A framework for partitioning plant rooting profiles from neighbours using multiple data types. (14th January 2016)

• Record Type: Journal Article
• Authors: Kropp, Heather; Ogle, Kiona; Wojciechowski, Martin F.
• Editors: Kikvidze, Zaal
• Is Part Of: Journal of vegetation science. Volume 27:Number 3(2016:May)
• Page Start: 587
• Page End: 595
• Publication Date: 2016-01-14
• Subjects: Bayesian modelling -- Below‐ground competition -- Molecular identification -- Plant–plant interactions -- Root biomass -- Root ecology -- Stable isotopes -- Vertical root distribution; Plant ecology -- Periodicals; Plant communities -- Periodicals; Plant populations -- Periodicals; 581.7
• DOI: 10.1111/jvs.12377
• Languages: English
• ISSNs: 1100-9233

Abstract

Abstract: Aim: Vertical root distributions ('profiles') influence plant water use and productivity, and the differentiation of root profiles between neighbouring species can indicate the degree of plant interactions and niche partitioning. However, quantifying multiple species' root distributions in the field can be labour intensive and highly destructive to the soil and plants. We describe a method for partitioning multiple species roots using minimally destructive methods to determine if neighbour interactions alter the root profile of a common desert shrub, Larrea tridentata (creosote bush). Location: Sonoran Desert, central Arizona, USA. Methods: We obtained root and soil samples from soil cores collected around Larrea growing alone and next to three different neighbouring species. Bulk root mass was measured for each soil sample, and Larrea and neighbouring species root presence was determined with molecular identification methods. Water extracted from the soil and paired stem samples was analysed for its stable isotope composition (D and 18 O). Species‐specific (i.e. Larrea and neighbouring species) root biomass and fractional active root area were estimated through a hierarchical statistical modelling approach that combined all three data sets and accounted for detection errors. Results: The combined data model successfully partitioned Larrea root biomass from neighbouring plants and provided biologically relevant estimates of rooting profiles with greater certainty than individual analyses of each data source. The data model results indicate that plant neighbours alter Larrea 's root profile; Larrea growing under tree species had significantly higher root biomass in shallow soil layers than Larre a growing alone. Conclusions: Our framework requires minimally destructive sampling methods, and accounts for sampling errors associated with different methods. We demonstrate the utility of our approach with a common desert shrub species, which illustrated that plant neighbours can alter the Larrea vertical root profile. Our approach is useful in problematic study systems fraught with sample collection issues or supporting species with inhibitory compounds that prohibit the use of more sophisticated molecular methods to identify the presence of other species' roots. Abstract : The quantification of multiple species' root distributions can be labor‐intensive and destructive. We describe a modeling framework that combines multiple minimally‐destructive methods to partition species' roots. We examine the effect of plant neighbors on the root profile of a desert shrub, and demonstrate that the combined model has reduced uncertainty compared to individual analyses.