OBSERVATIONS ON DOUBLE-CRESTED CORMORANTS (PHALACROCORAX AURITUS) AT SPORTFISHING WATERS IN SOUTHWESTERN UTAH

15 Pages·1994·7.6 MB·English

by  Michael J Ottenbacher

GreatBasinNaturalist54(3),© 1994,pp.272-286
OBSERVATIONS ON DOUBLE-CRESTED CORMORANTS
{PHALACROCORAXAURITUS) AT SPORTFISHING WATERS
IN SOUTHWESTERN UTAH
Michael Ottenbacher^, Dale K. Hepworth^ and Louis N. Berg^
J.
—
Abstract. Counts ofDouble-crested Cormorants {Phalacrocoraxaiiritiis) were made at 13 reservoirs and lakes in
southwestern Utah during 1989-91 to detennine current abundance ofthat species. Food habits ofcormorants were
studiedatthreeofthereservoirs in 1989. Datawerealsocollectedontroutabundanceduringstandardizedgill-netting
tomakecomparisonsbetweencomiorantnumbersandtroutabundance. Cormorantswereobservedatallwatersstudied
exceptoneandweregenerallymostnumerousduringthespringastheymigratedthroughthearea. Estimatedcormorant
abundance ranged from to34bird-days perhaand was highestat the larger, lower-elevation reservoirs. Cormorants
weresummerresidentsatseveralofthelargerreservoirsandnestedsuccessfullyatPiuteReservoir.Troutaccountedfor
24—81%ofthedietofcormorants,with Utahchubsconstitutingmostoftheremainderofthediet. Estimatesoftheannual
consumptionoffishbycormorantsrangedfrom to 15.8kgperha. Theindexoftroutabundancewasinverselyrelated
tocormorantabundance[P < .01)atthewatersstudied.Cormorantsapparentlyhaveincreasedinnumbersandextend-
ed theirrange in southwestern Utah during the past decade. This change maybe the resultoffactors that have led to
similarchanges throughout North America as well as some factors unique to Utah. Methods to mitigate the impact of
predation bypiscivorous birds on sportfisheries are discussed. The Utah Division ofWildlife Resources has initiateda
new management plan at Minersville Reservoir that incorporates piscivorous birds into sportfish management at that
reservoir.
Keywords: cormorants, Phalacrocorax auritus, trout, abundance,foodhabits, predation, management, sportfishing,
reservoirs, Utah.
Various factors influencing survival of fish and potentially impact important fisheries
stocked trout were examined at Minersville (Ayles et al. 1976, Myers and Peterka 1976,
Reservoir, Utah, in 1985-88 (Hepworth and Christie et al. 1987, Campo et al. 1988).
Duffield 1991, Wasowicz 1991). During that Others have felt that cormorants have had lit-
study we observed an increase in the number tle impact on economically valuable species of
ofDouble-crested Cormorants {Phalacrocorax fish (Baillie 1947, Carroll 1988, Findholt
auritus, hereafter referred to as cormorants) at 1988). To evaluate the potential impact ofcor-
Minersville Reservoircomparedwith previous morants on fisheries in southwestern Utah, we
years. An apparent increase in the abundance continued to document the number of cor-
ofcormorants at several other reservoirs was morants at Minersville Reservoir and 12 addi-
also noted, and we received reports of cor- tional waters. We also collected data on trout
morants at some waters where they previously abundance at these waters during standard-
izedannual gill-nettingand initiated a studyof
had not been reported (Walters and Sorenson
1983). Apparent changes in abundance and the food habits of cormorants at three of the
larger reservoirs. Based on these observations,
distribution of this species in Utah coincided we determined current abundance of cor-
wmiotrhanrtepsoritnedmainncyreapsaerstsinotfheNonrutmhbeArmeofricocra- morants at local waters, compared estimates of
cormorant abundance to indices oftrout abun-
(Price and Weseloh 1986, Christie et al. 1987, dance, and estimated annual consumption of
Campo etal. 1988, Findholt 1988). As the rela- fish bycormorants.
tive abundance of cormorants has increased,
there have been conflicting reports concern- StudyArea
ing their impact on recreational fisheries. A
number of authors have concluded that cor- Data on distribution, relative abundance,
morants take considerable numbers of game and seasonal occurrence of cormorants were
^UtahDivisionofWildlifeResources,622N.Main,Box606,CedarCity,Utah84720.
272
1994] CormorantObservations in Southwestern Utah 273
collectedat 13 reservoirs in southwestern Utah variety in fishing opportunity. Recruitment
(Table 1). Reservoirs ranged in size from 36 to from spawning in tributaries associated with
1020 ha, and elevations from 910 to 2695 m reservoirs also provided a small number of
above MSL. Most reservoirs were originally trout in addition to those stocked at some of
constructed for irrigation storage and have the study waters. Stocked trout were harvest-
water levels that fluctuate substantially on an ed by anglers after they reached a catchable
mm
annual basis. Highest water levels occurred in size (>230 TL), generally after they had
late winter and spring, with minimum levels been in the reservoirs for 7-11 months. Few
in the fall following the irrigation season. Fish rainbow trout survive longer than 2 years fol-
Lake and Panguitch Lake are natural lakes lowing stocking (Stuberet al. 1985, Hume and
where storage has been increased by the addi- Tsumura 1992). Most reservoirs contained few
tion of small dams. All the reservoirs except fish species other than trout, and five con-
Quail Creek and Gunlock had ice cover for a tained primarily stocked trout (Enterprise,
period of 2-5 months during winter and Kolob, Koosharem, Lower Bowns, and New-
spring. castle). Three ofthe reservoirs contained only
Sportfishing is a major activity at all ofthe stocked trout and Utah chubs (Gila atraria;
waters since they are open year-round to Minersville. Otter Creek, and Panguitch). Two
angling by the general public with various re- were primarily warm-water fisheries where
strictions (State of Utah 1992). Sportfisheries trout abundance was not evaluated (Gunlock
at all reservoirs except Gunlock are managed, and Quail Creek). The remaining three waters
at least in part, as put-grow-and-take trout (Fish Lake, Johnson and Piute reservoirs) con-
fisheries. Various sizes and numbers of rain- tained more than two other fish species
bow trout {Oncorhynchus mykiss) were besides trout. Only two or three ofthese other
stocked annually at the different reservoirs. species were abundant, while the rest were of
mm
Fingerling rainbow (76 total length [TL]) minoroccurrence.
were stocked at waters where numbers of A number of the waters in which Utah
competing species were low and predation chubs and Utah suckers {Catastomus ardens)
was not a concern. Larger rainbow (127-178 occurred were periodically treated with
mm TL) were stocked at reservoirs where sur- rotenone to remove all fish when those non-
vival ofsmall trout was poor because ofcom- game species became abundant. When recla-
petition with nongame species and/or preda- mation projects were conducted, chubs and
tion. Limited numbers of other species of suckers often outnumbered trout by hundreds
trout were stocked at some waters to provide to one. Following treatments, trout were the
Table 1. Description ofwaters in southwestern Utah where scheduled counts ofDouble-crested Cormorants
wereconducted, 1989-91.
274 Great Basin Naturalist [Volume54
predominant species for at least a \'ear or two. fall. The number ofbird-days for a stratum at
In situations where undesirable nongame a given water was estimated using the follow-
species of fish could not he completely ing formula:
removed from a drainage, Utah chubs and
Utah suckers would gradually increase and
D = K(—^)
eventualK' return to pre-reclamation densities.
n
In addition to cormorants, otherpiscivorous
birds observed at the study waters included
Common Loons {Gavia immer). Western = n
Var (D)
Grebes {AechmopJiorousoccidcniulis), American (n-1) n
White Pelicans {Pelecanus crfhrorychos).
Mergansers {Mergus merganser and M. serra- /h
tor), and Great Blue Herons {Ardea herodias D = estimated total bird-days;
treganzai). K = numberofdays within asampling
stratum;
Methods oCj = numberofcormorants counted on the
ith day;
Counts ofcormorants were made at 1- to 3- n = numberofdays sampled within a
week intervals at 11 resei^voirs during 1989. In stratum;
1990 we made biweekly counts at four of the Var = estimated variance;
larger reservoirs. Counts were made again in N = total cormorants counted per stratum.
1991 at the four reservoirs surveyed in 1990,
as well as three additional ones. Counts gener- 95%confidence intei"val = ± 2 yVar (D).
ally began following ice-out at each reservoir
and continued through November at most The estimate of annual cormorant abundance
waters. We discontinued counts early at sever- was the sum of estimated bird-days for strata
al reservoirs that were drained during the within a sampling year divided by the mean
summer or chemically treated to remove surface areaofthe reservoir.
nongame fish. At most locations counts were Gill-nets were used to estimate trout abun-
made from shore using binoculars or a spot- dance at each reservoir (Bennett 1962, Hubert
ting scope. At larger resei"voirs we often used 1983) during early spring, 2-4 weeks after
a boat to facilitate counting. Technicians mak- winter ice cover was completely gone. Net
ing counts were instructed using a standard numbers, styles, and locations were based on
training program by the authors. The same long sampling histories at each water. Gill-net
one or two technicians counted birds at all data have been collected on most ofthe study
waters during any one year of the study. waters for 10 years or more. We followed stan-
Cormorants were easily identified. Knowledge dardized netting practices used by the Utah
of the birds' feeding and resting patterns, as Division ofWildlife Resources (UDWR). Two
well as other behaviors, also aided in making to six nets, depending on lake or reservoir
accurate counts. size, were set at each water in areas less than
An annual estimate of cormorant abun- 30 ft deep. Nets were set during the afternoon
dance (bird-days per ha per year) was made and retrieved the following morning. Each net
for each reservoir studied. The estimate of was 1.8 m deep by 38.1 m long and consisted
abundance was calculated using methods of five monofilament nvlon panels with bar
commonly employed to estimate sportfishing mesh sizes of 19.1, 24.4, 31.8, 38.1, and 50.8
pressure in creel surveys of anglers (Robson mm.
1960, Lambou 1961). A bird-day was defined Data recorded for fish gill-netted at each
as one day spent by one cormorant at a given water included numbers, species, and individ-
water. The sampling period was stratified by ual lengths. Gill-net samples generally consist-
3-month intei^vals, March-May, June-August, ed of trout stocked the previous year and a
and September-November. The number of few from stocking 2 years earlier. The trout
days within a stratum varied among waters, abundance index used for each reservoir in
depending on the time ofice-out and whether the study was the mean number of trout col-
agiven reservoirwas drained or treated in the lected per net, set overnight (trout per net-
1994] CormorantObservations in Southwestern Utah 275
night). When comparing trout abundance and was obtained from six fledgling cormorants at
estimates ofcormorant abundance, we paired Piute Reservoir in 1989 by approaching active
the trout abundance index for a given water nests and collecting regurgitated stomach con-
with the estimate ofcormorant abundance for tents. In total, diet datawere obtained from 52
the previous year. Because spring gill-net cormorants, with samples taken in mid-April,
catches consisted primarily of trout stocked late April, late June, late July, early August,
the previous year, the relationship between and lateAugust.
the cormorant abundance estimate and trout Stomach contents were identified to fish
abundance index reflected impacts of preda- species using flesh color, peritoneum color, fin
tion on one cohort of stocked trout over one rays, and pharyngeal teeth as key characteris-
year Large trout were excluded from the data tics. We made TL measurements of ingested
at two waters when calculating the trout abun- fish when possible. TL estimates were also
dance index. These larger fish represented based on a measurement from the front ofthe
older cohorts that were not vulnerable to cor- dorsal fin to the front ofthe anal fin. Estimates
morant predation during the study period. ofbiomass of ingested fish were made using
Large trout occurred at Minersville Reservoir length-weight relationships for each species
and Fish Lake as the result ofunusual circum- (Carlander 1969, Varleyand Livesay 1976).
stances or the presence ofunique trout popu- Annual trout consumption by cormorants
lations. At Fish Lake a few large lake trout was estimated by multiplying values for bird
{Salvelinus namaijacush) were not used in the abundance (bird-days) by a daily biomass con-
index. One cohortofcutthroat trout {Oncorhyn- sumption rate of465 gper day (afterWasowicz
chus clarki) at Minersville Reservoir was not 1991), and by the percentage of trout in the
used in the trout abundance index for that diet (this study, Wasowicz 1991). The daily
reservoir. This 1986 cohort grew rapidly to a biomass consumption rate used by Wasowicz
large size following a chemical renovation in (1991) and this study was based on an average
1985 and comprised a substantial portion of adult body weight for cormorants of 1860 g
the annual spring gill-net catches through (Ross 1977) and a daily biomass consumption
1991. A simple linear regression was used to rate of 25% of body weight. Dunn (1975)
compare estimates of cormorant abundance reported that daily consumption rates for free-
(bird-days) and the trout abundance index livingadults andjuveniles ofseveral species of
(troutper net-night) usingboth untransformed cormorants averaged approximately20-30% of
dataand log-transformed data. body weight. When information on diet com-
Data on cormorant diet were collected at position ofcormorants was not available for a
three large, lower-elevation reservoirs where particular water, we made a conservative esti-
birds were relatively abundant. At Minersville mate ofthe percentage oftrout in the diet by
and Otter Creek resei'voirs, primaiy potential determining relative abundance of trout and
fish prey species were stocked rainbow trout otherforage species in thatwater.
and Utah chubs, with lesser numbers of cut- Season-long creel surveys of sport fisher-
throat trout and brown trout {Salmo trutta). At men and chemical treatment projects to
Piute Reservoirprimaiy prey species included remove undesirable nongame fish were con-
rainbow trout, Utah chubs, and Utah suckers. ducted at a number ofstudy waters. Although
Piute Reservoir also contained limited num- not directly related to this study, data collect-
bers ofredside shiners {Richardsonius haltea- ed during these activities provided a means to
tiis), smallmouth bass {Microptenis dolomieiii), validate trout abundance indices and verify
cutthroat trout, and brown trout. We collected relative abundance of different fish species.
10 cormorants each at Minersville, Otter We estimated total annual trout harvest by
Creek, and Piute reservoirs (30 birds total). anglers and the percent return to the creel of
Birds were collected using shotguns during the total numbers of fish stocked (Robson
July and August at 1000-1100 h following 1960, Lambou 1961) during creel surveys.
morning feeding periods. We also used food- High and low harvest estimates corresponded
habit data collected by Wasowicz (1991) at widi higli and lowtroutabundance as measured
Minersville Reservoir in April 1988, which by standardized gill-netting. Visual inspections
included cormorants collected during after- following chemical treatments provided anoth-
noon hours. Additional food-habit information er way ofverifying relative fish abundance and
276 Great Basin Naturalist [Volume54
species composition. Following a chemical 26 June 1989, 45 fledgling cormorants were
treatment, we could be certain that stocked observed in nests in flooded cottonwood trees
trout dominated a fishery for a year or two. in the south end ofthat reservoir. In 1990, 55
Creel surveys were conducted at Fish Lake in pairs of nesting birds were observed in the
1989, Johnson Reservoir in 1984 and 1989, same area on 11 April. Young cormorants were
Kolob Reservoir in 1991, Lower Bowns observed in 16 ofthe nests on 26 May 1990, in
Reservoir in 1991, Minersville Reservoir in spite ofrapidly dropping water levels that had
1986 and 1988, Newcastle Reservoir in 1991, left nesting trees well above the shoreline.
and Otter Creek Reservoir in 1985. Chemical Piute Reservoirwas drained in the fall of1990
treatments were conducted at Johnson causing water levels to remain low in 1991
Reservoir in 1986, Kolob Reservoir in 1985, and exposing the ground below trees used for
Koosharem Reservoir in 1985, Minersville nesting the previous 2 years. No nesting activ-
Reservoir in 1984 and 1991, Otter Creek ity was observed at any of the locations stud-
Reservoir in 1989, Panguitch Lake in 1991, ied in 1991.
and Piute Reservoirin 1985 and 1990. Estimates ofcormorant abundance atthe 13
reservoirs ranged from bird-days at Lower
Results Bowns Reservoir in 1991 to 20,329 bird-days
at Otter Creek Reservoir in 1989 (Tables 2 and
Cormorant Distribution andAbundance 3). When we accounted for the size ofvarious
Eight to 35 counts were made at each of waters surveyed, cormorant abundance was
the 13 reservoirs (Table 2). Individual counts highest at Minersville Reservoir where we
of cormorants ranged from to 264 birds. estimated 34 bird-days per ha for 1989 (Table
Cormorants were observed early in the year (2 4). Cormorant abundance was low at most of
February 1989) at Quail Creek Reservoir, the higher-elevation waters, such as Kolob
which was the lowest in elevation and most Reservoir,Johnson Reservoir, and Fish Lake.
southern reservoir studied. At most other TroutAbundance
waters, cormorants were first observed soon
after ice-out, usually in March. Numbers of Stocking rates ranged from 186 to 669 trout
cormorants were generally highest in spring per ha per year at the waters studied, except
or early summer. At lower-elevation waters, at Gunlock Reservoir, which was managed
cormorants were often absent during midsum- only forwarm-water species. In general, num-
mer but were observed again in late summer bers and sizes of trout stocked at each reser-
or fall. At some higher-elevation waters, high- voir or lake were considered sufficient to pro-
est counts occurred in midsummer They were duce high numbers of catchable-size trout
present throughout the summer at several of providing that survival was adequate. Trout
the larger reservoirs. Cormorants were abundance indices at the waters surveyed
observed at all waters surveyed except one. ranged from 1 to 91 trout per net-night (Table
Lower Bowns Reservoir, the smallest and 4). Our past experience indicates that trout
most easterly located. abundance indices ofat least 25-30 fish per
Cormorants attempted to nest at 2 ofthe 13 net-night yield a population of trout that will
locations studied. In 1988 and 1989 nesting produce good fishing during the year.
was initiated at Minersville Reservoir. Rainbow trout accounted for the majority of
Cormorants constructed nests in a flooded the gill-net catch at most waters. The trout
grove ofCottonwood trees in the shallow north abundance index was inversely related to esti-
end of the reservoir. The nests were aban- mates of cormorant abundance (P < .01, Fig.
doned, however, in late springwhen the water 1). Although alog transformation ofcormorant
level receded beyond the nesting trees. Water abundance data statistically improved the fitof
levels at Minersville Reservoir remained low the regression line, the negative relationship
during the spring of 1990 and 1991. The area was also significant [P < .05) for the original,
in which nesting had been attempted the pre- untransformed data. Trout abundance indices
vious 2 years remained some distance above were low when bird abundance was greater
the shoreline, and cormorants made no further than 15 cormorant-days per ha. Both high and
attempts to nest. Cormorants did nest success- low trout abundance indices occurred with
fully at Piute Reservoir in 1989 and 1990. On low cormorant abundance; however, there
1994] Cormorant Observations in Southwestern Utah 277
Table2. Statisticsfiomcormorantcountsat 13reservoirsinsouthwesternUtah, 1989-91.
278 Great Basin Natufl\list [Volume 54
Table2.Continued.
Ti )f\far
WaterAVar/Statistic Mar-\la\ Jun-Aug .St'p-Xo\' Total
Piute Reservoir, 1989
Totaldaysinintenid 82
Numberolcoiuits
Meanbirdspercount
Estimatedbird-days
Standarderror(]:)ird-days)
95%confidenceinterval
PiuteReser\'oir, 1990
Totaldaysininterval
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceinterval
PiuteReservoir, 1991
Totaldaysininter\'al
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceintei^val
Fisli Lake, 1989
Totaldaysininterval
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceintei'val
Panguitch Reservoir, 1989
Totaldaysinintei"val
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceinterval
Panguitch Reservoir, 1990
Totaldaysinintei"val
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceinterval
PanguitchReservoir, 1991
Totaldaysininterval
Numberofcounts
Meanbirdspercount
Estimatedbird-days
Standarderror(bird-days)
95%confidenceinterval
1994] CormorantObservations in Southwestern Utah 279
Table2.Continued.
280 Great Basin Naturalist [Volume54
Table3. Estimatedannualc()n.sinnpti(jn offishl)ycormorantsat 13reservoirs insouthwestern Utah, 1989-91.
1994] CormorantObservations in Southwestern Utah 281
Table 4. Estimates ofcoiTnorant abundance compared to trout abundance indices at southwestern Utah reservoirs,
1989-92.