Effects of fasting on collective movement and fission–fusion dynamics in both homogeneous and heterogeneous shoals of a group‐living cyprinid fish species

Abstract

The present study aimed to reveal the effect of fasting (21 days) on collective movement and interaction dynamics in both homogeneous (eight members fed a commercial diet or deprived of food) and heterogeneous (four fed + four starved members) shoals of juvenile qingbo (Spinibarbus sinensis). The authors of this study measured the shoaling behaviour in both a commonly used rectangular open arena with no spatial complexity and a radial arm maze. When measured in the open arena, the starved shoals had a faster swimming speed and acceleration rate and a longer interindividual distance than the fed shoals, possibly because of the elevated foraging motivation. Nonetheless, the values of the heterogeneous groups were similar to those of the fed groups. Furthermore, in contrast to the fish in homogeneous shoals, the starved fish in heterogeneous shoals showed a slower acceleration rate and speed than fed members in heterogeneous shoals. These results, combined with the relationships of variables at the among‐ and within‐shoal levels, suggested that starved fish limited their motion in heterogeneous shoals to maintain group cohesion but that the fed fish contributed more to maintaining shoal structure, possibly because of the higher energy expenditure required for movement changes. When monitored in a radial arm maze, starved shoals showed more fission–fusion episodes without sacrificing group cohesion, as they adaptively adjusted the frequency and duration of each majority choice. The among‐shoal variation revealed that the heterogeneous groups showed less variation in the open arena but more variation in the radius maze than did the homogeneous groups. This difference might arise because dominant members have opposite effects on shoal behaviour and consensus decisions. In conclusion, the present study showed opposite effects of feeding states on collective behaviour between homogeneous and heterogeneous shoals, possibly because of the complicated interactions among members with different energy storage levels and foraging motivations. Furthermore, the heterogeneous groups showed a difference between shoal behaviour in the open area and exploration in the radial arm maze. Future studies manipulating the personality composition of starved and fed members of heterogeneous groups might yield interesting results.

Fasting and digestion exert opposite effects on collective behavior in goldfish across contexts

“…Goldfish in the digestion groups swam slower than those in either the fasted or control groups, with the two mixed heterogeneous groups swimming at an intermediate level between the rates of those groups. Similarly, previous studies also found that hungry fish swim faster than relatively satiated fish within or among groups (Hansen et al, 2015a(Hansen et al, , 2015bKrause et al, 1998;Wilson et al, 2019;Zheng & Fu, 2021). These findings suggest that the internal nutritional state is one of the key drivers of the general activity levels of individual fish and hence results in individuals with different nutritional states showing specific swimming speed patterns.…”

“…For example, the whole group can break into subgroups according to different individual needs when the whole group is composed of a majority of hungry fish (Hansen et al, 2015a). However, the frequency and duration of decision making of the heterogeneous groups were not different from those of the homogeneous groups in Qingbo (Spinibarbus sinensis) (Zheng & Fu, 2021). Our study did not measure or compare the frequency and duration of group fission-fusion between the heterogeneous and homogeneous groups, but increasing either interindividual or nearest neighbor distance in the fasting group provides more choices for group fission-fusion than in the control or digestion groups.…”

“…For example, animal groups become more cohesive when facing predation risk but more scattered and less coherent when foraging (Hoare et al, 2004;Schaerf et al, 2017); additionally, hungry individuals tend to occupy anterior positions to increase their food capture probability (Krause et al, 1992(Krause et al, , 1998. Previous studies on collective motion and behavior have used a single and stable context (Herbert-Read, 2016;Zheng & Fu, 2021), but an increasing number of recent studies have used different ecological contexts to examine how animal groups change considerably in response to these contexts (Jolles et al, 2017(Jolles et al, , 2018Schaerf et al, 2017;Yang et al, 2021). However, little is still known about whether animal groups composed of individuals with different nutritional statuses maintain repeatable collective behavioral differences between groups across different ecological contexts.…”

Hiding in the bush, in fear of a predator! Vegetation and predators influence shoaling among wild zebrafish

“…Ecological factors like habitat, prey availability and predation strongly control fission-fusion dynamics among schooling fishes [43][44][45]. As in guppies, the analyses of fission-fusion dynamics revealed that in the presence of predator cues, the tendency of individuals to leave the largest subgroup declined significantly [46].…”

Density-dependent attributes of schooling in small pelagic fishes

“…Therefore, the models were defined as follows: where y i,j , is the observed value of a school characteristic of species i at station j in school k, α i is a shape parameter (>0), β i,j is a rate parameter (>0), μ i,j is the expected (mean) value of the school characteristic, x i,j is local population density, a i is a coefficient, b i is a power exponent, and r l is a random effect of date l . Factors other than population density have also been reported to affect schooling behaviour (Borner et al ., 2015; Romenskyy et al ., 2020; Zheng & Fu, 2021), and these factors may cause spatiotemporal autocorrelation in school characteristics. To consider this spatiotemporal autocorrelation, survey date was incorporated into the model as a multiplicative random effect following a normal distribution with mean 1 and standard deviation .…”