Skip to main content
Dryad

Data for: Spatial variability in the contribution of termites to the decay of plant detritus

Cite this dataset

Wijas, Baptiste; Letnic, Mike; Cornwell, Will (2024). Data for: Spatial variability in the contribution of termites to the decay of plant detritus [Dataset]. Dryad. https://doi.org/10.5061/dryad.rr4xgxdh3

Abstract

Drylands are characterized by high spatial variability in resource availability due to sporadic rainfall, topography of the landscape and important effects of animals. Resource availability gradients may trigger patterns in decomposer population abundances and activity which could affect ecosystem functions such as decomposition. Here, we examined the influence of resource availability gradients on the importance of termites in the decomposition of wood and grass litter. We placed wood blocks and grass litter baits in bags accessible and inaccessible to termites across wood and grass resource gradients as determined by the presence or absence of a top mammalian predator and across topographic gradients during a 9-month period in arid Australia. We hypothesized that grass-eating termite activity would track grass abundance and wood-eating termite activity would track wood abundance. Termites were the predominant decomposition agent at these sites. Termites contributed to 99.5% of wood decomposition and 83.9% of grass decomposition during our study period. For wood, the termite effect was spatially variable and increased with habitat wood availability which was greatest on dunes and where top predators were absent. However, the contribution of termites to grass litter decomposition did not track grass availability or termite abundance. The highest effects of termites on grass decomposition rates were found in habitats where the absence of top predators led to low grass availability. Our findings highlight how spatial variability in resources in addition to other factors that we do not document but are known to be influenced by the presence of top predators, such as insectivore predation rates, across the landscape could affect ecosystem functions such as decomposition.  

README

Description of columns for each dataset found in this Dryad repository. The code necessary to run the anlyses and plot the graphs can be found in the file "clean_script_publish.Rmd".

The different datasets descrobe decomposition rates of wood and grass litter over a 9 month period in an Australian arid ecossystem in addition to termite abundances and vegetation surveys.

Dataset: decompt_woodgrass_weight.csv
Columns:

  • tag: individual ID of litter bag
  • substrate: type of litter (wood or grass)
  • transect: dune or swale ID sampled
  • state: side of the fence (NSW = dingo absent, SA = dingo present)
  • set: station number
  • habitat: topographic position
  • treatment: type of litter bag (exclude = inaccessible to termtes, open = accessible to termites)
  • termites: discovred by termites or not (Y = yes, N = no)
  • ternum: numerical termite discovery (1 = discovered by termites, 2 = not discovered by termites)
  • initwweight: initial wet weight of litter in grams
  • initdweight: initial dry weight of litter in grams
  • finalwweight: final wet weight of litter after being harvested in grams
  • finaldweight: final dry weight of litter after being harvested in grams
  • finalpropmassloss: proportion mass loss of litter

Dataset: veg_time_terdingo_final.csv
Columns:

  • transect: dune or swale ID sampled
  • survey: indication of whether the decomposition experiment was run on the dune (y if decomposition was run)
  • state: side of the fence (nsw = dingo absent, sa = dingo present)
  • habitat: topographic position
  • grassdead: percentage cover of dead grass
  • shrubdead: percentage cover of dead wood
  • proptermiteswood: average percentage of wood consumed by termites
  • proptermitesgrass: average percentgae of grass consumed by termites

Empty cells represent sample IDs with no measured variable.

Dataset: terproj_final.csv
Columns:

  • ID: individual ID of sample
  • Transect: dune or swale ID sampled
  • State: side of the fence (NSW = dingo absent, SA = dingo present)
  • Habitat: topographic position
  • woodsamp: number of wood samples with termites under
  • grasssamp: number of grass clumps with termites under
  • hetero: number of occurences of Heterotermes spp.
  • micro: number of occurences of Microcerotermes spp.
  • ami: number of occurences of Amitermes spp.
  • nasut: number of occurences of Nasutitermes spp.

Dataset: grass_cons_ter_final.csv
Columns:

  • transect: dune or swale ID sampled
  • state: side of the fence (NSW = dingo absent, SA = dingo present)
  • set: station number
  • habitat: topographic position
  • termasslossperc: termite consumption of grass in percentages

Dataset: wood_cons_ter_final.csv
Columns:

  • transect: dune or swale ID sampled
  • state: side of the fence (NSW = dingo absent, SA = dingo present)
  • set: station number
  • habitat: topographic position
  • termasslossperc: termite consumption of wood in percentages

Methods

The precise method for how the data was collected can be found in the article presenting the data. 

Funding

Australian Research Council, Award: DP180101477