1. Network theory allows us to understand complex systems by evaluating how their constituent elements interact with one another. Such networks are built from matrices that describe the effect of each element on all others. Quantifying the strength of these interactions from empirical data can be difficult, however, because the number of potential interactions increases non-linearly as more elements are included in the system, and not all interactions may be empirically observable when some elements are rare.

2. We present a novel modelling framework that uses measures of species performance in the presence of varying densities of their potential interaction partners to estimate the strength of pairwise interactions in diverse horizontal systems.

3. Our method allows us to directly estimate pairwise effects when they are statistically identifiable and to approximate pairwise effects when they would otherwise be statistically unidentifiable. The resulting interaction matrices can include positive and negative effects, the effect of a species on itself, and allows for non-symmetrical interactions.

4. We show how to link the parameters inferred by our framework to a population dynamics model to make inferences about the effect of interactions on community dynamics and diversity.

5. The advantages of these features are illustrated with a case study on an annual wildflower community of 22 focal and 52 neighbouring species, and a discussion of potential applications of this framework extending well beyond plant community ecology.

This dataset comes from an annual wildflower community in Western Australia. This system is a diverse and well-studied community of annual plants which germinate, grow, set seed and die within approximately 4 months every year. Individual fecundity data were collected in 2016 when 100 50 x 50 cm plots established in the understory of West Perenjori Reserve were monitored over the length of the full field season. The resulting dataset includes between 29 to over 1,000 counts of individual plant seed production from 22 different focal species (with a median of 108 observations per species), in addition to the identity and densities of all neighbouring individuals within the interaction neighbourhood of each focal plant. Interaction neighbourhoods varied in radius from 3 to 5 cm depending on the size of the focal species. Total neighbourhood diversity was 71 wildflower species, 19 of which were recorded fewer than 10 times across the whole dataset. The species-specific effects of this latter group of species on focals were deemed negligible due to their extremely low density; they were thus grouped into an 'other' category and their effects on focals averaged. This resulted in 53 potential neighbour identities. Plots were randomly positioned across the reserve to capture natural environmental variation in the system. Major factors known to impact plant composition and within species abundances are soil P concentrations, shade, and the presence of woody debris. Natural levels of soil P are very low in this system with elevated P concentrations occurring predominantly on the edge of the reserve near agricultural fields that use P fertiliser. All study plots were placed far from reserve edges to avoid artificially elevated levels of P, which are easy to detect due to an association between high P levels and extensive exotic grass abundances, which are largely absent from the core reserve where this study was conducted. Thus, environmental variation captured by this dataset reflects natural variation in tree cover, soil nutrients and woody debris. To account for confounding effects between plot location and plant density, half of all plots were thinned (a quarter to 60% abundance and a quarter to 30%). Thinning did not target any particular species. This dataset also contains seed germination rates and seed survival rates for the focal species collected from a set of field experiments conducted by T. Martyn, M. Raymundo and I. Towers at Perenjori reserve between 2015 and 2019.

Further details on the collection of this data are available in the Methods sections of the associated manuscript, 'Estimating interaction strengths for diverse horizontal systems using performance data' (Bimler et al. 2023, Methods in Ecology and Evolution).

Files are intended to be opened and used with R. Associated scripts and functions are available at https://github.com/malbion/JointModelFramework/.