) analysis of rDNA internal transcribed spacer (ITS) regions was conducted. Ordination analysis of the T-RFLP data revealed significant separation of the fungal communities according to forest type along the first canonical axis, with the native rainforest samples separating from the three Araucariaceae plantations along the second axis. Overall, the most abundant ITS sequences in clone assemblages from the four forest types were Ascomycota, followed by Basidiomycota, Zygomycota and Chitridomycota, however their relative importance varied in individual forest types. The results indicate that conversion of tropical rainforest to monoculture plantations of native trees can significantly alter soil fungal diversity.     

BORON UPTAKE AND DISTRIBUTION IN FIELD GROWN CITRUS TREES

In low fertility tropical soils, boron (B) deficiency impairs fruit production. However, little information is available on the efficiency of nutrient application and use by trees. Therefore, this work verified the effects of soil and foliar applications of boron in a commercial citrus orchard. An experiment was conducted with fertigated 4-year-old ‘Valencia’ sweet orange trees on ‘Swingle’ citrumelo rootstock. Boron (isotopically-enriched ¹⁰B) was supplied to trees once or twice in the growing season, either dripped in the soil or sprayed on the leaves. Trees were sampled at different periods and separated into different parts for total B contents and ¹⁰B/¹¹B isotope ratios analyses. Soil B applied via fertigation was more efficient than foliar application for the organs grown after the B fertilization. Recovery of labeled B by fruits was 21% for fertigation and 7% for foliar application. Residual effects of nutrient application in the grove were observed in the year after labeled fertilizer application, which greater proportions derived from the soil supply.     

Matrix analysis to evaluate sustainability: The tropical tree Aquilaria crassna, a heavily poached source of agarwood

We used size-structured Lefkovitch projection matrix analysis to predict future trends in the survival of Aquilaria crassna (Thymelaeaceae), a tropical evergreen forest tree that has been highly sought after for its valuable aromatic wood (agarwood) for millennia. Data on growth, damage to trees by poachers, fruit production, seed dispersal and seedling recruitment were collected from a 30-ha plot in Khao Yai National Park, central Thailand that had seen moderate poaching. The population asymptotic growth rate, λ_a, and transient growth rate, λ_tr, found from using the matrix to project the observed population 48 years, were 1.002 and 1.005, respectively, with 95% confidence intervals of [0.985, 1.016] and [0.985, 1.023]. The stable size distribution obtained from the matrix was reasonably similar to the observed size distribution, suggesting that at the time of the study the population was not far from equilibrium. Sensitivity analysis and simulations of poaching on adult trees indicate that both λ’s are very sensitive to the removal of adult trees and the growth of preadult trees. In particular, the increase in tree mortality and decrease in preadult growth rate caused by agarwood collection seen in some areas could cause extirpation of the population. While in such cases the population appears to be poised on the brink of decline, several factors that we could not evaluate could alter its fate. First, adult deaths could stimulate recruitment of young under the parent trees; second, there is marked year-to-year variation in fruiting success and possibly recruitment; third, there is considerable spatial variation in tree density and recruitment, suggesting that attempts to evaluate future success need to encompass larger spatial and time scales.     

Remnant sugar maple (Acer saccharum subsp. skutchii) populations at their range edge: Characteristics,environmental constraints and conservation implications in tropical America

Relationships between environmental conditions and persistence of populations at edges of ranges should be useful for predicting consequences of climate change. We characterized environmental conditions of six disjunct sites in Mexico and Guatemala that contain all known relict populations of the most southern subspecies of sugar maple, cloud forest sugar maple (Acer saccharum subsp. skutchii). We also sampled abundance and basal area of all tree species, recording 164 species, 92 genera (61% tropical), and 54 families at these sites. Temperate genera decreased in number from north to south, but mixtures of temperate and tropical species were co-dominant with cloud forest sugar maple at all sites. For all species, basal area was 24–52 m² ha^?1 and density was 990–2986 trees/ha at the six sites. Cloud forest sugar maple comprised 7–43% of total basal area and 1–16% of total densities at each site. Populations of cloud forest sugar maple currently are vulnerable to environmental change. Anthropogenic disturbance is negatively affecting four populations, and regeneration is successfully occurring in only two sites. As a result, densities and basal area are much reduced compared to more northern temperate populations. NMS-ordination indicated that elevation and hurricane frequency are major environmental constraints related to abundance. Maple recruitment is associated with short hurricane return intervals that maintain higher elevation cloud forests in states conducive for regeneration. We establish the conservation status of cloud forest sugar maple at the southern end of its range relative to expected climate change and propose this subspecies be included in the IUCN-Red List.     

How much do site age, habitat structure and spatial isolation influence the restoration of rainforest beetle species assemblages?

Reforestation has been suggested as a way to mitigate the effects of tropical rain forest loss. However, factors influencing the successful colonisation of biota into newly created rainforest are poorly understood. We assessed beetle species assemblages with flight interception traps, in sites undergoing rainforest restoration across the largely cleared landscape of the Atherton Tablelands in north-eastern Queensland, Australia. There were two levels of site age (2-4 years and 6-17 years) and two distances from intact rainforest (adjacent and >0.9 km), with six sites in each treatment, together with six reference sites in each of pasture and small rainforest remnants (n = 36 sites). Multivariate metrics were used to assess the restored sites’ relative similarity to pasture and rainforest, in terms of both physical habitat structure and beetle species composition within ten family groups. Older restoration sites were structurally most similar to rainforest. Older sites and those adjacent to rainforest had a more rainforest-like beetle species composition (without significant interaction). However, even the closer and older sites had a substantially lower abundance and richness of rainforest-associated beetles than did rainforest. Age effects were generally stronger than distance effects, with the latter appearing to be entirely driven by rare rainforest species. Beetle assemblage similarity to rainforest was more strongly correlated with structural similarity to rainforest than with age, except within older sites. The use of revegetation techniques which lead to more rainforest-like structural conditions appears to be of over-riding importance in catalysing the rapid acquisition of volant rainforest beetle assemblages in the initial stages of restoration.     

Old manor parks as potential habitats for forest flora in agricultural landscapes of Estonia

Forest biodiversity conservation in intensively managed agricultural landscapes is a constant concern. The dispersal ability of forest plants is, hypothetically, the major limiting factor in fragmented forest landscapes and, therefore, we tested the validity of the theoretical dispersal scheme for plants in fragmented landscapes: ancient forest > woody corridor > emerging forest patch. To this end, we examined the distribution pattern of forest-dwelling plant species in rural landscapes, specifically the occurrence of common forest plant species in old historical forests and in closed-canopy stands of rural ornamental parks, planted on an agricultural land one–two centuries ago.We found that (i) the shade tolerant plant flora in parks’ stands more resemble forests than woody linear habitats (corridors), (ii) nearly 50% of the local forest species pool was present in parks, (iii) the abundance of seed source habitats and the ecological quality of the target habitat determine success rate of colonization. Models predicted that optimal stand characteristics for forest herbs are a minimum area 2.5 ha, canopy closure 0.7–0.8, basal area of trees 10–20 m² ha^?1 and the presence of moderate understory management.We conclude that only patch-type habitats provide suitable environmental conditions for forest-specific plant species. Many common forest plant species are capable of long-distance dispersal between habitat patches across hostile agricultural land, and accordingly, their dispersal follows a modified scheme of the island biogeography, without intermediary role of corridor habitats. Old rural manor parks provide an ecosystem service for nature conservation by harbouring forest biodiversity, and should be considered as potential refugium habitats.     

Morphological characterization of African bush mango trees (Irvingia species) in West Africa

The variation of the morphological characters of bitter and sweet African bush mango trees (Irvingia species) was investigated in the Dahomey Gap which is the West African savannah woodland area separating the Upper and the Lower Guinean rain forest blocks. African bush mangoes have been rated as the highest priority multi-purpose food trees species that need improvement research in West and Central Africa. A total of 128 trees from seven populations were characterized for their bark, fruits, mesocarp and seeds to assess the morphological differences between bitter and sweet trees and among populations. Multivariate analysis revealed that none of the variables: type of bark, mature fruit exocarp colour, fruit roughness and fresh mesocarp colour, could consistently distinguish bitter from sweet trees in the field. The analysis of the measurements of fruits, mesocarps and seeds demonstrated that bitter fruits have the heaviest seeds and this consistently distinguishes them from sweet fruits. However, the measurements of the fruit, mesocarp and seed did not have a joint effect in grouping types and populations of ABMTs. This indicates high diversity with a potential for selection existing across all phytogeographical regions investigated. The sweet trees of Couffo and those of Dassa in Benin are clearly different from all other populations. This can be attributed to traditional domestication (bringing into cultivation) and climate, respectively. The large fruits and the heavy seeds of the cultivated populations are evidence of successful on-going traditional selection of sweet trees in the Dahomey Gap.     

The effect of habitat fragmentation on communities of mutualists: Amazonian ants and their host plants

The consequences of fragmentation for communities of mutualist partners are for the most part unknown; moreover, most studies addressing this issue have been conducted on plant-pollinator communities. We evaluated how the experimental fragmentation of lowland Amazonian rain forest influenced a community of ant-plant mutualists. We inventoried a total of 1057 myrmecophytes in four fragments and four continuous forest sites; the twelve plant species recorded were occupied by 33 ant morphospecies, of which 11 were obligate plant inhabitants. Neither plant species richness, ant species richness, nor total ant-plant density were significantly lower in forest fragments. However, eight of the plant species, including three of the four most common, had higher mean densities in continuous forest than fragments. Of these four species, only one (Cordia nodosa) had significantly different colonization rates between habitat types, with higher colonization rates of plants in fragments. This may be because the Azteca species it is associated with increases in abundance in forest isolates. Although our results suggest that communities of ant-plant mutualists are likely to persist in fragmented tropical landscapes 25 years after fragment isolation, most species are rare and populations sizes in fragments are extremely low. Environmental and demographic stochasticity could therefore limit long-term population viability. We suggest future studies focus on evaluating how fragmentation has altered herbivore pressure and the dispersal of ants and plants to fragments, since the interaction of these factors is likely to have the greatest impact on long-term patterns of population persistence.