Widespread temporal niche partitioning in an adaptive radiation of cichlid fishes

The partitioning of ecological niches is a fundamental component of species diversification in adaptive radiations. However, it is presently unknown if and how such bursts of organismal diversity are influenced by temporal niche partitioning, wherein species avoid competition by being active during different time windows. Here, we address this question through profiling temporal activity patterns in the exceptionally diverse fauna of cichlid fishes from African Lake Tanganyika. By integrating week-long longitudinal behavioural recordings of over 500 individuals from 60 species with eco-morphological and genomic information, we provide two lines of evidence that temporal niche partitioning facilitated this massive adaptive radiation. First, Tanganyikan cichlids exhibit all known circadian temporal activity patterns (diurnal, nocturnal, crepuscular, and cathemeral) and display substantial inter-specific variation in daily amounts of locomotion. Second, many species with similar habitat an..., , # Widespread temporal niche partitioning in an adaptive radiation of cichlid fishes

Dataset DOI: [10.5061/dryad.j0zpc86sv](10.5061/dryad.j0zpc86sv)

## Description of the data and file structure

These speed and position data were collected from individual fishes of different Lake Tanganyika cichlid species. In short data covers 6 days of data, with a 12h:12h light:dark cycle. Except for the dark:dark experiments (prefix: *dark-dark*), which had a 14h:10h light:dark cycle for the first 3 days, then 3 days of constant darkness. Data was acquired at 10fps. Data has been smoothed and outliers removed.

### Files and variables

For each species, there are two file types: als.csv and meta.csv

There is one als.csv and one meta.csv file for each individual animal. The als.csv data contains the tracks. The file name (FishID) is made in the following format:

```
FISH{date collected}_c{camera number}_r{row number}_{species name - not always the same as in Ronco et al., use Supplemental Data 1...,