Life history and essential habitats of humpback whitefish in Lake Clark National Park, Kvichak River watershed, Alaska
(continued)

OBJECTIVES

1) Determine basic life history characteristics of Lake Clark National Park humpback whitefish populations including age and size, age at maturity, fecundity, and anadromy.

2) Determine seasonal migration patterns and habitat use of Lake Clark National Park humpback whitefish populations.

METHODS

Study Site

The Lake Clark watershed (60° 01’ N, 154° 45’ W) drains an area of about 7,620 km2 and is part of the greater Kvichak River watershed in southwest Alaska (Figure 1).  Lake Clark is the sixth largest lake in Alaska with a surface area of 267 km2, length of 66 km, width of 5 km, and an average depth of 103 m, and a maximum depth of 322 m (Anderson 1969, Wilkens 2002).  Glaciers, steep mountains, glacial rivers, and high precipitation (average 203 cm annually) characterize the upper watershed; lowland tundra, small mountains, clear and stained streams, and low precipitation (average 64 cm annually) characterize the lower watershed (Jones and Fahl 1994, Brabets 2002).  Six primary tributaries feed Lake Clark, with glacier-fed tributaries contributing about half the annual water budget and up to a million tons of suspended sediment annually (Brabets 2002).  High suspended sediment inputs result in reduced water clarity, which impedes visual fish surveys.

Sockeye salmon are a known food resource for more than 40 different species of animals (Willson et al. 1998), including humpback whitefish.  Sockeye salmon are an important cyclical resource to the Lake Clark National Park ecosystem.  Escapements have shown an increasing trend since 2000 and have ranged from 200,000 to more than 700,000 sockeye salmon (Woody 2004, Young and Woody 2006, National Park Service, unpublished data).

Life History

Capture Methods

We tested various capture methods in 2005 including seines, hook (#12) and line, gillnets with uniform and variable mesh sizes, fyke nets, minnow traps, and hoop nets.  Sampling was conducted over a randomly selected range of available fish habitats and primarily in littoral areas (<10 m) due to gear constraints.  Based on input from local subsistence fishers, fish eggs were used to pre-bait sampling areas and to bait traps in an effort to attract whitefish to gear.  Fishing with seines and hook and line was conducted during daylight hours, while 24 hr sets were made with gillnets, fyke nets, minnow traps, and hoop nets.  We used an ACCESS database to document sampling locations, general habitat type and humpback whitefish captures (Appendix I). 

Size and Age

Each captured humpback whitefish was measured (total and fork length) and weighed, and three scales were collected for age estimates.  Additionally, about 100 humpback whitefish over a range of sizes were sacrificed to obtain otoliths (ear bones) for age estimates.  We later compared ages estimated from both scales and otoliths from the same individuals.  We hoped that similar estimates would be obtained from both structures, or that a correctable bias was found, so that captured humpback whitefish would not have to be sacrificed to obtain age data.  Otolith aging criteria followed Mills and Beamish (1980), Chilton and Beamish (1982), and Howland et al. (2004).  Scale aging followed Howland et al. (2004).  All otoliths and scales were aged twice.

Verification of Anadromy

Ten otoliths were selected across sample sites to determine whether Lake Clark humpback whitefish were anadromous.  We tested for anadromy by analyzing strontium (Sr) concentrations in otoliths (Figure 4).  This method has been used for humpback whitefish collected from the upper Tanana River drainage (Brown 2006).  It is based on the documented influence of salinity on the chemical composition of fish otoliths (Fowler et al. 1995a and b; Mugiya and Tanaka 1995; Secor et al. 1995; Farrell and Campana 1996), and studies on salmonid otoliths that showed a significant rise in Sr concentration when fish experience a change from freshwater to 6.3 ppm salinity (Zimmerman and Reeves 2000 and 2002).

Figure 4.  Strontium (Sr) concentration as a proportion of otolith core to margin transects for a selection of known freshwater resident (top 2 rows, F species) and anadromous (bottom 2 rows, A species) fish species.  Horizontal dashed lines are at the 1,700 ppm position.  Freshwater fish do not usually exceed this value, while anadromous fish do (From Brown 2006).

Otoliths selected for microchemical analyses were prepared and interpreted by R. Brown (U.S. Fish and Wildlife Service, Fairbanks) following methods he had used previously (Brown 2000). Each otolith was polished on a lapidary wheel with 1 mm diamond abrasive and coated with a thin layer of conductive carbon in preparation for microprobe analysis.  Microchemical analysis of otoliths was accomplished using a wavelength-dispersive electron microprobe capable of precise and accurate measurement of otolith Sr concentration (Campana et al. 1997).  The technology functions by bombarding points on a sample surface with a focused beam of electrons.  Atoms within the material are ionized by the electron beam and emit x-rays unique to each element.  Spectrometers are tuned to count the x-rays from elements of interest, in this case, Sr.  X-ray counts at each sample point are proportional to the elemental concentration in the material (Potts 1987, Reed 1997, Goldstein et al. 2003).

RESULTS

Capture Techniques

The most effective capture techniques for humpback whitefish in Lake Clark were seines and gillnets.  Seining in shallow (<5 m) areas, baited with preserved salmon eggs, yielded the best catches.  In rocky areas, where seining was not feasible, constantly monitored variable mesh gillnets were successfully used, but resulted in higher mortality to captured fishes.  Ice fishing during early spring in Sixmile Lake with a single egg on a #12 hook attached to a hand line was another effective, but much less efficient method, and one that is commonly used by subsistence fishers.

Juvenile humpback whitefish (age 0 to 3) were mostly captured by seine in shallow (<3 m) areas of both Chulitna Bay and Long Lake, whereas individuals older than age 4 were captured across a wider range of habitat types including shallow (≤ 2m) small (<6 km long) tributary lakes with abundant aquatic plants (Long and Pickeral Lakes) and deep large fjord lakes (Little Lake Clark; Figure 5).  Humpback whitefish were easily observed and captured in the Newhalen River and Chulitna Bay at subsistence salmon processing sites.

Size and Age

A total of 809 humpback whitefish were sampled during 2005, and 649 of those were categorized as juveniles (≤ 4 years old).  Total lengths of sampled fish ranged between 95 and 584 mm (Figure 6) with 454 individuals measuring less than 119 mm.

Estimated ages ranged from age 0 to 27.  Ages estimated from paired otolith and scale samples (N=110) were usually similar for individuals with total lengths between 82 to ≤ 230 mm, but differed, sometimes dramatically, for individuals longer than 400 mm (Figure 7).  In general, estimated otolith ages were greater than estimated scale ages for the same individual (Figure 8), and less variation was observed between two readings of the same otolith (SD = 0.24) than between two readings of the same scale (SD = 0.87).

Figure 5.  Sample sites (red triangles) fished for humpback whitefish, Lake Clark drainage, 2005.  Pickerel Lake system tributaries were also sampled.

Figure 6.  Size frequency distribution for Lake Clark humpback whitefish sampled in 2005.

Figure 7.  Differences in estimated ages from otoliths and scales by length for humpback whitefish, Lake Clark, 2005.  Deviations above 0 indicated that the otolith age was older than the corresponding scale age.

The differences in age estimates between otoliths and scales produced different von Bertalanffy growth equations (Figure 9) giving different maximum length estimates: otolith L∞ = 507 (95% confidence interval 488, 531); scale L∞ =562 (confidence interval 533,600).

Figure 8.  Age estimates from otoliths versus scales.  The points left of the 1-1 diagonal indicate that for most humpback whitefish the otolith age exceeded the scale age.

Figure 9.  Von Bertalanffy growth equations for age estimates from otoliths and scales.  Equations result in different maximum length estimates: otolith L∞ = 507 (95% confidence interval 488, 531); scale L∞ =562 (confidence interval 533,600).

Verification of Anadromy

No definitive spikes in Sr concentrations above 1,700 ppm were observed for the 10 Lake Clark humpback whitefish otoliths analyzed, although four fish (05-108, 05-122, 05-127, and 05-152) had values that were near or at this level (Figure 10).  Also, most individuals exhibited greater variation in Sr concentrations in comparison to values for freshwater fishes examined in another study (Figure 4).

Figure 10. Strontium (Sr) concentration as a proportion of otolith core to margin transects for 10 Lake Clark humpback whitefish. No definitive spikes of Sr concentration are apparent suggesting these fish migrated to marine habitats; however, it is possible that fish 05-108, 05-122, 05-127, or 05-152 did because of the greater variation here when compared to non-anadromous fish (e.g. Figure 4). Another explanation for the variation may be vertical movements of whitefish in Lake Clark.

(continued to Discussion)