Buzhdygan, Oksana Y; Meyer, Sebastian Tobias; Weisser, Wolfgang W; Eisenhauer, Nico; Ebeling, Anne; Borrett, Stuart R; Buchmann, Nina; Cortois, Roeland; De Deyn, Gerlinde B; de Kroon, Hans; Gleixner, Gerd; Hertzog, Lionel R; Hines, Jes; Lange, Markus; Mommer, Liesje; Ravenek, Janneke; Scherber, Christoph; Scherer-Lorenzen, Michael; Scheu, Stefan; Schmid, Bernhard; Steinauer, Katja; Strecker, Tanja; Tietjen, Britta; Vogel, Anja; Weigelt, Alexandra; Petermann, Jana S (2020): Multitrophic energy dynamics (energy-use efficiency, energy flow, and energy storage) in the Jena Experiment (Main Experiment) [dataset]. PANGAEA, https://doi.org/10.1594/PANGAEA.910659,

---

Supplement to: Buzhdygan, OY et al. (2020): Biodiversity increases multitrophic energy use efficiency, flow and storage in grasslands. Nature Ecology & Evolution, https://doi.org/10.1038/s41559-020-1123-8

This data set contains measures of energy-use efficiency, energy flow, and energy storage in units of dry biomass that quantify the multitrophic ecosystem functioning realized in grassland ecosystems of differing plant diversity. Given are both the measures integrated over whole ecosystems (total network measures) as well as the energy dynamics associated with individual ecosystem compartments including the entire biological community and detrital compartments across the above- and belowground parts of the ecosystem.

Data presented here is from the Main Experiment plots of a large grassland biodiversity experiment (the Jena Experiment, see further details below). In the main experiment, 82 grassland plots of 20 x 20 m were established from a pool of 60 species belonging to four functional groups (grasses, legumes, tall and small herbs). In May 2002, varying numbers of plant species from this species pool were sown into the plots to create a gradient of plant species richness (1, 2, 4, 8, 16 and 60 species) and functional richness (1, 2, 3, 4 functional groups). Study plots are grouped in four blocks in parallel to the river in order to account for any effect of a gradient in abiotic soil properties. Each block contains an equal number of plots of each plant species richness and plant functional group richness level. Plots were maintained in general by bi-annual weeding and mowing. Since 2010, plot size was reduced to 5.5 x 6 m and plots were weeded three times per year.

Trophic-network models were constructed for 80 of the experimental plots, and represent the ecosystem energy budget in the currency of dry-mass (g m-2 for standing stocks and g m-2 d-1 for flows). All trophic networks have the same topology, but they differ in the estimated size of the standing stock biomass of individual compartments (g m-2) and flows among the compartments (g m-2 d-1). Each trophic network contains twelve ecosystem compartments representing distinct trophic groups of the above- and belowground parts of the ecosystem (i.e., plants, soil microbial community, and above- and belowground herbivores, carnivores, omnivores, decomposers, all represented by invertebrate macro- and mesofauna) and detrital pools (i.e., surface litter and soil organic matter). Vertebrates were not considered in our study due to limitations of data availability and because the impact of resident vertebrates in our experimental system is expected to be minimal. Larger grazing vertebrates were excluded by a fence around the field site, though there was some occasional grazing by voles.

Compartments are connected by 41 flows. Flows (fluxes) constitute 30 internal flows within the system, namely feeding (herbivory, predation, decomposition), excretion, mortality, and mechanical transformation of surface litter due to bioturbation plus eleven 11 external flows, i.e. one input (flows entering the system, namely carbon uptake by plants) and ten output flows (flows leaving the system, namely respiration losses). The ecosystem inflow (a flow entering the system) and outflows (flows leaving the system) represent carbon uptake and respiration losses, respectively. In the case of consumer groups, the food consumed (compartment-wide input flow) is further split into excretion (not assimilated organic material that is returned to detrital pools in the form of fecesfaeces) and assimilated organic material, which is further split into respiration (energy lost out of the system to the environment) and biomass production, which is further consumed by higher trophic levels due to predation or returned to detrital pools in the form of mortality (natural mortality or prey residues). In case of detrital pools (i.e. surface litter and soil organic matter), the input flows are in the form of excretion and mortality from the biota compartments, and output flows are in the form of feeding by decomposers and soil microorganisms (i.e. decomposition). Surface litter and soil organic matter are connected by flows in the form of burrowing (mechanical transportation) of organic material from the surface to the soil by soil fauna. Organism immigration and emigration are not considered in our study due to limited data availability.

Flows were quantified using resource processing rates (i.e. the feeding rates at which material is taken from a source) multiplied with the standing biomass of the respective source compartment. To approximate resource processing rates, different approaches were used: (i) experimental measurements (namely the aboveground decomposition, fauna burial activity (bioturbation), microbial respiration, and aboveground herbivory and predation rates); (ii) allometric equations scaled by individual body mass, environmental temperature and phylogenetic group (for the above- and belowground fauna respiration rates and plant respiration); (iii) assimilation rates scaled by diet type (for quantification of belowground fauna excretion and natural mortality); (iv) literature-based rates scaled by biomass of trophic groups (for microbial mortality); and (v) mass-balance assumptions (carbon uptake, plant and aboveground fauna mortality, belowground decomposition, belowground herbivory, and belowground predation). Mass-balance assumption means that the flows are calculated assuming that resource inputs into the compartment (i.e. feeding) balance the rate at which material is lost (i.e. the sum of through excretion, respiration, predation, and natural death). We used constrained nonlinear multivariable optimization to perturb the initial flow rates estimated from the various sources. We assigned confidence ratings for each flow rate, reflecting the quality of empirical data it is based on. We then used the 'fmincon' function from Matlab's optimization toolbox, which utilizes the standard Moore-Penrose pseudoinverse approach to achieve a balanced steady state ecological network model that best reflects the collected field data. Measured data used to parameterize the trophic network models were collected mostly in the year 2010.

Network-wide measures that quantify proxies for different aspects of multitrophic ecosystem functioning were calculated for each experimental plot using the 'enaR' package in R. In particular, total energy flow was measured as the sum of all flows through each ecosystem compartment. Flow uniformity was calculated as the ratio of the mean of summed flows through each individual ecosystem compartment divided by the standard deviation of these means. Total-network standing biomass was determined as the sum of standing biomass across all ecosystem compartments. Community maintenance costs were calculated as the ratio of community-wide respiration related to community-wide biomass.

Most of the data used to parameterize these trophic networks were collected in 2010.

A diagram depicting the conceptual trophic-network model developed to describe multitrophic ecosystem functioning can be found in the paper (REF to the NEE paper). This paper also shows the relationship between the individual flows and compartment sizes as well as the network-wide measures with plant species richness. Further sensitivity analyses for the influence of including the highest diversity level are also provided in the paper.