Interaction network of vascular epiphytes and trees in a subtropical forest

The commensalistic interaction between vascular epiphytes and host trees is a type of biotic interaction that has been recently analysed with a network approach. This approach is useful to describe the network structure with metrics such as nestedness, specialization and interaction evenness, which can be compared with other vascular epiphyte-host tree networks from different forests of the world. However, in several cases these comparisons showed different and inconsistent patterns between these networks, and their possible ecological and evolutionary determinants have been scarcely studied. In this study, the interactions between vascular epiphytes and host trees of a subtropical forest of sierra de San Javier (Tucuman, Argentina) were analysed with a network approach. We calculated metrics to characterize the network and we analysed factors such as the abundance of species, tree size, tree bark texture, and tree wood density in order to predict interaction frequencies and network structure. The interaction network analysed exhibited a nested structure, an even distribution of interactions, and low specialization, properties shared with other obligated vascular epiphyte-host tree networks with a different assemblage structure. Interaction frequencies were predicted by the abundance of species, tree size and tree bark texture. Species abundance and tree size also predicted nestedness. Abundance indicated that abundant species interact more frequently; and tree size was an important predictor, since larger-diameter trees hosted more vascular epiphyte species than small-diameter trees. This is one of the first studies analyzing interactions between vascular epiphytes and host trees using a network approach in a subtropical forest, and taking the whole vascular epiphyte assemblage of the sampled community into account.     

The network structure of plant-arbuscular mycorrhizal fungi

Ecological network theory predicts that in mutualistic systems specialists tend to interact with a subset of species with which generalists interact (i.e. nestedness). Approaching plant-arbuscular mycorrhizal fungi (AMF) association using network analyses will allow the generality of this pattern to be expanded to the ubiquitous plant-AMF mutualism. Based on certain plant-AMF specificity recently suggested, networks are expected to be nested as a result of their mutualistic nature, and modular, with certain species interacting more tightly than others. Network analyses were used to test for nestedness and modularity and to compare the different contribution of plant and AMF to the overall nestedness. Plant-AMF networks share general network properties with other mutualisms. Plant species with few AMFs in their roots tend to associate with those AMFs recorded in most plant species. AMFs present in a few plant species occur in plant species sheltering most AMF (i.e. nestedness). This plant-AMF network presents weakly interlinked subsets of species, strongly connected internally (i.e. modularity). Both plants and AMF show a nested structure, although AMFs have lower nestedness than plants. The plant-AMF interaction pattern is interpreted in the context of how plant-AMF associations can be underlying mechanisms shaping plant community assemblages.     

Invariant properties of mycobiont-photobiont networks in Antarctic lichens

Aim

Lichens are often regarded as paradigms of mutualistic relationships. However, it is still poorly known how lichen-forming fungi and their photosynthetic partners interact at a community scale. We explored the structure of fungus-alga networks of interactions in lichen communities along a latitudinal transect in continental Antarctica. We expect these interactions to be highly specialized and, consequently, networks with low nestedness degree and high modularity.

Location

Transantarctic Mountains from 76° S to 85° S (continental Antarctica).

Time Period

Present.

Major Taxa Studied

Seventy-seven species of lichen-forming fungi and their photobionts.

Methods

DNA barcoding of photobionts using nrITS data was conducted in 756 lichen specimens from five regions along the Transantarctic Mountains. We built interaction networks for each of the five studied regions and a metaweb for the whole area. We explored the specialization of both partners using the number of partners a species interacts with and the specialization parameter d'. Network architecture parameters such as nestedness, modularity and network specialization parameter H₂' were studied in all networks and contrasted through null models. Finally, we measured interaction turnover along the latitudinal transect.

Results

We recovered a total of 842 interactions. Differences in specialization between partners were not statistically significant. Fungus-alga interaction networks showed high specialization and modularity, as well as low connectance and nestedness. Despite the large turnover in interactions occurring among regions, network parameters were not correlated with latitude.

Main Conclusions

The interaction networks established between fungi and algae in saxicolous lichen communities in continental Antarctica showed invariant properties along the latitudinal transect. Rewiring is an important driver of interaction turnover along the transect studied. Future work should answer whether the patterns observed in our study are prevalent in other regions with milder climates and in lichen communities on different substrates.     

Evaluating factors that predict the structure of a commensalistic epiphyte–phorophyte network

A central issue in ecology is the understanding of the establishment of biotic interactions. We studied the factors that affect the assembly of the commensalistic interactions between vascular epiphytes and their host plants. We used an analytical approach that considers all individuals and species of epiphytic bromeliads and woody hosts and non-hosts at study plots. We built models of interaction probabilities among species to assess if host traits and abundance and spatial overlap of species predict the quantitative epiphyte–host network. Species abundance, species spatial overlap and host size largely predicted pairwise interactions and several network metrics. Wood density and bark texture of hosts also contributed to explain network structure. Epiphytes were more common on large hosts, on abundant woody species, with denser wood and/or rougher bark. The network had a low level of specialization, although several interactions were more frequent than expected by the models. We did not detect a phylogenetic signal on the network structure. The effect of host size on the establishment of epiphytes indicates that mature forests are necessary to preserve diverse bromeliad communities.     

Factors influencing epiphytic bryophyte and lichen species richness at different spatial scales in managed temperate forests

The effect of management related factors on species richness of epiphytic bryophytes and lichens was studied in managed deciduous-coniferous mixed forests in Western-Hungary. At the stand level, the potential explanatory variables were tree species composition, stand structure, microclimate and light conditions, landscape and historical variables; while at tree level host tree species, tree size and light were studied. Species richness of the two epiphyte groups was positively correlated. Both for lichen and bryophyte plot level richness, the composition and diversity of tree species and the abundance of shrub layer were the most influential positive factors. Besides, for bryophytes the presence of large trees, while for lichens amount and heterogeneity of light were important. Tree level richness was mainly determined by host tree species for both groups. For bryophytes oaks, while for lichens oaks and hornbeam turned out the most favourable hosts. Tree size generally increased tree level species richness, except on pine for bryophytes and on hornbeam for lichens. The key variables for epiphytic diversity of the region were directly influenced by recent forest management; historical and landscape variables were not influential. Forest management oriented to the conservation of epiphytes should focus on: (i) the maintenance of tree species diversity in mixed stands; (ii) increment the proportion of deciduous trees (mainly oaks); (iii) conserving large trees within the stands; (iv) providing the presence of shrub and regeneration layer; (v) creating heterogeneous light conditions. For these purposes tree selection and selective cutting management seem more appropriate than shelterwood system.     

Network Dynamics Contribute to Structure: Nestedness in Mutualistic Networks

Both ecological and evolutionary timescales are of importance when considering an ecological system; population dynamics affect the evolution of species traits, and vice versa. Recently, these two timescales have been used to explain structural patterns in host-parasite networks, where the evolution of the manner in which species balance the use of their resources in interactions with each other was examined. One of these patterns was nestedness, in which the set of parasite species within a particular host forms a subset of those within a more species-rich host. Patterns of both nestedness and anti-nestedness have been observed significantly more often than expected due to chance in host-parasite networks. In contrast, mutualistic networks tend to display a significant degree of nestedness, but are rarely anti-nested. Within networks with different interaction types, therefore, there appears to be a feature promoting non-random structural patterns, such as nestedness and anti-nestedness, depending on the interaction types involved. Here, we invoke the co-evolution of species trait-values when allocating resources to interactions to explain the structural pattern of nestedness in a mutualistic community. We look at a bipartite, multi-species system, in which the strength of an interaction between two species is determined by the resources that each species invests in that relationship. We then analyze the evolution of these interactions using adaptive dynamics. We found that the evolution of these interactions, reflecting the trade-off of resources, could be used to accurately predict that nestedness occurs significantly more often than expect due to chance alone in a mutualistic network. This complements previous results applying the same concept to an antagonistic network. We conclude that population dynamics and resource trade-offs could be important promoters of structural patterns in ecological networks of different types.     

Trait-mediated interaction leads to structural emergence in mutualistic networks

As asymmetric structures of mutualistic networks can potentially contribute to system resilience, elucidating drivers behind the emergence of particular network architectures remains a major endeavour in ecology. Here, using an eco-evolutionary model for bipartite mutualistic networks with trait-mediated interactions, we explore how particular levels of connectance, nestedness and modularity are affected by three network assembly forces: resource accessibility, tolerance to trait difference between mutualistic pairs and competition intensity. We found that a moderate accessibility to intra-trophic resources and cross-trophic mutualistic support can result in a highly nested web, while low tolerance to trait difference between interacting pairs leads to a high level of modularity. Network-level trait complementarity leads to low connectance and high modularity, while network-level specialization can result in nested structures. Consequently, we argue that the interplay of ecological and evolutionary processes through trait-mediated interactions can explain these widely observed architectures in mutualistic networks.     

A commensal network of epiphytic orchids and host trees in an Atlantic Forest remnant: A case study revealing the important role of large trees in the network structure

Application of metrics derived from network theory could elucidate the structural organization of orchid assemblages, and help identify the host tree species on which they depend, as well as predicting the impacts of removing host tree species. In this study, we used nestedness, modularity, connectance and robustness, to identify the factors that predict the structure of a quantitative orchid–host tree network in a remnant of Atlantic Forest in Brazil. The network exhibited low nestedness (NODF = 14.07; P = 0.03; WNODF = 5.3; P = 0.02) and no modularity. It was highly robust to the random elimination of host tree species, but showed low robustness when host trees with more interactions started to be eliminated. The nested pattern found was attributed to a combination of two main host tree traits, height and diameter (given by DBH measure). Thus, our analyses reflect the importance of tall and large host tree species, which seem to play an important role in the network structure, providing a substrate for orchid species with different habitat associations and representing a decisive factor in both nested and robust patterns found.     

Relative impacts of environmental variation and evolutionary history on the nestedness and modularity of tree–herbivore networks

Nestedness and modularity are measures of ecological networks whose causative effects are little understood. We analyzed antagonistic plant–herbivore bipartite networks using common gardens in two contrasting environments comprised of aspen trees with differing evolutionary histories of defence against herbivores. These networks were tightly connected owing to a high level of specialization of arthropod herbivores that spend a large proportion of the life cycle on aspen. The gardens were separated by ten degrees of latitude with resultant differences in abiotic conditions. We evaluated network metrics and reported similar connectance between gardens but greater numbers of links per species in the northern common garden. Interaction matrices revealed clear nestedness, indicating subsetting of the bipartite interactions into specialist divisions, in both the environmental and evolutionary aspen groups, although nestedness values were only significant in the northern garden. Variation in plant vulnerability, measured as the frequency of herbivore specialization in the aspen population, was significantly partitioned by environment (common garden) but not by evolutionary origin of the aspens. Significant values of modularity were observed in all network matrices. Trait-matching indicated that growth traits, leaf morphology, and phenolic metabolites affected modular structure in both the garden and evolutionary groups, whereas extra-floral nectaries had little influence. Further examination of module configuration revealed that plant vulnerability explained considerable variance in web structure. The contrasting conditions between the two gardens resulted in bottom-up effects of the environment, which most strongly influenced the overall network architecture, however, the aspen groups with dissimilar evolutionary history also showed contrasting degrees of nestedness and modularity. Our research therefore shows that, while evolution does affect the structure of aspen–herbivore bipartite networks, the role of environmental variations is a dominant constraint.     

Contrasting structures of plant–mite networks compounded by phytophagous and predatory mite species

Differences in the feeding habits between phytophagous and predatory species can determine distinct ecological interactions between mites and their host plants. Herein, plant–mite networks were constructed using available literature on plant-dwelling mites from Brazilian natural vegetation in order to contrast phytophagous and predatory mite networks. The structural patterns of plant–mite networks were described through network specialization (connectance) and modularity. A total of 187 mite species, 65 host plant species and 646 interactions were recorded in 14 plant–mite networks. Phytophagous networks included 96 mite species, 61 host plants and 277 interactions, whereas predatory networks contained 91 mite species, 54 host plants and 369 interactions. No differences in the species richness of mites and host plants were observed between phytophagous and predatory networks. However, plant–mite networks composed of phytophagous mites showed lower connectance and higher modularity when compared to the predatory mite networks. The present results corroborate the hypothesis that trophic networks are more specialized than commensalistic networks, given that the phytophagous species must deal with plant defenses, in contrast to predatory mites which only inhabit and forage for resources on plants.     

Species traits and abundance influence the organization of liana–tree antagonistic interaction

The interaction among species can be influenced by neutral processes, in which more abundant species have high effect on the structure of interaction, or can be influenced by trait matching. Despite both variables (abundance and species traits) influencing the interaction of species in mutualistic networks, few studies showed their importance in antagonistic networks. Here, we posed the question: what are the main predictors of the liana–tree interactions: species abundance, biological traits or both? In a savanna woodland fragment in south‐eastern Brazil, we sampled lianas and trees in 1 ha, where we recorded the abundance, maximum height and bark roughness of tree species, as well as abundance, maximum diameter and climbing system of liana species. For each species, we calculated their contribution to nestedness (n_i), which is a measure of network structure, and performed simple linear regressions between n_i and abundance and species traits. Abundant species contribute more to n_i than rare species, indicating that neutral processes affect interactions between lianas and trees, probably because lianas are opportunistic and climb trees in their neighbourhood. The only trait related to n_i was tree height, which can indicate that light availability can have a considerable role on network structure between both growth forms. Therefore, the importance of species abundance and tree height can be related to opportunism of lianas on climbing the most suitable tree in their neighbourhood.     

Hidden crown jewels: the role of tree crowns for bryophyte and lichen species richness in sycamore maple wooded pastures

Tree crowns typically cover the vast majority of the surface area of trees, but they are rarely considered in diversity surveys of epiphytic bryophytes and lichens, especially in temperate Europe. Usually only stems are sampled. We assessed the number of bryophyte and lichen species on stems and in crowns of 80 solitary sycamore maple trees (Acer pseudoplatanus) at six sites in wooded pastures in the northern Alps. The total number of species detected per tree ranged from 13 to 60 for bryophytes, from 25 to 67 for lichens, and from 42 to 104 for bryophytes and lichens considered together. At the tree level, 29 % of bryophyte and 61 % of lichen species were recorded only in the crown. Considering all sampled trees together, only 4 % of bryophyte, compared to 34 % of lichen species, were never recorded on the stem. Five out of 10 red-listed bryophyte species and 29 out of 39 red-listed lichen species were more frequent in crowns. The species richness detected per tree was unexpectedly high, whereas the proportion of exclusive crown species was similar to studies from forest trees. For bryophytes, in contrast to lichens, sampling several stems can give a good estimation of the species present at a site. However, frequency estimates may be highly biased for lichens and bryophytes if crowns are not considered. Our study demonstrates that tree crowns need to be considered in research on these taxa, especially in biodiversity surveys and in conservation tasks involving lichens and to a lesser degree also bryophytes.     

Vertical distribution of epiphytic bryophytes and lichens emphasizes the importance of old beeches in conservation

During storms in 2005, a number of beech trees fell over at Biskopstorp, SW Sweden, offering the opportunity to study epiphytes along entire stems. In total 16 beech trees in four beech stands representing three different age classes were included. For each tree, 2 m segments from the base to the top were surveyed. In total 115 species were found (76 lichens, 39 bryophytes), of which 30 were considered to be of conservation concern (22 lichens, 8 bryophytes). For lichens significantly more species were recorded above 2 m in height, whereas more bryophytes were recorded below 2 m in height. Certain red-listed lichens were recorded only above 2 m in height on old trees. In a second data set from the same area 140 age-determined beech trees were surveyed for species of conservation concern at the heights 0–2 and 2–5 m, respectively. These species were found almost exclusively on old beech trees, and presence at 2–5 m was recorded, with one exception, only on those trees which also had species of conservation concern at 0–2 m. Records of these species correlated significantly to microhabitat variables, i.e. the presence of rough bark and moss cover higher up the stems on the old trees. This study indicates that surveying only the base in really old beech forests can underestimate both the number of species of conservation concern and their population sizes. However, surveys restricted to the base in rather even-aged beech stands catch a large proportion of the trees with species of conservation concern.     

Diurnal foraging ant–tree co-occurrence networks are similar between canopy and understorey in a Neotropical rain forest

Discussion of the vertical stratification of organisms in tropical forests has traditionally focused on species distribution. Most studies have shown that, due to differences in abiotic conditions and resource distribution, species can be distributed along the vertical gradient according to their ecophysiological needs. However, the network structure between distinct vertical strata remains little-explored. To fill this gap in knowledge, we used baits to sample ants in the canopy and understorey trees of a Mexican tropical rain forest to record the ant–tree co-occurrences. We examined the ant–tree co-occurrences in the canopy and understorey using complementary network metrics (i.e., specialization, interaction diversity, modularity, and nestedness). In addition, we evaluated co-occurrence patterns between ant species on trees, using C-score analysis. In general, we found no differences in the network structure, although the interaction diversity was greater in the understorey than in the canopy networks. We also observed that co-occurrence networks of each vertical stratum featured four ant species in the central core of highly co-occurring species, with three species unique to each stratum. Moreover, we found a similar trend toward ant species segregation in the both strata. These findings reveal a similar pattern of ant–ant co-occurrences in both vertical strata, probably due to the presence of arboreal-nesting ants in the understorey. Overall, we showed that despite the marked differences in species composition and environmental conditions between understorey and canopy strata, ant–tree co-occurrences in these habitats could be governed by similar mechanisms, related to dominance and resource monopolization by ants.     

A neutral‐niche theory of nestedness in mutualistic networks

Recently, there has been a vigorous interest in community ecology about the structure of mutualistic networks and its importance for species persistence and coevolution. However, the mechanisms shaping mutualistic networks have been rarely explored. Here we extend for the first time the neutral theory of biodiversity to a multi trophic system. We focus on nestedness, a distinctive pattern of mutualistic community assembly showing two characteristics, namely, asymmetrical specialization (specialists interacting with generalists) and a generalist core (generalists interacting with generalists). We investigate the importance of relative species abundance (RSA) for the nested assembly of plant–animal mutualistic networks. Our results show that neutral mutualistic communities give rise to networks considerably more nested than real communities. RSA explains 60–70% of nested patterns in two real communities studied here, while 30–40% of nestedness is still unexplained. The nested pattern in real communities is better explained when we introduce interaction‐specific species traits such as forbidden links and intensity of dependence (relative importance of fruits for the diet of a frugivore) in our analysis. The fact that neutral mutualistic communities exhibit a perfectly nested structure and do not show a random or compartmentalized structure, underlines the importance of RSA in the assembly of mutualistic networks.     

High richness of ectomycorrhizal fungi and low host specificity in a coastal sand dune ecosystem revealed by network analysis

Ectomycorrhizal (EM) fungi are ubiquitous in temperate and boreal forests, comprising over 20,000 species forming root symbiotic associations with Pinaceae and woody angiosperms. As much as 100 different EM fungal species can coexist and interact with the same tree species, forming complex multispecies networks in soils. The degree of host specificity and structural properties of these interaction networks (e.g., nestedness and modularity) may influence plant and fungal community assembly and species coexistence, yet their structure has been little studied in northern coniferous forests, where trees depend on EM fungi for nutrient acquisition. We used high‐throughput sequencing to characterize the composition and diversity of bulk soil and root‐associated fungal communities in four co‐occurring Pinaceae in a relic foredune plain located at Îles de la Madeleine, Québec, Canada. We found high EM fungal richness across the four hosts, with a total of 200 EM operational taxonomic units (OTUs), mainly belonging to the Agaricomycetes. Network analysis revealed an antinested pattern in both bulk soil and roots EM fungal communities. However, there was no detectable modularity (i.e., subgroups of interacting species) in the interaction networks, indicating a low level of specificity in these EM associations. In addition, there were no differences in EM fungal OTU richness or community structure among the four tree species. Limited shared resources and competitive exclusion typically restrict the number of taxa coexisting within the same niche. As such, our finding of high EM fungal richness and low host specificity highlights the need for further studies to determine the mechanisms enabling such a large number of EM fungal species to coexist locally on the same hosts.     

A waterfowl seed-dispersal network from the Neotropical region is nested and modular

Seed dispersal by vertebrates is fundamental for the persistence of plant species, forming networks of interactions that are often nested and modular. Networks involving angiosperms and frugivorous birds are relatively well-studied in the Neotropical region, but there are no previous studies of networks involving waterbirds. Here, we describe the structure of a Neotropical waterfowl seed-dispersal network and identify the species that have an important role for the network structure. We used information on 40 plant taxa found in fecal samples of five common waterfowl species to calculate the nestedness (NODF), weighted nestedness (WNODF), modularity, and weighted modularity of the network. We found that the network was nested, with yellow-billed teal showing the highest contribution both to nestedness and weighted nestedness. Twenty-four plant species contributed positively to weighted nestedness, with Salzmann's mille graines presenting the highest influence both to nestedness and weighted nestedness. The network was modular, but the weighted modularity was not significant. These results need to be considered with caution due to incomplete interaction sampling for two species. Ringed teal, Brazilian teal, and yellow-billed teal were considered hub modular species. Among plants, beak sedges and water snowflake were considered modular hub species, while Salzmann's mille graines and spikerush were network connectors. The structure of this Neotropical waterbird seed-dispersal network differed from the only previous waterfowl network study, from Europe, which found similar level of nestedness but no significant modularity. We include several possible explanations for this discrepancy and identified priorities for future research into waterbird–plant interaction networks. Abstract in Portuguese is available with online material.     

Temporal variation in bat-fruit interactions: Foraging strategies influence network structure over time

Mutualistic interactions, such as seed dispersal, are important for the maintenance of structure and stability of tropical communities. However, there is a lack of information about spatial and temporal variation in plant-animal interaction networks. Thus, our goal was to assess the effect of bat's foraging strategies on temporal variation in the structure and robustness of bat-fruit networks in both a dry and a rain tropical forest. We evaluated monthly variation in bat-fruit networks by using seven structure metrics: network size, average path length, nestedness, modularity, complementary specialization, normalized degree and betweenness centrality. Seed dispersal networks showed variations in size, species composition and modularity; did not present nested structures and their complementary specialization was high compared to other studies. Both networks presented short path lengths, and a constantly high robustness, despite their monthly variations. Sedentary bat species were recorded during all the study periods and occupied more central positions than nomadic species. We conclude that foraging strategies are important structuring factors that affect the dynamic of networks by determining the functional roles of frugivorous bats over time; thus sedentary bats are more important than nomadic species for the maintenance of the network structure, and their conservation is a must.     

Patterns of interactions of a large fish-parasite network in a tropical floodplain

1.?Describing and explaining the structure of species interaction networks is of paramount importance for community ecology. Yet much has to be learned about the mechanisms responsible for major patterns, such as nestedness and modularity in different kinds of systems, of which large and diverse networks are a still underrepresented and scarcely studied fraction. 2.?We assembled information on fishes and their parasites living in a large floodplain of key ecological importance for freshwater ecosystems in the Paraná River basin in South America. The resulting fish-parasite network containing 72 and 324 species of fishes and parasites, respectively, was analysed to investigate the patterns of nestedness and modularity as related to fish and parasite features. 3.?Nestedness was found in the entire network and among endoparasites, multiple-host life cycle parasites and native hosts, but not in networks of ectoparasites, single-host life cycle parasites and non-native fishes. All networks were significantly modular. Taxonomy was the major host's attribute influencing both nestedness and modularity: more closely related host species tended to be associated with more nested parasite compositions and had greater chance of belonging to the same network module. Nevertheless, host abundance had a positive relationship with nestedness when only native host species pairs of the same network module were considered for analysis. 4.?These results highlight the importance of evolutionary history of hosts in linking patterns of nestedness and formation of modules in the network. They also show that functional attributes of parasites (i.e. parasitism mode and life cycle) and origin of host populations (i.e. natives versus non-natives) are crucial to define the relative contribution of these two network properties and their dependence on other ecological factors (e.g. host abundance), with potential implications for community dynamics and stability.     

Network properties of arboreal plants: Are epiphytes,mistletoes and lianas structured similarly?

Network analyses provide a unified framework to evaluate different types of species interactions. We used a network approach to comparatively evaluate three types of arboreal plant metacommunities. Interactions between mistletoes, lianas and epiphytes and their host trees were quantified in two New Zealand forests and individual-based null models were used to test for non-random patterns in network degree, nestedness and negative co-occurrence patterns. Arboreal plants were simulated to randomly occur on individual host trees to derive ‘null’ interaction matrices, which were then compared to the observed matrix. Results showed that mistletoes, lianas and epiphytes had very different network properties. Mistletoe and liana degree distributions exhibited fewer links than expected under the null model, indicating strong host preferences. Conversely, degree distributions for epiphytes were consistent with randomised expectations. Mistletoes and lianas were less nested than null model expectations and instead showed support for negative co-occurrence patterns, meaning mistletoe and liana species tended to have mutually exclusive host preferences. Conversely, epiphytes were more nested than expected by chance and showed positive co-occurrence patterns. Overall results indicate that plant–plant interactions exhibited by different types of arboreal plants have very different network properties. We hypothesize that these differences result from (1) differences in coevolutionary dynamics between arboreal plants and their hosts, which range from parasitic (mistletoes) to commensal (epiphytes), and (2) biotic interactions among arboreal plant species for access to host trees.