9697

• UCREFRP Catalog Number: 9697
• Author: Bestgen, K.R., D.W. Beyers, J.A. Rice and G.B. Haines.
• Title: Factors affecting recruitment of young Colorado pikeminnow
• USFW Year: 2006.
• USFW - Doc Type: synthesis of predation experiments, field studies, and individual-based modeling.
• Copyright Material: YES

1726

levels: 38, 47, 53, 61, 63, and 69 rom. The second
factor, prey size (mean 11.), also had six levels: 9.8,
11.1, 13.6, 15.3, 17.2, and 18.7 rom. We completed
three trials for each size-class of prey per each size-
group of predator (or 3 X 36 = 108 trials), plus 12
additional trials for the 11. I-rom prey size. In each trial,
five pikeminnow larvae of the same size category were
introduced into a 38-L glass aquarium containing three
red shiners from one of the six different predator size-
groups; subsequent trials were completed such that
each larval size was separately tested with each
predator size. Each predator was watched by a separate
observer, and the number of attacks and successful
captures during two consecutive lO-rnin intervals were
recorded. If three or more larvae were captured during
a trial, an additional five larvae were added to the
aquarium. Predators were assumed to act independently
because no cooperation was observed between red
shiners when capturing Colorado pikeminnow larvae.
Logistic regression analysis (SAS 1993) was used to
estimate the probability of capture as a function of
predator 11., prey 11., predator 11.2, prey 11.2, and the
predator-prey 11. interaction (pPI). A general linear
model was fit using the method of maximum
likelihood, and Akaike's corrected information criteri-
on (AlCc; Akaike 1981; Burnham and Anderson 1998)
was used to identify a reduced model that best
explained variation in probability of capture. Model
selection suggested that the best predictors of proba-
bility of capture, P(capture), were prey 11.2 and the PPI
term. The form of the size-dependent capture equation
was

10gitP(capture) = -2.26 - 0.0136(prey 11.)2
+ 0.0036(PPI), (1)

which can be transformed to probability of capture as
follows:

P(capture) = elogitP(capture) /(1 + elOgitPlcaptureJ). (2)

Attack rate.-Direct estimation of the rate of
encounter and the rate of attack given an encounter
requires extensive data about the behavior and physical
abilities of the species of interest (Gerritsen and
Strickler 1977; Bailey and Batty 1983; Fuiman and
Gamble 1989). As an alternative to gathering that
information, we used a methodology that combined the
probabilities of encounter and attack into a single
paranaeter that estimated the number of attacks per day
on individual larvae (hereafter called attack rate) as a
function of predator length, Colorado pikeminnow
larva length, and environmental conditions that includ-
ed presence or absence of both alternative prey and
turbidity. DaIa for estimating attack rate were obtained

BESTGEN ET AL.

from outdoor mesocosm experiments. Mesocosms
were intended to mimic characteristics of backwaters
typically occupied by Colorado pikeminnow larvae in
Green River nursery habiIat (Tyus and Haines 1991).
Plastic wading pools (diameter, 1.5 m) were positioned
on a sloping surface so that water depth in each
mesocosm ranged from 0 to 25 cm. Pools were filled
with well water and substrate was 1-2 em of washed
sand. Mesocosms had a water surface area of 1.48 m2
and contained about 225 L of water. A factorial design
similar to that described for estimating probability of
capture was used, except that the 38-rom 11. predator
group was not included because observations showed
that capture rate was very low. A total of 69 individual
mesocosm trials were completed over all size-groups of
larvae (mean prey 11. = 9.8-20.1 mm) and conditions;
we tested effects of turbidity and alternative prey with
several prey size-classes. In each trial, three red shiners
from each size-group were introduced into pools and
allowed to acclimate overnight Colorado pikeminnow
larvae were floated in plastic bags in mesocosms and
allowed to acclimate for approximately 30 min before
being released at the start of trials. Standard trials were
conducted under daylight conditions for 6 h using clear
water and no alternative prey, conditions intended to
maximize predation by red shiners. Additional trials
were conducted using environmental conditions hy-
pothesized to reduce predation by red shiners: turbid
water, and presence of alternative prey (chironomid
larvae). Mesocosm turbidity was manipulated by
adding a suspension obtained from mixing unwashed
sand and water until a Secchi depth of about 5 em was
achieved. Observations showed that water turbidity
remained high for the duration of the trials. Chirono-
mid larvae were used as alternative prey because they
are known prey items of early life stages of Colorado
pikeminnow in the wild (Muth and Snyder 1995).
Preliminary studies showed that maximum consump-
tion of chironomid larvae by red shiners over a 6-h
period was about 10% of body mass. To avoid predator
satiation, the total mass of prey provided (Colorado
pikeminnow larvae or chironomids) never exceeded
10% of combined predator mass. Consequently, the
number of pikeminnow larvae used in each trial varied
as a function of predator mass. and ranged from 227 for
the smallest larvae to 15 for the largest The number of
chironomid larvae ranged from 25 to 100, with an
equal mass of Colorado pikeminnow larvae in each
treatment Water temperature during trials was usually
25-310C. At the conclusion of a trial, a standardized
protocol was used to remove predators and remaining
prey from pools. Five preliminary tests showed that the
protocol recovered 100% of a known number of 10-
mrn-11. Colorado pikeminnow or chironomid larvae.