Restored streams recover food web properties but with different scaling relationships when compared to natural streams

Despite extensive studies revealing differences in the composition of aquatic assemblages between restored streams and natural or pre-restoration states, understanding the ecological consequences and trajectories of stream restoration remains challenging. Food webs are an important way of mapping biodiversity to ecosystem functioning by describing feeding linkages and energy transfer pathways. Describing food webs can provide ecological insights into stream restoration. This study analyzed an unprecedentedly large quantity of food web data (> 1,700 webs) based on long-term (2008–2018) biomonitoring data in South Korea using a feeding link extrapolation. By doing so, we aimed to describe general patterns for the reassembly of aquatic food webs in restored streams. Specifically, we analyzed 12 indices related to the food web structure and robustness of restored streams and compared them with those of natural streams. First, the species richness, link numbers, link density, and connecta..., This dataset was generated using data from the National Aquatic Ecological Monitoring Program (NAEMP) in South Korea, which includes species-level and some higher taxonomic group data for fish, benthic macroinvertebrates, and epilithic diatoms collected from 284 stream survey sites. Trophic links between these taxa were extrapolated using the Global Biotic Interactions (GloBI) database to construct food webs for each of the survey sites., , # Data from: Restored streams recover food web properties but with different scaling relationships when compared to natural streams

[https://doi.org/10.5061/dryad.mgqnk998s](https://doi.org/10.5061/dryad.mgqnk998s)

## Description of the data and file structure

This dataset contains food web data provided in the community object format of the R cheddar package. A total of 1,773 food webs are included, representing data from 284 survey locations, constructed annually over the period from 2008 to 2018. Please note that data is not available for all years.

Each food web is stored as a community object in the R cheddar package format and is named sequentially as Web1, Web2, ..., Web1773. Each community object contains three main components:

nodes: Taxonomic information for each observed taxon in the food web.

properties: Metadata including the survey year and stream type (restored or natural).

trophic.links: Consumer-resource interactions in the food web.

## Files and variables

####...