Maximum stem diameter predicts liana population demography

Determining population demographic rates is fundamental to understanding differences in species life-history strategies and their capacity to coexist. Calculating demographic rates, however, is challenging and requires long-term, large-scale censuses. Body size may serve as a simple predictor of demographic rate; can it act as a proxy for demographic rate when those data are unavailable? We tested the hypothesis that maximum body size predicts species’ demographic rate using repeated censuses of the 77 most common liana species on the Barro Colorado Island, Panama (BCI) 50-ha plot. We found that maximum stem diameter does predict species’ population turnover and demography. We also found that lianas on BCI can grow to the enormous diameter of 635 mm, indicating that they can store large amounts of carbon and compete intensely with tropical canopy trees. This study is the first to show that maximum stem diameter can predict plant species’ demographic rates and that lianas can attain extremely large diameters. Understanding liana demography is particularly timely because lianas are increasing rapidly in many tropical forests, yet their species-level population dynamics remain chronically understudied. Determining per-species maximum liana diameters in additional forests will enable systematic comparative analyses of liana demography and potential influence across forest types.

Methods
We measured the diameter of all lianas ³ 10 mm diameter in the BCI plot in 2007 and again in 2017 using liana-specific census methods developed by Gerwing et al. (2006) and Schnitzer et al. (2008). Diameter measurements were taken 1.3 m along the stem from the roots using calipers for stems < 50 mm diameter, and diameter tape for stems ³ 50 mm. If the stem was deformed by bulges, nodes, damage, or other defects at the 1.3 m measurement point, we measured 50 mm below the deformity. In total, there were 175 liana species in the BCI 50-ha plot (Schnitzer et al. 2021); we selected the 77 species that had 100 more individuals in the 50- ha plot (Appendix S1: Table S1) to calculate species population dynamics (measured as turnover). We defined species-specific turnover rate as the geometric mean of individual liana recruitment (Lr) and mortality (Lm) between the two censuses (Phillips & Gentry 1994):  (ln Lr + ln Lm)/2.  Both recruitment and mortality were highly correlated with turnover (Pearson’s r = 0.97 & 0.92, respectively); thus, our findings for turnover apply to its component parts (Lr and Lm).