Interactions between epiphytes during canopy soil formation: an experiment in a lower montane cloud forest of southeast Mexico.

In several montane forests around the world, epiphytes coexist in mats, sharing the rhizosphere and forming histosol-type soils rich in nutrients. The role of these epiphytes in the formation of canopy soil and the fitness costs that epiphytes face when cohabiting in these mats are unknown. In a lower montane cloud forest in central Veracruz, Mexico, a 2-year factorial experiment was carried out with the presence/absence of ramets of Phlebodium areolatum (Polypodiaceae), Tillandsia kirchhoffiana, T. multicaulis and T. punctulata (Bromeliaceae). We examined (i) which epiphyte species contribute to the formation of canopy soil, (ii) the role of epiphyte composition in the soil nutrient composition, and (iii) the fitness costs faced by epiphytes when cohabiting. Canopy soil formation highest when P. areolatum is present. Soil nutrient content does not change with epiphyte composition, is influenced by the microbiota, and P content decreases with the presence of epiphytes. The fitness costs show that the species compete, decreasing their survival and growth, but the competitive capacity differs between the species. We conclude that P. areolatum is an ecosystem engineer that promotes the creation of canopy soil but is a poor competitor. The results coincide with the model of succession by facilitation. Canopy soil is a slow-created component whose nutrient content does not depend on the epiphytic flora. In epiphyte mats, the dominant interactions are competitive, but there is also facilitation.

Epiphyte associations and canopy soil volume: nutrient capital and factors influencing soil retention in the canopy.

Canopy soil (CS) volume reflect epiphyte community maturity, but little is known about the factors that retain CS or species succession within it. Humus fern species (e.g. Phlebodium areolatum) appear capable of retaining CS. In ten Quercus spp. we sampled 987 epiphyte mats to examine the role of the common epiphyte species and crown traits determining CS volume, in order to infer successional stages and identify pioneer and late successional species. Branch traits (height, diameter and slope), CS volume and cover of the epiphyte species were determined for each mat. Nutrient content was determined in CS random samples of 12 epiphyte associations and sizes (one sample from each size quintile). A total of 60% of the mats lack CS. Cover of P. areolatum was the main variable explaining CS volume, and this species was present in 46.8% of those with CS. Epiphyte composition was highly variable, but pioneer (species appearing in monospecific mats, without CS) and late successional species could be identified. Canopy soil nutrient content was similar among the associations of epiphytes. Magnesium, Ca and pH decreased with CS volume, while P and N increased. Phlebodium areolatum is associated with high CS volumes and could act as a key species in its retention. Monospecific mats of pioneer species lack CS or have low volumes, while CS is much higher in mats with late successional species, but the mechanisms of CS formation and nutrient retention in response to interactions between epiphyte species remain to be tested.

Orchid seed removal by ants in Neotropical ant-gardens.

Most plants that inhabit ant-gardens (AGs) are cultivated by the ants. Some orchids occur in AGs; however, it is not known whether their seeds are dispersed by AG ants because most orchid seeds are tiny and dispersed by wind. We performed in situ seed removal experiments, in which we simultaneously provided Azteca gnava ants with seeds of three AG orchid species and three other AG epiphyte species (Bromeliaceae, Cactaceae and Gesneriaceae), as well as the non-AG orchid Catasetum integerrimum. The seeds most removed were those of the bromeliad Aechmea tillandsioides and the gesneriad Codonanthe uleana, while seeds of AG orchids Coryanthes picturata, Epidendrum flexuosum and Epidendrum pachyrachis were less removed. The non-AG orchid was not removed. Removal values were positively correlated with the frequency of the AG epiphytes in the AGs, and seeds of AG orchids were larger than those of non-AG orchids, which should favour myrmecochory. Our data show that Azt. gnava ants discriminate and preferentially remove seeds of the AG epiphytes. We report for the first time the removal of AG orchid seeds by AG ants in Neotropical AGs.

Does structural parasitism by epiphytes exist? A case study between Tillandsia recurvata and Parkinsonia praecox.

The effects that epiphytes have on their hosts have been poorly explored in an experimentally. Correlational evidence suggests that epiphytes may be either mutualists or structural parasites, as has been proposed for Tillandsia recurvata on Parkinsonia praecox. To test the effect of T. recurvata upon P. praecox, the epiphyte load on branches of P. praecox was measured and two 1-year experiments were performed to detect the effect of transplantation/removal of epiphytes and shade (0%, 35%. 50% and 80%) on shoot dynamics. If T. recurvata represents a selective pressure for P. praecox, then the frequency of branches carrying large epiphyte loads will be high, and host shoot survival will decrease in the presence of T. recurvata and with increased shade. A weak inverse relationship between epiphyte load and percentage of dead shoots in the host was detected. Shoot survival was independent of epiphyte presence. Shade decreased shoot survival by 35-72%. Results suggest that at the study site, T. recurvata is a commensalist of P. praecox. An alternative hypothesis to explain the correlation between high epiphyte load and branch/tree decay is that older branches carry more epiphytes, receive more shade from neighbouring branches and could be undergoing a natural decline process.