Spatial and temporal variation in the diversity, abundance, and energetic returns of underground storage organs on the coastal lowlands of the Cape Floristic Region, South Africa

Description

This package contains data and related R script (and README file explaining the organisation of this package) which accompany the research paper with the same title submitted to the Journal of Archaeological Science, with submission ID JASC25-178. This research assessed species with underground storage organs (USOs) on the coastal lowlands of the Cape Floristic Region (CFR) in terms of their richness, abundance, and energetic returns across regions, seasons and biome types. We surveyed 212 sites equally in the growing (spring) and dormant (autumn) seasons and across three biomes (Fynbos, Renosterveld, Strandveld) in each of three regions (west, south, and southeast coasts). In all the surveyed regions, richness, abundance, and energetic returns of edible USOs peaked in winter/spring and were dominated by high-energy cormous species. On an annual basis, energetic returns were highest on the south coast, owing to USO apparency peaking in autumn, a feature absent at the other sites. In all regions, hotspots of USO energetic returns were rare, but most common on the south coast. Energetic returns were mostly higher in Strandveld than in Fynbos and Renosterveld.

Steps to reproduce

We assessed seasonal variation in USO abundance in each of the three regions (west, south, and southeast coasts) by conducting field surveys at the time of lowest (Mar–Apr - autumn) and highest geophyte apparency (Aug–Nov - spring) in three major biomes (Fynbos, Renosterveld, Strandveld) of the Cape Floristic Region coastal lowlands. In each plot, we counted the number of species and the number of individuals per species (abundance) of all noticeable geophytes, both edible and non-edible. In total, we surveyed 212 plots of 5×5 m2, half in spring and half in autumn. Species-specific mean values of biomass and energy content were obtained from a parallel study (Singels et al. 2025 in press). Based on these data, the following variables were calculated per plot: * Richness all – a count of the number of USO species per plot. * Richness edible – a count of the number of edible (Singels et al., in presssubmitted) USO species per plot. * Abundance all – a count of the number of all USO plants per plot. * Abundance edible – a count of the number of edible USO plants per plot. * Biomass – an estimate of the biomass of edible USOs per plot, calculated from plot-level species abundances multiplied by the mean biomass of edible species. The derived values were extrapolated from the plot scale (25 m2) to obtain an estimate of the biomass per hectare (kg.ha-1). * Energy – an estimate of the kilojoules contained in edible USOs in each plot (kJ.25 m-2), calculated from plot-level species abundances multiplied by the mean biomass of edible species, and multiplied by the mean kilojoule content of edible species. The derived values were extrapolated from the plot scale to obtain an estimate of the kilojoules per square meter (kJ.m-2). * Relative return – the energy variable derived above but reduced in relation to relative harvesting and preparation costs pertaining to different USO types (after Singels et al. 2025 in press). Thus, the energy contents of species classified as small corms (which are easiest to harvest, have highest energetic returns and require little or no preparation to render them edible) were reduced by 5%, big corms by 10%, bulbs by 15%, small rhizomes by 20%, big rhizomes by 25%, and tubers by 30%. We stress the relative nature of this reduction factor; we have no direct evidence for discounting the returns by these levels. The response variables did not conform to normality and the experimental design was not perfectly balanced. We thus employed rank-based multiple test procedures and simultaneous confidence intervals for general factorial designs (Brunner et al., 2018; Konietschke et al., 2012) to assess the effects of region, season, biome, and their interactions on the respective response variables pertaining to edible USOs, namely richness, abundance, biomass, energy, and relative return. For this purpose, we used the rankFD package (Konietschke et al., 2022) in the open-source R software version 4.5.1. Details in the README file.

Institutions

• University of Cape Town

  Western Cape, Rondebosch
• Nelson Mandela University

  Port Elizabeth

Categories

Funders

• National Research Foundation

  SG.01, Singapore

  Grant ID: 93487 (NRF of South Africa)
• Palaeontological Scientific Trust

  Gauteng, South Africa
• Nelson Mandela University

  Eastern Cape, South Africa

Licence