Ontogenetic habitat shifts and habitat use in an endangered minnow, Notropis mekistocholas

Henderson AR, Johnston CE. Ontogenetic habitat shifts and habitat use in an endangered minnow, Notropis mekistocholas . Ecology of Freshwater Fish 2010: 19: 87–95.
© 2009 John Wiley & Sons A/S
Abstract –  Information on larval and juvenile habitat use is virtually absent for fishes. Our objective was to document habitat use of larval and juvenile Cape Fear shiners Notropis mekistocholas , an endangered species, in both natural and lab settings so that a better understanding of habitat requirements in all life stages can be achieved. We measured habitat parameters of areas used by all life stages of N. mekistocholas in the summers of 2007–2008 in the Rocky River, North Carolina. Field data suggests larvae use more shallow depths (mean = 31.6 cm) and reduced water velocities (mean = 0.02 m·s⁻¹) than adults (mean = 56.3 cm, 0.10 m·s⁻¹), and juveniles 15–25 mm often school with adults. Results of lab studies support field observations. In the mesolarval and metalarval stages N. mekistocholas were selective in their use of water velocity ( P  = 0.013) and depth ( P  = 0.001). In multiple juvenile stages, N. mekistocholas were selective for water velocity ( P  = 0.015), and depth ( P  < 0.001), choosing deeper depths and higher water velocities than larvae. These results demonstrate ontogenetic shifts in habitat use of N. mekistocholas and suggest that water velocity and depth are critical to successful recruitment for this species.     

Habitat use and movement of river blackfish (Gadopsis marmoratus R.) in a highly modified Victorian stream, Australia

Abstract –  Movement and habitat use of river blackfish ( Gadopsis marmoratus R.) was studied in a highly modified central Victorian stream (Birch Creek), Australia, using radio telemetry ( N  = 10) and mark–recapture ( N  = 113) between October 2001 to May 2002, and November to December 2002 respectively. Blackfish had a small home range of 10–26 m with strong affinity to a pool. However, small movements within a pool were common which resulted in relatively large cumulative movements. At a diel scale there were no significant differences in blackfish movement between day and night (Kruskal–Wallis test, P  = 0.95). Six of seven blackfish translocated 40 m upstream or downstream of their original positions returned within 48 h, suggesting strong affinity not only to a pool, but also to a small region within a pool. Blackfish utilised slow flowing (0–20 cm s⁻¹) and deep waters (40–60 cm). Blackfish were strongly associated with the instream cover habitats of undercut banks and boulders. Significant differences were recorded in instream cover and water velocity used by blackfish between day and night (Kruskal–Wallis test, P  = <0.01). At night blackfish used open areas with high water velocities whereas during the day blackfish used complex undercut banks with low water velocities.     

Movements and habitat use by pikeperch (Stizostedion lucioperca (L.)) in a hydropeaking reservoir

Abstract –  We used radio telemetry to measure the habitat use and movements of 20 tagged pikeperch (44–59 cm) in the hydropeaking Pyhäkoski Reservoir, Finland, during June 1999–2000. The hydrological conditions in the 8-km-long reservoir were measured and then modeled over eight discharge rates between 30 and 700 m³ · s⁻¹. Pikeperch preferred relatively low water velocities and deep water depths especially during the winter. During the summer, pikeperch moved actively in different parts of the reservoir. No relationship was found between hydroelectric plant operations and distances moved by pikeperch during the summer. Movements peaked in autumn but decreased during the winter to a restricted area. After 1 year of monitoring, 4 out of 20 tagged pikeperch remained in the Pyhäkoski Reservoir, three had died when descending from hydroelectric dams, 10 had descended downstream to other reservoirs or sea, two were captured by local fishermen, and one fish's signal was lost for unknown reasons.     

Microhabitat use, movements and abundance of gilt darters (Percina evides) in southern Appalachian (USA) streams

Abstract –  We examined microhabitat use by gilt darters ( Percina evides ) in two streams in the south-eastern USA. Darters were over-represented in erosional microhabitats with higher average velocities and more cobble. Male darters tended to show stronger selection than females. Size-based analyses showed that larger (≥60 mm) gilt darters tended to use microhabitats with more heterogeneous substrata and more boulder than smaller (≤59 mm) darters. We also conducted a short-term movement study and calculated population estimates based on mark–recapture data in autumn 2005. Darters moved both long and short distances with 40% of all recaptures occurring within 5 m of the initial capture point. Using Program mark and model-averaged parameter estimates gilt darter density was 0.31 darters·m⁻² (225 darters/730 m²). Conservation of this species will require the preservation of erosional habitats in streams.     

Production and predation rates in a cannibalistic Arctic char (Salvelinus alpinus L.) population

Abstract – Production of cannibalistic Arctic char ( Salvelinus alpinus (L.)) and their prey conspecifics were estimated in an allopatric population, situated in an alpine ultraoligotrophic lake in central Norway. Estimates of population abundance were obtained by mark recapture and successive removal experiments. Assuming stable age distribution, we estimated mean yearly biomass (±95% CL) of char in sampled size groups vulnerable to predation (60 to 150 mm) to 7.93 (5.11–14.30) kg · ha⁻¹. Similarly, mean yearly biomass (±95% CL) of cannibalistic char (> 250 mm) was estimated to 0.62 (0.50–1.06) kg · ha⁻¹. Annual production (±95% CL) of char in length groups 60 to 150 mm was estimated to be 4.31 (2.74–8.03) kg · ha⁻¹ · year⁻¹, and production (±95% CL) of cannibalistic char to 0.19 (0.15–0.33) kg · ha⁻¹ · year⁻¹. Depending on the food conversion factor (set to vary from 0.1 to 0.4), the cannibalistic char removed from 10% to 40% of the production of char in sampled size-classes vulnerable to predation yearly. The overall ecological efficiency in energy transfer between the prey and predator population was 4.4%.     

Vertical distribution of jack mackerel Trachurus japonicus larvae in the southern part of the East China Sea

ABSTRACT:   The vertical distribution of jack mackerel Trachurus japonicus larvae was described based on discrete depth sampling using a MTD net system from the surface down to 100 m depth at 13 stations in the southern East China Sea between 27 February and 10 March 2002. Of the 20 782 fish larvae collected, T. japonicus larvae were most abundant, accounting for 37.0% of the total catch. The average abundance of the larvae was 419.0 individuals/10 m², with the average (± standard deviation [SD]) body length of 2.6 ± 0.3 mm (range 1.8–7.5 mm). The larvae were concentrated in the mixed layer, with peak densities in the 10–30 m layer. The average (±SD) weighted mean depth was 21.5 ± 7.8 m. There was no evidence of either diel or ontogenetic vertical migration for the early larvae of <5 mm. More than 90% of the T. japonicus larvae were collected in the water temperature ranging from 21 to 23°C. Vertical profiles of the larval densities and chlorophyll a coincided with each other, which might reflect the abundance of their main prey organisms, copepod nauplii and copepodites, since copepod production is known to be closely related with the chlorophyll a concentration.     

Improving the upstream passage of two galaxiid fish species through a pipe culvert

Abstract  Movement between habitats in river fish assemblages is often restricted by instream structures such as culverts. The ability of diadromous common jollytail, Galaxias maculatus (Jenyns), and spotted galaxias, Galaxias truttaceus (Val.), to pass upstream through an in situ pipe culvert modified through the installation of baffles was assessed. Spoiler baffles (100 × 70 × 28 or 56 mm) were installed in three spatial arrangements along a 5.5-m section of the pipe, and individual fish passage assessed at three flow velocities (0.35, 0.70 and 1.0 m s⁻¹). Common jollytails (43–169 mm fork length, FL) were 10 times more successful in passing when baffles were present than under control conditions (baffles absent). Baffle size did not influence success, which increased with the spatial complexity of the baffle arrangement. Across all velocities, common jollytails (46–132 mm FL) and spotted galaxias (55–190 mm FL) were, respectively, 86 and 73 times more successful with the most complex baffle arrangement (overall 80% success) compared with control conditions (overall 13.5% success). Success for both species decreased at higher velocities under control conditions; however, when baffles were present, this trend persisted only for common jollytails. Installing small spoiler baffles may provide a simple, cost-effective solution to passage problems at culverts.     

Size-dependent predation on marine fish larvae by Ctenophores, Scyphomedusae, and Planktivorous fish

Size-dependent predation rates on marine fish larvae by the ctenophore Mnemiopsis leidyi , scyphomedusa Chrysaora quinquecirrha , and planktivorous fish Anchoa mitchilli were estimated via experiments in 3.2 m³ me-socosms. Predation rates on larvae of the goby Gobio-soma bosci were estimated in relation to 1) length of larval prey, 2) presence or absence of alternative < 1 mm zooplankton prey, and 3) a predator-prey interaction between the scyphomedusa and ctenophore. The consumption rate of larvae by the three predators generally declined as larval length increased. The ctenophore ate fewer (1.0 d^-1 per predator) larvae than did the medusa (7.8 d ^-1 per predator) or bay anchovy (4.0 d ^-1 per predator), but it consumed larvae in all size classes tested (3.0–9.5 mm SL). Predation by bay anchovy and the medusa was more size-dependent: 3.0–5.0 mm larvae were most vulnerable. Fewer larvae were eaten by the ctenophore and bay anchovy when zooplankton alternative prey was available, but predation on larvae by the medusa was not influenced by alternative zooplankton prey. Consumption rate of fish larvae by the medusa was reduced 20–25% when the ctenophore was present as its alternative prey. An encounter-rate model was developed and its parameters estimated from the experimental results. Model simulations indicated that the relationship between larval size and vulnerability is dependent on size, swimming speed, and behavior of both predators and larvae, and that bigger or faster-growing fish larvae are not always less vulnerable to predation.     

Growth model for the endangered cyprinid fish Tribolodon nakamurai based on otolith analyses

A growth model for the endangered cyprinid fish Tribolodon nakamurai was derived following otolith analyses of 16 wild and 53 reared specimens. The asteriscus was the most appropriate to measure size among three otolith elements, and its height OH  mm was used as size index of otolith. Standard length L  cm was best back-calculated using the Gompertz model, L  = 70.0·exp[–exp{−0.553 (OH   –  2.73)}]. Translucent zones on the lapilli, analyzed from 5-year-old-reared fish, were regarded as winter slow-growing zones. The ages of 10 wild specimens of 37.0–48.1 cm standard length were calculated as 7–10 years by counting the translucent zones on the lapilli. Age t was best back-calculated using the allometry model, t  = 1.33· OH ^1.37. The growth trajectory of T.   nakamurai followed a slender S curve, three typical growth models, von Bertalanffy, Logistic and Gompertz, and Richards' model, which is a general formula of the above three, being fitted using the maximum likelihood method. The Gompertz model, L_t  = 60.2·exp[–exp{−0.258( t  − 4.68)}], was found by Akaike's information criterion (AIC) to be the statistically most acceptable growth model.     

Growth of juvenile tilapia, Oreochromis niloticus L. from Lakes Zwai, Langeno and Chamo (Ethiopian rift valley) based on otolith microincrement analysis

Abstract – Age and growth of juvenile tilapia, Oreochromis niloticus , from Lakes Zwai, Langeno and Chamo (Ethiopia) were studied from microincrements in otoliths. Growth in length was best described by the Gompertz model. Average growth rate of the fish was most rapid in Lake Chamo (0.39 mm  ·  day⁻¹, 1.14%  ·  day⁻¹), intermediate in Lake Zwai (0.20 mm  ·  day⁻¹, 0.72%  ·  day⁻¹) and slowest in Lake Langeno (0.16 mm  ·  day⁻¹, 0.62%  ·  day⁻¹). Similarly, back-calculation from otolith increment widths gave growth rates of 0.28 to 0.43 mm  ·  day⁻¹, 0.15 to 0.32 mm  ·  day⁻¹ and 0.11 to 0.28 mm  ·  day⁻¹ for Chamo, Zwai and Langeno fish, respectively. In addition, Fulton's condition factor was largest for Chamo tilapia and smallest for Langeno tilapia; the difference between fish from Langeno and Zwai was small. Rapid growth of juvenile O. niloticus in Lake Chamo was attributed to warm temperature and better food quality. ^Note     

The potential of liposomes as a nutrient supplement in first-feeding marine fish larvae

The ingestion rate (ng liposome larva^–1 h^–1) of extruded [1–¹⁴C] palmitic acid-labelled liposomes containing physiological saline (PHS) or cod fish extract (CFE), was tested in 5-day-old gilthead seabream Sparus aurata and white grouper Epinephelus aeneus larvae. A follow-up study compared the assimilation of radioactive free fatty acid (FFA) label of these two liposome treatments into six phospholipid and neutral lipid fractions as well as the nonlipid fraction in 5-day-old seabream. In seabream larvae, there was a 50-fold ( P  < 0.05) increase in the net consumption rate when fed CFE liposomes (2305.8 ng liposome larva^–1 h^–1) compared with liposomes containing physiological saline (42.7 ng liposome larva^–1 h^–1). A similarly significant ( P  < 0.05) but less marked pattern was also observed in the grouper larvae where the CFE treatment larvae ingested 238.5 ng liposome larva^–1 h^–1 compared with 54.3 ng liposome larva^–1 h^–1 in larvae fed the PHS liposomes. In seabream larvae ingesting CFE and PHS liposomes, radioactivity was found in all larval fractions analysed. However, marked treatment differences ( P  > 0.05) in assimilation were found only in the triacylglycerol fraction (3.4 and 0.6 dpm larva^–1 h^–1, respectively) and nonlipid fraction (11.2 and 15 dpm larva^–1 h^–1, respectively).     

Effects of a peroxide-based commercial product on bacterial load of larval rearing water and on larval survival of two species of Sparidae under intensive culture: preliminary study

Larvae of two Mediterranean Sparidae species, Sparus aurata and Dentex dentex , were used to test the efficacy of a peroxide-based product (Ox-Aquaculture^©) on the reduction in bacterial load in larval rearing water and its effects on larval survival. Eleven-day-old S. aurata larvae and 15-day-old D. dentex larvae were exposed to different concentrations of Ox-Aquaculture^© (50, 100 and 200 mg L⁻¹, and 20 and 50 mg L⁻¹ respectively) for 1 h. Results indicated that 50 and 20 mg L⁻¹ were the most effective concentrations for the reduction in bacterial load (at least one order of magnitude) after 1 h treatment, without affecting larval survival and/or vitality in 11 dph S. aurata and 15 dph D. dentex larvae respectively. Ox-Aquaculture^© concentrations of 200 and 50 mg L⁻¹ during 1 h affected negatively final survival rate of the larvae of S. aurata and D. dentex respectively.     

Ingestion of Artemia nauplii by Chinese mitten crab Eriocheir sinensis zoea larvae

A series of ingestion trials were conducted to determine the ingestion rate of Artemia nauplii by Eriocheir sinensis zoea larvae with increasing densities of Artemia and with or without rotifers as a co-feed. At each zoeal stage, 10 groups of 10 larvae were reared individually in glass beakers and fed with increasing densities of newly hatched Artemia nauplii (0.5, 2.5, 5, 10 and 20 individual (ind.) mL⁻¹) with or without rotifers (15–25 ind. mL⁻¹) as a co-feed. The average number of ingested Artemia was measured over 24 h. In addition, the average larval development rate (Larval Stage Index, LSI) over a longer period (time needed for the best treatment to reach 100% moult or metamorphosis to the next larval stage) was compared. The results showed that Artemia ingestion rate of E. sinensis larvae increased with increasing prey densities and larval development, and had a significantly negative correlation with rotifer consumption for all zoeal stages. Rotifers as an alternative prey significantly affected the intake of Artemia at early larval stages (Z1 and Z2) and promoted LSI at a lower Artemia density. Further experiments are needed to clarify the effect of prey density on survival and larval development when larvae are reared communally.     

Exchange and flux of larval fishes across the western Gulf Stream front south of Cape Hatteras, USA, in winter

To describe exchange and to frame the magnitude of the flux of larval fishes across the Gulf Stream front in winter, collections were taken on the inshore and offshore sides of the frontal zone in 1991, 1993, 1994 and 1995, off Onslow Bay, North Carolina. Collections yielded larval fishes representing 75 families. Of these, the larvae of the families Gonostomatidae or Phosichthyidae consistently occurred on both sides of the frontal zone, yet are exclusively of mesopelagic, deep-oceanic origin; hence, they indicate Gulf Stream-to-shelf water-exchange. The larvae of the clupeid, Brevoortia tyrannus , and the sciaenid, Leiostomus xanthurus , occurred on both sides of the frontal zone, yet spawn on the outer continental shelf in or near the Gulf Stream front; hence these taxa indicate shelf-to-Gulf Stream water-exchange. The mechanisms of exchange, as indicated by the distribution of heat and salt across the front, along with satellite images of sea-surface temperature, were not completely resolved, largely because of the two-dimensional ( x , z ) rather than three-dimensional ( x , y , z ) coordination employed. The observed distributions of indicator taxa did not always conform with expected, given the distribution of heat and salt across the front. Parcels of water exchanged across the front upstream (to the south) and encountered during observation along the cross-shelf transects explains the lack of conformity. Where the expected and observed distributions agreed, the flux was 0.0009 larvae m⁻¹ s⁻¹ for the indicators of Gulf Stream-to-shelf exchange, and 0.0007–0.0030 larvae m⁻¹ s⁻¹ for the indicators of shelf-to-Gulf Stream exchange.     

Use of neuroactive catecholamines to chemically induce metamorphosis of hatchery-reared flat oyster, Ostrea angasi, larvae

Low numbers and unreliable wild catch of the native flat oyster, Ostrea angasi , spat has resulted in the NSW flat oyster industry being reliant on hatchery-produced spat. The need to produce culchless spat in the hatchery stimulated investigation of several catecholamines to induce metamorphosis in O. angasi larvae. Larvae were treated with one of four neuroactive catecholamines (epinephrine, epinephrine bitartrate, l -Dopa and GABA) at one of four concentrations (10⁻³, 10⁻⁴, 10⁻⁵ or 10⁻⁶  m ) for one of three treatment durations (0.5, 1–2 h) to determine morphogenic action for culchless spat production. Epinephrine bitartrate at 10⁻³ and 10⁻⁴  m and epinephrine at 10⁻³, 10⁻⁴ and 10⁻⁵  m , for a treatment duration of 1–2 h, produced significantly greater numbers of spat and culchless spat, compared with any other treatment combination. The other catecholamines tested did not induce a significant increase in the total number of spat or culchless spat, over untreated controls. Separate trials found that long-term treatment (24 h) with epinephrine bitartrate and epinephrine at morphogenic concentrations inhibited metamorphosis. Consecutive daily use of epinephrine bitartrate increased the numbers of spat and culchless spat produced, but did not affect larval or short-term post-larval survival. Treatment with 10⁻³  m epinephrine bitartrate or 10⁻⁴  m epinephrine for 1 h is recommended for routine commercial production of culchless flat oyster spat.     

Timing of brown trout spawning in Alpine rivers with special consideration of egg burial depth

Timing of spawning, habitat use and egg burial depths of brown trout were studied in seven Swiss (alpine and prealpine) rivers, which differed in size, altitude and flow regime. In general, we observed brown trout spawning activity between the end of October and the beginning of January. The spawning duration differed significantly, however, between rivers, ranging from 28 to 72 days. Analysis of environmental parameters for their influence on spawning activity revealed mean water temperature and altitude as the most explanatory variables. Detailed investigation of redd characteristics based on water velocity, water depth and substrate size clearly showed differences between positions on the redd. Brown trout in Alpine rivers preferred to use velocities of 30–40 cm·s^?1, water depths of 10–20 cm and substrates of 16–32 mm for spawning. It has to be noted, however, that recorded values cover almost the whole range of data on spawning habitats that has been reported in literature so far. A special focus of this study was on egg burial depths, which were surprisingly not found to differ significantly between the rivers despite their different flow regimes. Recorded egg burial depths were, however, found to be distinctly lower (mean burial depth: 3.8 cm) than reported by almost any study so far. We see this observation of low burial depths in Alpine rivers as useful in the context of scouring effects, especially when evaluating the influence of scouring on fish populations.     

Swimming performance of wild and F1-hatchery-reared Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) smolts

Abstract –  The swimming performance of wild and hatchery-reared smolts of two salmonid species was investigated. Wild Atlantic salmon smolts (WS) and brown trout smolts (WT) of equal size were caught in fish traps during migration. Hatchery-reared smolts of both species (HS and HT for salmon and trout respectively) were first generation offspring from wild broodstock. The swimming performance of individual smolts from the four groups (WS, HS, WT, HT) was tested three consecutive times using a swimming flume with water flowing at a start rate of 0.16 m·s⁻¹ and a constant acceleration rate of 0.167 cm·s⁻² (10 cm·s⁻¹·min⁻¹). Wild caught smolts of both species performed significantly better than those reared in hatchery conditions. The WS group were observed to maintain an average swimming speed ( U _burst) that was 30% faster than the HS group, whereas the wild trout smolts were superior to HT by approximately 25%. Repeated measures revealed species-specific exhaustion patterns. Brown trout smolts maintained consecutive U _burst indicating significant stamina compared with Atlantic salmon smolts that were found to be exhausted by the initial trial.     

Growth of 0+ perch (Perca fluviatilis) predating on 0+ bream (Abramis brama)

Abstract – Larvae and juveniles of perch Perca fluviatilis and bream Abramis brama of Lake Speldrop, a highly eutrophicated gravel-pit lake in the floodplain of the Lower Rhine, were used in laboratory experiments to study predation of perch on bream. In the first series of experiments (control), 0+ perch of 30 mm total length (TL) did not prey on 0+ roach of about 24 mm TL. The perch fed only on zooplankton, resulting in low growth rates of 0.17 mm  ·  day⁻¹. In the second series of experiments, perch (30 mm TL) were combined with 0+ bream with an average TL of 14 and 19 mm at different ratios. As in the first series, even a nine-fold higher amount of zooplankton food had no significant influence on the growth rate of perch, regardless of the perch:bream ratio. At a perch: bream ratio of 7:1 and 4:4 all bream were eaten by the perch and restocked twice daily. At the end of the experiment, the TL of perch at a perch:bream ratio of 7:1 was significantly lower compared to a perch:bream ratio of 4:4 (mean growth rates at 7:1 of 0.16 mm  ·  day⁻¹ and at 4:4 of 0.35 mm  ·  day⁻¹). At the perch:bream ratio of 1:7, perch fed on bream in varying amounts, averaging between 3 and 12 bream per day and per individual. After 20 days, the final length of these fish was always significantly higher than the TL of perch at other perch:bream ratios and increased in correlation to number of bream eaten per day (mean growth rate 1.13 mm  ·  day⁻¹). The results are discussed with respect to the piscivore-dominated fish community of the eutrophied Lake Speldrop. ^Note     

Angling pressure, yield and catch per effort of grayling, Thymallus thymallus (L)., and brown trout, Salmo trutta L., on the rivers Glomma and Rena, southeastern Norway

Abstract Postal questionnaires, interviews and accurate reports from anglers were used to study angling activity, yield and catch per effort of grayling, Thymallus thymallus (L.), and brown trout, Salmo trutta L., on three stretches of the regulated river Glomma and two stretches of its tributary, the river Rena in southeastern Norway. Angling activity was expressed as hours angling per kilometre of river and varied from 364 (CL 245–598) to 1337 (CL 912–1991) h km^-1. Grayling and brown trout dominated the catches. The catches were estimated between 476 (CL 334–659) and 753 (CL 517–1155) grayling and brown trout per kilometre river stretch. The mean weights were estimated between 216 and 353 g for grayling, and 260 and 395 g for brown trout. The yield of grayling and brown trout was estimated between 8.7 (CL 6.1–12.4) and 46.4 (CL 34.9–62.3) kg ha^-1 yr^-1. The quotient grayling/brown trout in the catches varied between 0.41 and 5.1, and was found to be significantly negatively correlated with angling effort ( r = - 0.838, P <0.05). Catch per effort was estimated between 132 (CL 82–218) and 382 (CL 287–513) g h^-1, and was also negatively correlated to angling effort, according to: Catch per effort (g h^-1) = - 0.210 × effort (h km^-1) + 405 ( r = and 0.899, P <0.02).
Angling activity in earlier years was determined from number of permits sold where data were available. On stretches where angling activity was moderate, there was a tendency towards increase, whereas on stretches where activity had been high for several years, the activity seemed to have stagnated, indicating that a saturation had been reached.     

Seasonal variation in body composition of Arctic char, Salvelinus alpinus, from an ultraoligotrophic alpine lake

Abstract  – Seasonal pattern of energy content was determined in a population of Arctic char, Salvelinus alpinus (L.), living in an ultraoligotrophic alpine lake (altitude 1100 m) located in central Norway (62°5'N). Specific somatic energy content varied between 4.3 and 6.1 kJ · g⁻¹ in spawners and between 4.8 and 6.6 kJ · g⁻¹ in immatures. Corresponding values for specific somatic lipid energy were 0.5–1.6 kJ · g⁻¹ for spawners and 0.7–2.1 kJ · g⁻¹ for immatures. The temporal pattern in storage energy residuals (deviation from mean storage energy, all sampling periods pooled) indicated that total, protein and lipid energy accumulated in the autumn and early winter. This increase was followed by a winter decrease in somatic energy (January to April). A new increase in total somatic energy and somatic lipid energy occurred during late winter (March/April to June), while the lake was still ice covered and water temperatures ranged between 0.5 and 2.5 °C. The observed seasonal pattern of specific energy storage and lipid deposition demonstrated a considerable potential for energy accumulation in Arctic char at low temperatures. It is discussed whether freshwater lakes at high altitudes or latitudes represent a temperature-stabilised and predictable environment in contrast to the variable temperature experienced in the corresponding terrestrial system.