Search

Table S1: Sampling Site Information Table S2: Mean host tree measurements at each elevation and plot data used for biomass calculations. Host trees: dbh (diameter at breast height) and ht¬tot (total tree height) were measured in the field (n=3). Ahost (tree surface area) was calculated using the “Tree A-dbh model”, SE given in parentheses. Plots: No stems = total number of stems (>10 cm dbh) per 1 ha forest plot, Aplot = total tree area per 1 ha plot, calculated using the “Tree A-dbh model”. Table S3: Summary of meteorological data from weather stations located in close proximity to the sampling sites. MAT: Mean Annual Temperature, MAP: Mean annual precipitation, VPD: Vapour pressure deficit, SENHAMI: Servicio Nacional de Meteorologia e Hidrologia (http://www.senhami.gob.pe/), ABERG: Andes Biodiversity and Ecosystem Research Group, Atrium: database from The Botanical research Institute of Texas (http://atrium.andesamazon.org/index), * Data collected at 2750 m Table S4: Number of liverwort and moss species, per genera, recorded across the transect Table S5: Percentages of moss and liverworts species recorded across altitudinal transect Table S6: Most abundant epiphytic bryophytes and all vascular plants recorded on three host trees at each elevation Fig S1. Non-linear regression function (“Tree A-dbh model”), based on allometric calculations and existing plot census data (dbh for all trees >10 cm) used to estimate total surface area in one hectare plot: y =7.96e0.029x Supplementary Information Continued: Data used for Figures 2-5

Bryophytes are unable to control tissue water content although physiological adaptations allow growth in a wide range of habitats. Carbon isotope signals in two mosses (Syntrichia ruralis and Chorisodontium aciphyllum) and two liverworts (Conocephalum conicum and Marchantia polymorpha), whether instantaneous (real-time, Δ13C), or organic matter (as δ13COM), provide an assimilation-weighted summary of bryophyte environmental adaptations. In mosses, δ13COM is within the measured range of Δ13C values, which suggests that other proxies, such as compound specific organic signals will be representative of historical photosynthetic and growth conditions. The liverworts were photosynthetically active over a wider range of relative water contents (RWC) than the mosses. There was a consistent 5‰ offset between Δ13C values in C. conicum and M. polymorpha, suggestive of greater diffusion limitation in the latter. Analysis of a C. aciphyllum moss-peat core showed the isotopic composition over the past 200 years reflects recent anthropogenic CO2 emissions. Once corrected for source-CO2 inputs, the seasonally integrated ∆13COM between 1350 and 2000 AD varied by 1.5‰ compared with potential range of the 12‰ measured experimentally, demonstrating the relatively narrow range of conditions under which the majority of net assimilation takes place. Carbon isotope discrimination also varies spatially, with a 4‰ shift in epiphytic bryophyte organic matter found between lowland Amazonia and upper montane tropical cloud forest in the Peruvian Andes; associated with increased diffusion limitation.