9663

• UCREFRP Catalog Number: 9663
• Author: Johnson, B. M. and e. al.
• Title: Ranking Predatory Threats by Nonnative Fishes in the Yampa River, Colorado, via Bioenergetics Modeling
• USFW Year: 2008
• USFW - Doc Type: North American Journal of Fisheries Management
• Copyright Material: YES

NONNATIVE FISH PREDATION THREAT
were converted to wet mass using family- or order-
specific functions .derived from the literature (Smock
1980; Burgherr and Meyer 1997; Benke et aI. 1999).
Crayfish cazapace lengths were also converted to wet
mass (Roell and Orth 1992). Species-specific functions
were used to convert fish vertebral column length to
wet mass (B.M.J., unpublished data), which allowed us
to compute diet composition on a wet-mass basis.
We estimated predator consumptive demand by
performing simulations with Fish Bioenergetics soft-
ware (Hanson et al. 1997). A revised parameter set for
smallmouth bass was derived from Whitledge et al.
(2003). Parameters for channel catfish simulations were
obtained by adjusting temperature-dependent physio-
logical inputs reported for flathead catfish Pylodictus
olivaris (Roell and Orth 1993) to approximate the
thermal preferences of channel catfish (Becker 1983),
as described by Hanson et al. (1997). Northern pike
parameters were not changed from Fish Bioenergetics
defaults. Per-capita consumption was simulated for the
average adult of each population and was computed
from the annual growth increment, predator diet, river
temperature, predator energy density, and prey energy
density. Growth increment used in simulations was
calculated from the geometric mean weight of the fish
collected in the mark-recapture samples and the von
Bertalanffy growth function fitted to weight at age.
Each simulation encompassed 1 year. The thermal
experience of each species was estimated from mean
temperatures of the Yampa River recorded at the
Maybell gauge station (U.S. Geological Survey,
Station 09251000; RKM 126) during 1996-2002.
Energy density was set at 3.6 kJ/g of wet weight for
northem pike and at 4.2 kJ/g for smallmouth bass and
channel catfish (Hanson et al. 1997). Energy lost to
spawning (smallmouth bass: 7% loss on 20 May;
northem pike: 10% on 15 March; channel catfish: 7%
on 1 July) was incorporated into the simulations.
Literature-based estimates of energy density (wet-mass
basis) were obtained for aquatic insects (4.3 kJ/g:
Cummins and Wuycheck 1971), fish prey (4.2 kJ/g:
Hanson et al. 1997), and crayfish (3.8 kJ/g: Roell and
Orth 1993).
We performed simulations using two diet scenarios.
The nominal run (realized piscivory) used the diet
information we collected from smallmouth bass,
northern pike, and channel catfish in the Yampa River
during 2003-2005, when the availability of small-
bodied fish prey was low. The second set of
simulations represented potential consumptive demand
before the observed decline of small-bodied fishes
(potential piscivory scenario). For these simulations,
we used northem pike diet information (90% fish, 10%
invertebrates) reported by Nesler (1995). Diet infor-
1945
mation was not available for smallmouth bass before
2003; diet composition for smallmouth bass collected
from GVR was used to represent the Yampa River diet
as if small-bodied fishes had not already been depleted.
We assumed that this set of simulations represented the
latent piscivory within the piscivore populations and
was an indicator of their potential to prevent the
recovery of small-bodied fishes via predation. Potential
consumption by channel catfish was not computed,
because the incidence of fish in the diet was low during
the late 1980s (Tyus and Nikirk 1990).
Annual per-capita consumption (c) by each predator
species (a) was scaled up to consumption by the entire
population (B) based on the mark-recapture abundance
estimate (1V~) and its confidence limits:
B; _ (c; X 1V,)±(to.os X c; x SEN,).
We apportioned the estimated biomass of fish
consumed by smallmouth bass and northem pike into
small-bodied fish equivalents (SBFs) to evaluate the
intensity of predation on native fish populations on a
numerical basis. The SBFs were computed for eight
native prey fishes: the bluehead sucker, flannelmouth
sucker, razorback sucker, humpback chub, Colorado
pikeminnow, roundtail chub, mottled sculpin, and
speckled dace. The number of prey of each species
consumed was computed from prey mean weight at a
specified size and the total biomass consumed per
predator population:
SBFrj = Bi ~ ~~7 X (Pi X pa )~' ~ e
where i is the predator species, j is the prey species, B
is fish biomass consumed per year, P is predator size
(mean TL, mm), p is the median prey :predator size
ratio observed in predator guts, and a and b are
coefficients of prey length-weight regressions (Caz-
lander 1969; Didenko and Bonaz 2004). The 5th and
95th percentiles of the prey :predator size ratio (p5 and
p9s) were used to compute a range of small-bodied fish
that could reasonably occur with changes in the size
structure of the extant prey assemblage. The number of
age-1 fish consumed was computed by dividing
consumption by prey weight at age (Bailey 1952;
Vanicek and Kramer 1969; Minckley 1983; McAda
and Wydoski 1985; Osmundson et al. 1997; Robinson
and Childs 2001;). The maximum age of each prey
species that was vulnerable to the smallmouth bass and
northern pike populations was computed from prey
length at age and predator gape limits (60% of predator
length for smallmouth bass: Katano and Aonuma
2001; 50% for northern pike: Mittelbach and Person
1998).