Inherent instability leads to high costs of hovering in near-neutrally buoyant fishes

Abstract

Hovering, the ability to maintain a stationary position in fluid, is essential for many fish species during prey capture, habitat exploration, and mating. While traditionally assumed to be energetically inexpensive for fishes with a swim bladder, the metabolic costs and morphological factors influencing postural stability during hovering remain poorly understood. Hovering requires fishes to counteract small instabilities in position and orientation, often through continuous adjustments using their fins and body. To examine the energetic consequences of this active stabilization, we measured body posture, fin kinematics, and metabolic rates in 13 near-neutrally buoyant fish species during both hovering and resting. Our results show that hovering nearly doubles metabolic rates compared to resting, and species with greater separation between the center of mass and center of buoyancy and increased caudal fin activity exhibit higher energetic costs. In contrast, species with more posteriorly positioned pectoral fins and lower length-to-depth ratios show reduced hovering costs. Our findings demonstrate that, despite morphological traits that promote instability, fishes maintain posture and position through fine-scale fin control-at a significant energetic expense. This study suggests that hovering is a costly behavior that likely plays a key role in shaping the evolution of fish morphology and locomotor strategies.

Keywords: dynamic stability; fish locomotion; hovering; inherent instability.

Fig. 1.

Hovering increases metabolic rates (MO₂) from resting across species (n = 5 to 8). (A) MO_2hover was significantly elevated from MO_2rest for all the species except for sand smelt and three-stripe gourami (asterisks denote statistically a significant difference between MO_2hover and MO_2rest for each species, P < 0.05). (B) Species are divided into two groups based on whether their MO_2net is above or below the average (horizontal dashed line). (C) The same grouping is applied for MO_2hover/MO_2rest.

Fig. 2.

Hovering requires the movement of all fins. (A) 3D fin movements during hovering. (B) Pectoral fin movements varied between in-phase and antiphase. (C) The distance traveled by each fin varied across species, with significant differences in caudal fin movement between the metabolic rate group, G1, or low metabolic group (species: Atherina presbyter, Carassius auratus, Chromis viridis, Dichotomyctere ocellatus, Gasterosteus aculeatus, Poecilia latipinna, and Trichopsis schalleri) and G2, or high metabolic group (species: Dermogenys pusilla, Devario aequipinnatus, Hemigrammus rhodostomus, Kryptopterus vitreolus, Lamprologus ocellatus, and Trigonostigma heteromorpha). Asterisks denote a statistically significant difference between G1 and G2 (P < 0.05). (D) Example of 3D caudal fin movement in a fish from the G1 and G2 groups.

Fig. 3.

Inherent instability vs. dynamic stability during hovering. (A) Near-neutrally buoyant fishes are inherently unstable due to the separation of the center of mass (COM) and the center of buoyancy (COB). (B) Measurements show that 12 out of 13 fishes are inherently unstable on either the dorsal–ventral, anterior–posterior axes, or both, given that the COM and COB are separated. (C) The roll angle across species is maintained low, and most fishes have an average roll angle at or below an average of about 4° (horizontal dashed line). (D) The pitch angle is also low across species, and most fishes show a pitch angle at or below the average of about 2° (horizontal dashed line).

Fig. 4.

Multilinear regression model showing measured and predicted values for MO_2net and MO_2hover/MO_2rest. The model explains 86% of the variation in MO_2net and 63% of the variation in MO_2hover/MO_2rest.

Fig. 5.

Energetic, morphological, and kinematic measurements mapped on phylogenetic relationships of fishes. MO_2hover / MO_2rest, MO_2net = MO_2hover − MO_2rest in mgO₂ kg⁻¹ h⁻¹. COM–COM A-P Axis = difference in position between the center of mass and center of buoyancy as a proportion of body length along the anterior–posterior axis. COM–COB D-V Axis = difference in position between the center of mass and center of buoyancy as a proportion of body depth along the dorsal–ventral axis. Pect. fin A-P axis = pectoral fin position along the anterior–posterior axis as a proportion of body length. Caud. fin A-P axis = caudal fin position along the anterior–posterior axis as a proportion of body length. All values have been centered and scaled for each variable. Color gradient ranges from light (low values) to dark (high values) blue. Phylogenetic relationships follow Betancur-R et al. (22).