Comparing bark thickness: testing methods with bark–stem data from two South African fire‐prone biomes. (12th March 2014)

• Record Type: Journal Article
• Authors: Hempson, Gareth P.; Midgley, Jeremy J.; Lawes, Michael J.; Vickers, Karen J.; Kruger, Laurence M.; Pausas, Juli
• Is Part Of: Journal of vegetation science. Volume 25:Number 5(2014)
• Page Start: 1247
• Page End: 1256
• Publication Date: 2014-03-12
• Subjects: Plant ecology -- Periodicals; Plant communities -- Periodicals; Plant populations -- Periodicals; 581.7
• DOI: 10.1111/jvs.12171
• Languages: English
• ISSNs: 1100-9233

Abstract

<abstract abstract-type="main" id="jvs12171-abs-0001"> <title>Abstract</title> <sec id="jvs12171-sec-0001" sec-type="section"> <title>Aims</title> <p>Bark thickness–stem diameter relationships are non‐linear above a stem diameter threshold in many woody species, which makes relative bark thickness measures dependent on the range of stem diameters that are sampled. This influences the appropriateness of different methods for comparing fire responses of woody plants across studies. Here we develop a framework for bark thickness comparisons by evaluating relative bark thickness estimates and bark thickness predictions obtained from linear and curved models fitted to raw and log‐transformed bark–stem data. We use this framework to contrast bark thickness among fynbos and savanna plant functional groups.</p> </sec> <sec id="jvs12171-sec-0002" sec-type="section"> <title>Location</title> <p>Fynbos (17 species) and savanna (21 species) systems in South Africa.</p> </sec> <sec id="jvs12171-sec-0003" sec-type="section"> <title>Methods</title> <p>The linear subset of bark–stem data was identified using a three‐step procedure. Linear regressions (with and without an intercept) and curved models (allometric and modified exponential models) were fitted to the linear subset and complete raw bark–stem data set, respectively. In addition, linear regression models were fitted to the log‐transformed complete bark–stem data set. Regression slopes and bark thickness predictions obtained from these different approaches were compared, to determine which method provides the most robust metric for comparing bark thickness. Bark thickness was compared among fynbos resprouter guilds and acacias from low‐ and high‐fire savannas.</p> </sec> <sec id="jvs12171-sec-0004" sec-type="section"> <title>Results</title> <p>The slope of the regression model fitted to the linear subset of raw bark–stem data provides a reliable metric for general comparisons of relative bark thickness. Bark thickness predictions from the curved and log models were comparable at 20‐cm stem diameter, but the log model underestimated bark thickness at 5 cm for certain species. Relative bark thickness was significantly higher in: (1) fynbos fire resisters and epicormic resprouters than in non‐resprouters; and (2) acacias from high‐ vs low‐fire savannas.</p> </sec> <sec id="jvs12171-sec-0005" sec-type="section"> <title>Conclusions</title> <p>The slope of the regression model fitted to the linear subset of raw bark–stem data is a useful metric for bark thickness comparisons across studies, and compares favourably with bark thickness predictions derived from models fitted to the complete bark–stem data set. Fynbos and savanna trends support the proposition that relative bark thickness reflects differences in woody plant responses to fire and indicate the modal fire regime of ecosystems.</p> </sec> </abstract>