Natural selection drives the evolution of traits to optimize organismal performance, but optimization of one aspect of performance can often influence other aspects of performance. Here, we asked how phenotypic variation between locally adapted fish populations affects locomotion and ventilation, testing for functional trade-offs and trait-performance correlations. Specifically, we investigated two populations of livebearing fish (Poecilia mexicana) that inhabit distinct habitat types (hydrogen-sulfide-rich springs and adjacent nonsulfidic streams). For each individual, we quantified different metrics of burst swimming during simulated predator attacks, steady swimming, as well as gill ventilation. Coinciding with predictions, we documented significant population differences in all aspects of performance, with fish from sulfidic habitats exhibiting higher steady swimming performance and higher ventilation capacity, but slower burst swimming. There was a significant functional trade-off between steady and burst swimming, but not between different aspects of locomotion and ventilation. While our findings about population differences in locomotion performance largely parallel the results from previous studies, we provide novel insights about how morphological variation might impact ventilation and ultimately oxygen acquisition. Overall, our analyses provided insights into the functional consequences of previously documented phenotypic variation, which will help to disentangle the effects of different sources of selection that may coincide along complex environmental gradients.

Study organisms and general experimental design

Several sulfide springs inhabited by P. mexicana occur in the states of Tabasco and Chiapas, Mexico (Palacios, et al. 2016). For the present study, we collected young adult fish (standard length: 28-38 mm) from a sulfidic (El Azufre I: N 17.442°/W 92.775°) and a nonsulfidic population (Río Tacotalpa: N 17.275°/W 92.462°) near Tapijulapa, Tabasco, and transported them to Kansas State University (KSU), where they were housed in 80-liter tanks and given two months to acclimate to standardized laboratory conditions, irrespective of their habitat of origin. Fish were kept at under nonsulfidic and normoxic conditions, a constant temperature of 25° C, and a 12:12 hr light:dark photoperiod. Fish were fed ad libitum and had access to flake food and frozen brine shrimp twice daily. We chose to acclimate all individuals to one standardized environment, because maintaining stable H₂S-rich water is difficult in the laboratory. In addition, the toxic effects of H₂S lead to high mortality of fish from nonsulfidic populations, preventing factorial experimental designs where the performance of fish from different habitats is measured under multiple environmental conditions. Consequently, our experiments only quantify the functional repercussions of traits that vary between sulfidic and nonsulfidic populations, and we cannot capture potential trait-environment interactions that might impact performance of fish in their natural habitats.

After acclimation, fish were separated into groups of 3-4 individuals and housed in 40-liter tanks, facilitating the tracking individual fish across different portions of the experiment. For each fish, we quantified three aspects of performance: 1) burst swimming upon a simulated predator attack, 2) steady swimming, and 3) gill ventilation. The order of performance trials was randomized across individuals, and individuals were allowed to recover for at least one week in between trials. Fish were fasted for 24 hours prior to each performance trial to ensure they were in a post-absorptive state (Niimi and Beamish 1974; Kieffer 2000). After the completion of all performance trials, individuals were euthanized using MS-222 (500 mg/l buffered to pH 7.5 with sodium bicarbonate) for morphological analyses. A total of N=71 fish were tested.

 

Burst swimming performance

We quantified burst swimming performance of fish during their reflexive escape response (c-start) to simulated predator attacks (Weihs 1973; Howland 1974; Eaton, et al. 1977; Harper and Blake 1990; Domenici and Blake 1997). Fish from high-predation environments have previously been documented to perform faster at c-starts than fish from low-predation environments (Walker 1997; Langerhans, et al. 2004; Langerhans, et al. 2007), and several studies have linked burst swimming performance with survival in the presence of predators (e.g., Ingley & Johnson, 2016; Walker, Ghalambor, Griset, Mckenny, & Reznick, 2005). 

Methods for the quantification of fast-start responses were adapted from previous studies (Langerhans, et al. 2004; Ingley, et al. 2016). For each trial, we placed a fish into a clear acclimation cylinder (5.5 cm in diameter) within a larger test arena (circular tank with a 40-cm diameter). To minimize vertical displacement and approximate two-dimensional escape responses, the water level was maintained at a depth of ~3 cm. After a 10-min acclimation period, the cylinder was removed, and we struck the arena with a probe (6 mm in diameter and 90 cm long) within ~1 body length of the fish’s caudal region to evoke an escape response. After the first trial, fish were placed back into the cylinder, given the same acclimation period, and tested twice more with the same procedures, yielding three burst swimming performance trials per fish. Each trial was filmed from above with a Sony NXCAM NEX-FS700 high-speed camera (Sony Corporation, Tokyo, Japan) at 120 frames per second (fps). 

Videos were analyzed frame by frame using the DLTdv6 tracking software (Hedrick 2008) in MATLAB 2016a (Mathworks Inc., Natick, MA, USA) to quantify four metrics of burst swimming performance (Walker, et al. 2005) following methods established by Langerhans (2009): (1) total distance traveled (d_net [cm]) is the net distance a fish traveled within 1/12 of a second after bending into the c-shape; (2) rotational velocity (w_S [°/s]) is the average rotational velocity of the head from the moment the fish begins bending into the c-shape until it has completely bent (rotational angle of bend divided by duration); (3) maximum velocity (v_max [cm/s]) is the greatest change in distance between two consecutive frames (1/120 of a second); (4) maximum acceleration (a_max [cm/s²]) is the greatest positive change in velocity between two consecutive frames.For each individual, we calculated the average of each metric across the three trials.

 

Steady swimming performance

We quantified metrics for steady swimming performance, an energy-efficient mode of locomotion that fish use during place-holding against water flow, foraging, and mate searching (Domenici 2003b; Blake 2004; Langerhans 2008). To quantify steady swimming performance, we measured the critical swimming speed as well as swimming kinematics at different swimming speeds for each individual.

Critical swimming speed is defined as the maximum speed at which a fish can maintain steady swimming (Plaut 2001; Domenici 2003a; Blake 2004) and can be quantified by incrementally increasing the swimming speed until an individual fatigues (Brett 1964). Critical swimming speed can then be calculated as U_crit = U_f + U_s × (t_f/t_s), where U_f is the highest flow velocity maintained for a full time-interval, U_s is the velocity increment, t_f is the time to fatigue at the last flow speed, and t_s is the time interval at which increases in speed occur (Brett 1964). To quantify U_crit, individual fish were placed in a 5-L swim tunnel (Loligo Systems ApS, Viborg, Denmark), in which they were exposed to laminar flow with adjustable speed (Ellerby and Herskin 2013){Ellerby, 2013 #83;Ellerby, 2013 #83}. Trials started with a 20-min acclimation period, including 10 min without flow and 10 min at a flow speed of one body length per second (BLs^-1). After acclimation, fish were incrementally exposed to higher flow speeds (U_s = 1 BLs^-1) every 10 min (t_s = 600 s). Acclimation times and speed increments were adopted from previous studies (Hammill, et al. 2004; Oufiero and Garland 2009; Sfakianakis, et al. 2011). Trials were immediately terminated once a fish reached fatigue.

During each steady swimming trial, fish were filmed at 120 (fps) using a high-speed camera twice during each interval (once immediately after the speed was increased and a second time halfway into a time interval). Four videos were selected at speeds ~25%, 50%, 75%, and 100% of each individual fish’s U_crit (Oufiero and Garland 2009). At each speed, we quantified five kinematic variables relevant for the hydrodynamics of steady swimming (see Langerhans, 2009; McHenry, Pell, & Long, 1995): (1) tail-beat frequency (f [Hz]) was measured as the inverse average period of ten complete tail-beat cycles; (2) rostral amplitude (R [mm]) was measured as half the distance between right and left excursions of the anterior tip of the rostrum; (3) tail-beat amplitude (H [mm]) was measured as half the distance between right and left excursions of the caudal fin; (4) propulsive wavelength (λ [mm]) was quantified as double the posterior half-wavelength; and (5) propulsive wave speed (c[mm/s]) was calculated by multiplying the propulsive wavelength with the tail-beat frequency (c= λ *f). For R, H, and λ, measurements were taken by averaging the values of each across three complete tail beats.

 

Gill Ventilation Capacity

Gill ventilation in teleost fishes has been described as a two-pump system, in which water is taken through the mouth into the buccal cavity, pumped into opercular cavity, and out through the opercular openings, resulting in a unidirectional flow across the respiratory surfaces (Hughes 1958). The rate at which water is pumped over the gills is consequently limited by the capacity of the first pump (i.e., the frequency of ventilation and the buccal volume). We quantified these variables for each individual to estimate the maximum gill ventilation capacity.

To measure the maximum ventilation frequency, fish were driven to maximum aerobic performance using chase trials (see Brennan, et al. 2016). Fish were placed in a circular arena (28 cm in diameter), given a 10-min acclimation period, and then chased along the edge of the arena for 6 min. After chasing, individuals were immediately transferred into a photo tank (18´10.5´10 cm) with aerated water and filmed at 60 fps. Fish were recorded eight times for 3 s each, 15 s apart. We counted the number of respiratory cycles for each 3-second interval (in pumps s^-1), retaining the highest value for subsequent analyses.

To estimate the volume of the buccal cavity for each fish, we performed a sagittal cut from the center of the body toward the mouth and a second cut in a planar fashion just behind the opercula, yielding in two separate halves of the head. Digital images were than taken of each sagittal plane using a Canon Rebel T5i camera (Canon USA Inc., Lake Successes, NY, USA), allowing for the measurement of the length of the buccal cavity. Additional images were taken of the planar planes at the anterior and posterior edge of the buccal cavity, allowing for the measurement of the corresponding radii of the buccal cavity. Measurements on the two halves were averaged. To approximate the buccal volume, we treated the shape of the cavity as a circular truncated cone, for which the volume can be calculated as V_b=(1/3)*h*π*(r₁²+r₁r₂+r₂²), where h is the length of the buccal cavity, and r₁ and r₂ are the radii of the anterior and posterior ends, respectively.

 

Body shape

To relate variation in performance to variation in morphology, we quantified the body shape based on lateral photographs of each individual using geometric morphometrics. We digitized 11 landmarks using the software program tpsDig version 2.10 (Rohlf 2006): (1) tip of the upper jaw; (2) anterior and (3) posterior insertions of the dorsal fin; the (4) dorsal and (5) ventral insertions of the caudal fin; (6) the anterior junction of the anal fin; (7) the bottom of the head where the operculum breaks away from the body outlin; (8) the dorsal endpoint of the opercular bone; (9) the dorsal and (10) ventral insertions of the pectoral fin; and (11) the center of the orbit.

 

References

Blake RW. 2004. Fish functional design and swimming performance. Journal of Fish Biology 65:1193-1222.

Brennan RS, Hwang R, Tse M, Fangue NA, Whitehead A. 2016. Local adaptation to osmotic environment in killifish, Fundulus heteroclitus, is supported by divergence in swimming performance but not by differences in excess post-exercise oxygen consumption or aerobic scope. Comparative Biochemistry and Physiology A - Molecular and Integrative Physiology 196:11-19.

Brett JR. 1964. The respiratory metabolism and swimming performance of young sockeye salmon. Journal of the Fisheries Research Board of Canada 21:1183-1226.

Domenici P. 2003a. Habitat, body design and the swimming performance of fish. In:  Bels VLVL, Gasc JP, Casinos AA, editors. Vertebrate biomechanics and evolution. Oxford, U.K.: BIOS Scientific. p. 136-160.

Domenici P. 2003b. Habitat, body design and the swimming performance of fish. In:  Bels VL, Gasc J-P, Casinos A, editors. Vertebrate Biomechanics and Evolution. Oxford: BIOS Scientific Publishers. p. 137-160.

Domenici P, Blake R. 1997. The kinematics and performance of fish fast-start swimming. Journal of Experimental Biology 200.

Eaton RC, Bombardieri RA, Meyer DL. 1977. The Mauthner-initiated startle response in teleost fish. The Journal of experimental biology 66:65-81.

Ellerby DJ, Herskin J. 2013. Swimming Flumes as a Tool for Studying Swimming Behavior and Physiology: Current Applications and Future Developments. In:  Palstra AP, Planas JV, editors. Swimming physiology of fish : towards using exercise to farm a fit fish in sustainable aquaculture: Springer.

Hammill E, Wilson RS, Johnston IA. 2004. Sustained swimming performance and muscle structure are altered by thermal acclimation in male mosquitofish. Journal of Thermal Biology 29:251-257.

Harper DG, Blake RW. 1990. Fast-Start Performance of Rainbow Trout Salmo Gairdneri and Northern Pike Esox Lucius. Journal of Experimental Biology 150.

Hedrick TL. 2008. Software techniques for two- and three-dimensional kinematic measurements of biological and biomimetic systems. Bioinspiration & Biomimetics 3.

Howland HC. 1974. Optimal strategies for predator avoidance: the relative importance of speed and manoeuvrability. Journal of theoretical biology 47:333-350.

Hughes GM. 1958. The Mechanism of Gill Ventilation in Three Freshwater Teleosts. Journal of Experimental Biology 35.

Ingley SJ, Camarillo H, Willis H, Johnson JB. 2016. Repeated evolution of local adaptation in swimming performance: population-level trade-offs between burst and endurance swimming in Brachyrhaphis freshwater fish. Biological Journal of the Linnean Society 119:1011-1026.

Kieffer JD. 2000. Limits to exhaustive exercise in fish. Comparative biochemistry and physiology. Part A, Molecular & integrative physiology 126:161-179.

Langerhans RB. 2008. Predictability of phenotypic differentiation across flow regimes in fishes. Integrative and Comparative Biology 48:750-768.

Langerhans RB, Gifford ME, Joseph EO. 2007. Ecological Speciation in Gambusia Fishes. Evolution 61:2056-2074.

Langerhans RB, Layman CA, Shokrollahi AM, DeWitt TJ. 2004. Predator-Driven Phenotypic Diversification in Gambusia Affinis. Evolution 58:2305-2318.

Niimi AJ, Beamish FWH. 1974. Bioenergetics and growth of largemouth bass ( Micropterus salmoides) in relation to body weight and temperature. Canadian Journal of Zoology 52:447-456.

Oufiero CE, Garland T. 2009. Repeatability and correlation of swimming performances and size over varying time-scales in the guppy (Poecilia reticulata). Functional Ecology 23:969-978.

Palacios M, Voelker G, Arias Rodriguez L, Mateos M, Tobler M. 2016. Phylogenetic analyses of the subgenus Mollienesia ( Poecilia , Poeciliidae, Teleostei) reveal taxonomic inconsistencies, cryptic biodiversity, and spatio-temporal aspects of diversification in Middle America. Molecular Phylogenetics and Evolution 103:230-244.

Plaut I. 2001. Critical swimming speed: its ecological relevance. Comparitive Biochemistry and Physiology Part A: Molecular & Integrative Physiology 131:41-50.

Rohlf FJ. 2006. tpsDig, version 2.10 

Sfakianakis DG, Leris I, Kentouri M. 2011. Effect of developmental temperature on swimming performance of zebrafish (Danio rerio) juveniles. Environmental Biology of Fishes 90:421-427.

Walker JA. 1997. Ecological morphology of lacustrine threespine stickleback Gasterosteus aculeatus L. (Gasterosteidae) body shape. BwlogialJournal of& Linnean SocieQ 61:3-50.

Walker JA, Ghalambor CK, Griset OL, McKenney D, Reznick DN. 2005. Do faster starts increase the probability of evading predators? Functional Ecology 19:808-815.

Weihs D. 1973. The mechanism of rapid starting of slender fish. Biorheology 10:343-350.

 

A ReadMe file explaining all variables is available for download.