RIVER RESEARCH AND APPLICATIONS
River Res. Applic. (2014)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/rra.2768
EFFICACY OF MECHANICALLY REMOVING NONNATIVE PREDATORS FROM A
DESERT STREAM
D. L. PROPSTa*, K. B. GIDOb, J. E. WHITNEYb, E. I. GILBERTc, T. J. PILGERa, A. M. MONIÉc, Y. M. PAROZd,
J. M. WICKc, J. A. MONZINGOe AND D. M. MYERSf
a
Museum of Southwestern Biology, University of New Mexico, Albuquerque, New Mexico, USA
b
Division of Biology, Kansas State University, Manhattan, Kansas, USA
c
Conservation Services Division, New Mexico Department of Game & Fish, Santa Fe, New Mexico, USA
d
US Forest Service, Southwest Region, Albuquerque, New Mexico, USA
e
Gila National Forest, Silver City, New Mexico, USA
f
New Mexico Fish & Wildlife Conservation Office, US Fish & Wildlife Service, Albuquerque, New Mexico, USA
ABSTRACT
Native fish faunas throughout the American Southwest have declined dramatically in the past century, mainly a consequence of habitat alteration
and alien species introductions. We initiated this 6-year study to evaluate the efficacy of mechanical removal of nonnative predaceous rainbow
trout Oncorhynchus mykiss, brown trout Salmo trutta, yellow bullhead Ameiurus natalis and smallmouth bass Micropterus dolomieu from an open
4.6-km reach of West Fork Gila River in southwest New Mexico, USA. Removal efforts involved intensive sampling with a 10- to 12-person crew
using backpack electrofishers and seines to capture fish over a 4- to 5-day period each year. Additionally, two reference sites were sampled with
similar methods to compare temporal changes in species mass in the absence of a removal effort. Results were mixed. Mass of yellow bullhead,
rainbow trout and brown trout declined in the removal reach from 2007 through 2012, but there was no change in smallmouth bass. Concurrently,
mass of Rainbow trout, yellow bullhead and smallmouth bass did not change at reference sites, but brown trout mass declined, indicating factors
other than removal were driving abundance of brown trout. Occurrence of several large flathead catfish Pylodictis olivaris in the removal reach
in 2012 changed what would have been a decline in overall nonnative mass to no change over the course of the study. Spikedace Meda fulgida
was the only native species positively responding to predator removal. Results of this study suggest that with moderate effort and resources applied
systematically, mechanical removal can benefit some native fish species, but movement of problem species from surrounding areas into removal
reaches necessitates continued control efforts. Copyright © 2014 John Wiley & Sons, Ltd.
key words: conservation; native fishes; nonnative predator control; Gila River; USA
Received 1 November 2013; Revised 1 April 2014; Accepted 15 April 2014
INTRODUCTION
Globally, indigenous fish faunas are threatened by an array of
human-caused activities (Dudgeon et al., 2006; Vorosmarty
et al., 2010). In addition to changes wrought by physical
modifications, introduction of alien species has further
challenged the persistence of native fishes (Clarkson et al.,
2005; Cucherousset and Olden, 2011). In most systems
where native fish faunas are imperilled, both habitat alteration
and introduced species are problematic (Hoagstrom et al.,
2011). The relative importance of habitat alteration as
opposed to nonnatives on persistence of native fishes varies
across systems and is dependent upon species involved, the
degree of alteration (i.e. extent, timing and severity) and the
traits of the native fauna (Olden et al., 2008; Gido et al.,
2013). For example, large-bodied migratory natives may be
*Correspondence to: D. L. Propst, Museum of Southwestern Biology,
University of New Mexico, Albuquerque, New Mexico 87131, USA.
E-mail: tiaroga@comcast.net
Copyright © 2014 John Wiley & Sons, Ltd.
less affected by nonnative predators than small short-lived
natives but highly vulnerable to habitat fragmentation.
Because of high water demand and low fish species
diversity in the arid American Southwest, the extent of
habitat alteration and species introductions is perhaps more
evident, if not more severe, than in mesic regions. Many
desert streams are seasonally dry throughout substantial
lengths (Blinn and Poff, 2005). Reaches that retain surface
flow often do so by reservoir releases that are not reflective
of pre-regulation flow regimes. Many streams are no longer
connected to their floodplains and thus are deprived of
essential nutrient infusions. The combined pressures of
habitat modification and nonnative organisms have left no
intact native fish faunas at the drainage scale and few at a
local scale of small tributary stream or spring system
(Hendrickson and Minckley, 1984; Fagan et al., 2005;
Minckley and Marsh, 2009). Even where physical alterations
are modest and comparatively few nonnative fishes occur,
survival of native fish faunas is not assured (Propst et al.,
2008; Pool and Olden 2014). Added to these more immediate
D. L. PROPST ET AL.
pressures are those of climate change, which will cause
increasing aridity and altered precipitation patterns across
the American Southwest (Seager et al., 2007).
Numerous approaches have been used to diminish or
reverse the effects of habitat alteration and nonnative
organisms for the benefit of native fish assemblages (Gozlan
et al., 2010). Some are rather broad in scope in that they
attempt to correct or ameliorate what is wrong with the
entire system and not focus on single elements of a larger
problem (Fausch et al., 2002). Such efforts might involve
mimicking a natural flow regime via reservoir releases in
concert with removal of nonnative fishes, as is occurring
in the Yampa-Green rivers of Colorado and Utah or San
Juan River of New Mexico, Colorado, and Utah (Bestgen
et al., 2007; Franssen et al. 2014). Realizing it is virtually
impossible to eradicate nonnative fishes or restore natural
flow regimes, these efforts are designed to tip the balance
back towards native fish faunas and make conditions less
suitable for undesirable nonnative fishes. Other efforts
involve using piscicides (rotenone or antimycin) to remove
nonnative fishes from a stream, constructing a barrier(s) to
prevent re-invasion of nonnative fishes and repatriating
native fishes to the renovated reach, as was carried out in
Fossil Creek, Arizona (Weedman et al., 2005). In many
systems, the means to eliminate or substantially reduce
problem species have been limited, for numerous practical
reasons, to non-chemical removal tactics, such as with
electrofishing gear (e.g. Kulp and Moore, 2000). Regardless
of removal method, achievement of management objectives
has been mixed (Meronek et al., 1996).
Eradication—through piscicides—of problem nonnative
fishes from stream reaches of the upper Gila River in
southwest New Mexico that supports an intact but
vulnerable native fish fauna was neither practical nor likely
because of high flow volume and environmental regulatory
permitting requirements. Thus, we tested if abundance of
nonnative fishes could be diminished using electrofishing
and seining and if a measureable response by the native fish
assemblage could be documented. We predicted that largebodied nonnative fishes would be most vulnerable to these
sampling methods and would decline over the study period.
However, we expected juveniles to continually recruit into
the reach and expected movement of individuals from
surrounding areas into the removal reach, albeit the rate
and extent at which this would occur was unknown. We also
predicted the most notable response by native fishes would
be from small-bodied species that typically co-occur in
low-velocity habitats with nonnative predators (Stefferud
et al., 2011).
A key requirement of our approach was that our effort be
equivalent to what cooperating resource agencies could
invest on an annual basis for an indefinite period. Thus,
we present estimates of project costs. Although it might be
Copyright © 2014 John Wiley & Sons, Ltd.
theoretically possible to quantitatively demonstrate that
removal of nonnative fishes would yield a desirable
outcome, this was of little practical importance if the effort
required was more than the typical resource agency(ies)
could invest.
METHODS
Study area
Nonnative fishes were removed from a 4.6-km reach of
lower West Fork Gila River in southwest New Mexico,
USA, within the New Mexico Department of Game and Fish
Heart Bar Wildlife Management Area (Heart Bar removal
reach, Figure 1). The Middle Fork (Gila River) was
confluent with the West Fork about 150 m above the
upstream removal reach terminus, and the East Fork joined
the West Fork 3.9 km below the downstream removal reach
terminus. Little Creek was the only perennial tributary
confluent to the river within the removal reach and marked
the downstream removal reach terminus. The Middle Fork
contributed considerably more to discharge within the Heart
Bar removal reach than did the West Fork (Table I). Within
the removal reach, the river meandered through a comparatively narrow floodplain vegetated largely by narrow-leaf
cottonwood Populus augustifolia, coyote willow Salix
exigua and Goodding willow Salix gooddingii. Surrounding
uplands were almost entirely within US Forest Service Gila
National Forest and New Mexico Department of Game and
Fish Heart Bar Wildlife Management Area. Domestic
livestock grazed adjacent uplands in the past, but none has
occurred for over 30 years. Although open to sport fishing,
anglers were rarely observed in the removal reach during
removal efforts. There were no impediments or barriers
to fish movement into or out of the 4.6-km Heart Bar
removal reach.
During early June, habitats within the removal reach
ranged from deep (1 to 2.5 m), slow-velocity (<0.1 m s 1)
pools associated with standing and uprooted cottonwoods,
scour pools (1 to 2 m deep with 0.5 m s 1 velocity water)
along cliff faces, moderate-velocity runs (0.25 to 0.5 m s 1
and 0.3 to 0.75 m deep), cobbled riffles (0.1 to 0.25 m deep
and 0.35 to 1.0 m s 1 velocity water), and shallow
(<0.1 m deep) to deep (1.0 m) backwaters. Vegetated
undercut banks were common. Mid-day water temperatures in early June were about 22°C, clarity was typically
high (>0.5 m), dissolved oxygen about 8 mg L 1, specific
conductance about 150 μmho cm 1 and pH slightly
alkaline (<8.0). Discharge was less than 1.0 m3 s 1 during
removal efforts.
Reference sites were located on the Middle (site length = 297
m) and West (site length = 234 m) forks about 1 and 2 km
upstream, respectively, of the upstream terminus of the removal
River Res. Applic. (2014)
DOI: 10.1002/rra
NONNATIVE PREDATOR REMOVAL
Figure 1. Location of Heart Bar nonnative fishes removal reach and reference sites in the upper Gila River, Catron County, New Mexico, USA
reach (Figure 1). The variety and general dimensions of habitats
within reference sites were similar to those in the removal
reach. Reference site sampling was initiated in 2008. Each
reference site had fewer species than the removal reach, but
collectively, they had all species collected in the removal reach,
except that loach minnow Tiaroga cobitis, although found prior
to this study, was not collected from either reference site.
Habitats in reference sites and the removal reach were similar,
Table I. Temperature and discharge at West Fork Gila River Heart Bar nonnative fishes removal reach and West Fork
Gila River and Middle Fork Gila River reference sites, June 2008–2012
Heart Bar
Temp (C°)
2008
2009
2010
2011
2012
West Fork
Middle Fork
Discharge (m3 s 1)
Temp (C°)
Discharge (m3 s 1)
Temp (C°)
Discharge (m3 s 1)
0.13
0.21
18.7
18.3
18.9
19.0
21.0
0.23
0.20
0.22
0.06
0.08
23.6
24.1
24.8
24.5
22.8
0.51
0.25
0.40
0.10
0.19
21.1
19.0
23.2
Temperature for West Fork and Middle Fork obtained at 1200 h on 24 June of each year. Discharge measurements made at time of sampling at each location
Copyright © 2014 John Wiley & Sons, Ltd.
River Res. Applic. (2014)
DOI: 10.1002/rra
D. L. PROPST ET AL.
but water temperature in the West Fork was usually about 5°C
less than the Middle Fork or Heart Bar reach. No other
accessible stream reaches having comparable fish assemblages
were within reasonable proximity (ca. 50 km) of the removal
reach.
Sampling methods
Removal of nonnative fishes and inventory of native fishes
in the Heart Bar removal reach was carried out by two 5- to
6-person crews over a 4- to 5-day period in early June each
year, 2007 through 2012. Sampling occurred in June
because snowmelt runoff had ceased, and summer monsoons had not commenced; thus, discharge was at or near
its annual base, and sampling efficiency was expected to
be highest. Sampling at reference sites (one 3-person crew)
occurred within 10 days± of removal reach collections. All
fishes captured at reference sites were returned alive to
habitat of capture. All habitats were sampled using
battery-powered backpack electrofishers and 3.0 and
4.6 × 1.2 m, 3.2-mm mesh seines. The specific method used
depended upon mesohabitat being sampled and was
selected to maximize probability of capture of species
likely to occupy that mesohabitat. All captured fishes were
held in 18.9-L buckets equipped with aerators. Total length
(TL, mm) and mass (g) were determined for all nonnative
predators and large-bodied native fishes >100 mm TL. A
subsample (ca. 50 per species) of large-bodied specimens
<100 mm, all small-bodied (< 130 mm maximum TL)
native and nonnative fishes, were also measured. Large-bodied
fishes (adults > 130 mm TL) were categorized as juvenile,
sub-adult and adult on the basis of length–frequency histograms
of specimens of each species collected within the study area
(Pilger et al., 2010). Length–mass relationships were calculated,
using the general formula, mass = aTLb. Each species relationship was used to estimate mass of un-weighed specimens. In
the removal reach, native fishes were returned to habitat of
capture, and all nonnative fishes were retained. Response metrics were number of specimens and mass of specimens per species, and age class (large-bodied species) per 10 m of stream.
To obtain a general estimate of capture efficiency, twopass closed population capture–recapture was conducted
for each species during June from 2008 through 2011 at
the Middle Fork reference site. A 3-person crew captured
fishes with a battery-powered backpack electrofisher and a
seine (4.6 × 1.2 m, 3.2 mm mesh). Efficiency was calculated
as the mass of individuals of a species captured in the
first pass divided by the total mass estimate (population
estimate × mean mass of captured specimens) for that
species. The mean for each species was used to evaluate
nonnative capture efficiency in the removal reach.
Data analysis
Pearson’s correlation coefficient was calculated to evaluate
changes in mass (g/10 m) of each species over time. Because
catostomids <30 mm TL could not be reliably field
identified, Sonora sucker and desert sucker specimens
Table II. Fishes captured in West Fork Gila River Heart Bar nonnative removal reach, 2007–2012, Catron County, New Mexico
Year
Species
Longfin Dace
Headwater chub
Spikedace
Speckled dace
Loach minnow
Desert sucker
Sonora sucker
Gila trout
Proportion native
Common carp
Red shiner
Fathead minnow
Yellow bullhead
Flathead catfish
Western mosquitofish
Rainbow trout
Brown trout
Green sunfish
Smallmouth bass
Proportion nonnative
2007
2008
2009
2010
2011
2012
115
38
—
17
1
263
511
—
0.81
—
—
—
99
—
15
48
36
1
24
0.19
207
46
27
59
8
360
641
13
0.92
—
—
1
30
—
—
14
62
—
8
0.08
3444
518
103
566
50
1427
5328
13
0.98
—
—
—
118
1
4
47
73
1
29
0.02
636
94
65
122
6
196
954
1
0.96
—
2
16
42
—
—
—
10
5
2
0.04
1821
105
881
984
99
901
1933
1
0.96
—
1
62
142
1
4
20
8
1
37
0.04
675
18
138
237
20
296
899
4
0.92
2
2
45
17
11
90
1
6
—
13
0.08
Counts include all catostomid specimens ≥30 mm total length.
Copyright © 2014 John Wiley & Sons, Ltd.
River Res. Applic. (2014)
DOI: 10.1002/rra
NONNATIVE PREDATOR REMOVAL
Table III. Mass (kg) of fishes collected in West Fork Gila River Heart Bar nonnative removal reach (length = 4600 m), Catron County, New
Mexico, 2007–2012
Year
Species
2007
2008
2009
Longfin Dace
Headwater chub
Spikedace
Speckled dace
Loach minnow
Desert sucker
Juvenile
Sub-adult
Adult
Sonora sucker
Juvenile
Sub-adult
Adult
Gila trout
Total native
Common carp
Red shiner
Fathead minnow
Yellow bullhead
Flathead catfish
Western mosquitofish
Rainbow trout
Brown trout
Green sunfish
Smallmouth bass
Total nonnative
0.240
2.764
—
0.052
0.007
0.393
0.563
0.040
0.116
0.018
5.096
2.941
0.172
1.805
0.130
1.387
1.153
0.062
0.289
0.017
0.505
3.454
4.250
1.609
0.397
4.141
1.231
2.162
9.931
0.146
1.975
182.413
—
195.806
—
—
—
10.476
—
0.022
7.993
4.926
0.034
2.004
25.455
1.837
1.212
165.937
6.175
182.438
—
—
0.003
4.387
—
—
1.128
6.994
—
0.059
12.571
5.413
6.172
177.556
8.513
221.122
—
—
—
2.406
2.500
0.006
6.509
5.479
0.035
0.576
17.511
<30 mm were not considered in assessing native response to
nonnative removal. Exclusion of individuals <30 mm TL
had negligible effect on estimates of juvenile mass because
they accounted for a small proportion of that age group’s mass.
Mass was used as a response variable because it was, we
believed, more informative of potential predator impact than
number of specimens. Mass was log10 + 1 transformed prior
to analysis. Because of the low power (n = 6 years) and
exploratory nature of our study, we set α at ≤0.1 so as to not
exclude trends that might be obscured by a single data point.
Cost estimation
Field crews were composed of summer interns, technicians,
staff biologists and supervisors, with a commensurate salary
range. Collectively, daily salary and benefits for the field
crew was about $3300 for a 10-h day. Per diem was $30
per person. Equipment and supply costs were not included,
but vehicle expenses were estimated at $250 each for the
five vehicles used each year. Expenses associated with
obtaining reference site data were not included in total
annual cost estimates, nor were planning and data compilation and synthesis costs.
Copyright © 2014 John Wiley & Sons, Ltd.
2010
2011
2012
Total
3.169
5.737
0.925
1.680
0.150
1.562
1.656
0.328
0.627
0.045
11.847
14.814
1.527
4.569
0.367
0.659
0.606
4.225
2.568
1.274
12.704
0.301
1.910
2.508
6.873
9.803
37.759
3.379
5.157
12.918
0.160
30.012
—
0.003
0.048
3.371
—
—
—
0.286
0.020
1.470
5.198
5.425
4.965
163.203
—
201.800
—
0.002
0.186
3.610
2.750
0.006
0.157
1.624
0.048
0.729
9.112
0.675
2.614
32.866
1.086
46.178
0.590
0.003
0.135
0.872
31.190
0.135
0.120
0.960
—
2.759
36.764
16.875
22.095
734.893
15.934
877.365
0.590
0.008
0.372
25.122
36.440
0.169
15.907
20.269
0.137
7.734
106.611
RESULTS
The native fish community in the Heart Bar nonnative
removal reach was composed of five cyprinids (longfin dace
Agosia chrysogaster, headwater chub Gila nigra, spikedace
Meda fulgida, speckled dace Rhinichthys osculus and loach
minnow T. cobitis), two catostomids (desert sucker
Catostomus clarkii and Sonora sucker Catostomus insignis)
and one salmonid (Gila trout Oncorhynchus gilae) (Table II).
Longfin dace and speckled dace were usually the most common small-bodied native fishes, and Sonora sucker was the
most common large-bodied native fish each year. During
the study, 10 nonnative species were collected: three cyprinids (common carp Cyprinus carpio, red shiner Cyprinella
lutrensis and fathead minnow Pimephales promelas), two
ictalurids (yellow bullhead Ameiurus natalis and flathead
catfish Pylodictis olivaris), two salmonids (rainbow trout
Oncorhynchus mykiss and brown trout Salmo trutta), one
poeciliid (western mosquitofish Gambusia affinis) and two
centrarchids (green sunfish Lepomis cyanellus and
smallmouth bass Micropterus dolomieu). Among nonnative
fishes, yellow bullhead, rainbow trout, brown trout and
smallmouth bass were the most frequently collected. In
River Res. Applic. (2014)
DOI: 10.1002/rra
D. L. PROPST ET AL.
2007, native fishes were 81% of the individuals collected,
but from 2008 through 2012, native fishes were at least
92% of total individuals.
Among native fishes, suckers, especially Sonora sucker,
had a disproportionally greater mass in all years, and desert
sucker had the second-greatest mass in all years (Table III).
Mass of headwater chub and Gila trout, the only other
large-bodied native species present, were considerably less
than that of either sucker in most years. Among small-bodied
native fishes, mass of longfin dace was the greatest. Mass of
each small-bodied native varied considerably, but synchronously from year to year across reference sites and the
removal reach (Figure 2). In contrast, mass of Sonora sucker
was comparatively less variable across years at reference sites
and the removal reach, that of desert sucker somewhat
variable but synchronous, and headwater chub mass was
variable and asynchronous across all locations. In 3 years,
yellow bullhead mass was greater than any other nonnative
species in the removal reach, but in 2008, brown trout mass
was greatest, and rainbow trout was greatest in 2009. In
2012, flathead catfish mass exceeded that of all other nonnative
fishes owing to 11 large individuals (mean TL = 477 mm and
mean mass = 3735 g). Changes in mass of nonnatives,
especially rainbow and brown trouts, were synchronous
among reference sites and the removal reach (Figure 3).
Collectively, the native fish assemblages at the reference sites
were similar in composition to that in the removal reach, but
individually, there were differences. Gila trout was not found
at Middle Fork, and loach minnow was not found at either
reference site (Table IV). Prior to this study, loach minnow
was irregularly collected at both reference sites (Stefferud
et al., 2011). Gila trout was stocked during autumn in West
Figure 2. Mass of native fishes in Heart Bar removal reach (HB) and Middle (MF) and West (WF) forks Gila River reference sites, Catron
County, New Mexico, 2007–2012. Zero values entered as 1. Note different scales
Copyright © 2014 John Wiley & Sons, Ltd.
River Res. Applic. (2014)
DOI: 10.1002/rra
NONNATIVE PREDATOR REMOVAL
at West Fork, but at Middle Fork, longfin dace had greatest
mass among small-bodied fishes in 3 of 5 years.
Capture efficiency estimates from a two-pass closed
population estimate at the Middle Fork reference site
indicated that at least 30% of nonnative predator mass was
captured each year (Table V). Because effort in the removal
reach involved two crews and at least four netters, it is likely
that 50%, or more, of predator species mass was taken each
year from the removal reach.
After 6 years of nonnative fish removal, mass of nonnative
yellow bullhead, rainbow trout and brown trout declined
(Table VI). Because brown trout also declined at the West
Fork site, we could not attribute its decline in the removal
reach to our effort. There also was no change in overall
nonnative mass in the removal reach over time, but if the
mass of the 11 flathead catfish captured in 2012 was removed
as an outlier, mass declined significantly. Total nonnative
mass declined at West Fork but increased at Middle Fork.
Contrary to our prediction, there was no consistent change
in size structure of nonnative fishes (Figure 4), as adults
accounted for a majority of mass in all years. Whereas adult
smallmouth bass was absent from collections in 2008, they
reappeared and accounted for the majority of the mass in
2010 and 2012.
Spikedace was the only native species to increase significantly with time in the removal reach (Table VI, Figure 2).
Loach minnow was present in all years in the removal reach
but was not collected at either reference site during the study,
despite occurring at both prior to this study and their proximity to the removal reach. At reference sites, the only significant
trend in mass of native species was a decline in mass of desert
sucker adults at West Fork and a decline of desert sucker subadults at Middle Fork.
Expenses (US $) were $13 200 (salary and benefits),
$1440 (per diem) and $1250 (vehicles) for a 4-day effort
(n = 3), and $16 500 (salary and benefits), $1800 (per diem)
and $1250 (vehicles) for a 5-day effort (n = 3). Thus, total
field costs were about $106 320 for the 6-year effort.
Expressed in terms of total nonnative fish mass removed,
the cost was about $995/kg of fish removed.
DISCUSSION
Figure 3. Mass of nonnative fishes at Heart Bar removal reach (HB)
and Middle (MF) and West (WF) forks Gila River reference sites,
Catron County, New Mexico, 2007–2012. Zero values entered as 1
Fork Gila River in the vicinity of the reference site, but not in
Middle Fork Gila River. Yellow bullhead and smallmouth bass
were not found at West Fork during the study, and neither trout
species was common at Middle Fork. By mass, Sonora sucker
was the dominant native at both reference sites in all years.
Among small-bodied natives, speckled dace held that position
Copyright © 2014 John Wiley & Sons, Ltd.
The results of this study suggest that with an annual 4- to 5-day
intensive removal effort in a river reach open to immigration,
mass of several nonnative species could be reduced, and in
doing so, at least one native species, spikedace, and perhaps
desert sucker, benefitted. These results contrast, to some extent,
with the findings of several others who reported that considerable and continuous effort was required to measurably impact
nonnatives and improve the status of native species (Knapp
et al., 2007; Pine et al., 2007; Coggins et al., 2011) and
River Res. Applic. (2014)
DOI: 10.1002/rra
D. L. PROPST ET AL.
Table IV. Mass (kg) of fishes collected at West and Middle forks Gila River reference sites (lengths = 234 and 297 m, respectively),
2008–2012, Catron County, New Mexico
West Fork
Middle Fork
Year
Species
2008
2009
2010
2011
2012
2008
2009
2010
2011
2012
Longfin dace
Headwater chub
Spikedace
Speckled dace
Desert sucker
Juvenile
Sub-adult
Adult
Sonora sucker
Juvenile
Sub-adult
Adult
Gila trout
Total native
Common carp
Red shiner
Fathead minnow
Yellow bullhead
Flathead catfish
Rainbow trout
Brown trout
Western mosquitofish
Smallmouth bass
Total nonnative
—
0.114
0.006
0.184
0.011
—
0.016
0.249
0.006
0.124
0.002
0.113
0.053
0.002
0.017
0.792
0.025
0.006
0.003
0.244
0.003
0.420
0.005
0.002
0.140
0.177
0.035
0.022
0.027
0.330
0.002
0.005
0.041
0.198
0.018
0.275
—
—
—
0.007
0.026
0.329
1.386
0.127
0.061
0.309
0.021
0.137
0.157
0.436
0.061
0.189
—
0.050
0.046
0.124
0.137
0.230
0.300
0.214
—
0.013
0.091
0.157
0.435
0.038
0.189
—
0.038
0.046
0.224
0.135
2.466
0.095
4.965
—
—
—
—
—
0.037
2.468
—
—
2.505
0.143
0.166
4.809
0.425
6.316
—
—
—
—
—
0.457
2.998
—
—
3.455
0.030
0.082
2.324
—
2.996
—
—
—
—
—
0.056
0.212
—
—
0.268
0.129
0.185
1.186
—
3.050
—
—
—
—
—
0.172
0.463
—
—
0.635
0.012
0.117
2.954
—
3.457
—
—
—
—
—
—
—
—
—
—
0.007
0.045
6.366
—
7.339
—
—
—
0.473
—
0.035
0.095
—
—
0.605
0.525
0.612
7.858
—
9.883
—
0.036
—
0.410
—
0.205
0.019
—
0.911
1.581
0.017
0.567
1.195
—
2.404
—
0.014
—
1.484
—
—
0.120
—
0.218
1.836
0.137
0.518
2.496
—
4.345
—
0.001
0.010
2.382
—
—
0.020
—
0.045
2.458
0.013
—
3.615
—
3.719
4.200
—
—
0.542
1.620
—
—
0.029
0.446
6.837
Loach minnow and green sunfish were not collected at either reference site during the study.
demonstrated that nonnative electrofishing mechanical
removal efforts are not always ‘quixotic enterprises’ (Meyer
et al., 2006). The relative success of our efforts was likely
related to the comparatively small size of the West Fork Gila
River in the removal reach. Mechanical removal efforts in
large systems have been less successful (Tyus and Saunders,
2000; Mueller, 2005), although these efforts were evaluated
early in their implementation. Recently, Franssen et al.
(2014) showed that sustained removal efforts yielded riverwide declines in common carp and reach-specific declines in
channel catfish Ictalurus punctatus in the comparatively large
San Juan River. Nonnative removal success in the West Fork
Gila River was also facilitated by the susceptibility of target
species to capture with the gear used (estimated > 50% mass
removal), initial low abundance in the removal reach and
persistent low abundance in adjoining reaches.
Several points merit consideration in evaluating the
efficacy of a nonnative control effort, such as that undertaken in this study. With only 6 years of data, the power to
detect statistically significant trends was influenced by
uncontrolled factors that were likely to either mask effects
of removal or be more powerful drivers of changes in
Copyright © 2014 John Wiley & Sons, Ltd.
assemblage structure than removal of nonnative fishes. Similar
trends in the removal reach and reference sites, as we found
with brown trout, suggested regional factors, rather than our
removal efforts were driving abundance trends for this species.
For other species, it was evident that changes in mass were
synchronous across the removal reach and reference sites.
Mass of native fishes at all sites, except headwater chub at
reference sites, was lower in 2010 than in the preceding or
Table V. Percent of total estimated mass captured in first pass
during two-pass capture–recapture population estimate at Middle
Fork reference site during June 2008–2011
Native
Species
Longfin dace
Headwater chub
Spikedace
Speckled dace
Desert sucker
Sonora sucker
Nonnative
Percent
Species
Percent
19
41
54
55
18
24
Yellow bullhead
Rainbow trout
Brown trout
Smallmouth bass
31
60
73
36
River Res. Applic. (2014)
DOI: 10.1002/rra
NONNATIVE PREDATOR REMOVAL
Table VI. Pearson correlation coefficients of fish mass (log10 mass/10 m + 1) change over time for West Fork Gila River Heart Bar removal
reach, West Fork Gila River reference site and Middle Fork Gila River reference site
Removal
Species
Longfin dace
Headwater chub
Spikedace
Speckled dace
Loach minnow
Desert sucker
Juvenile
Sub-adult
Adult
Sonora sucker
Juvenile
Sub-adult
Adult
Total native
Yellow bullhead
Rainbow trout
Brown trout
Smallmouth bass
Total nonnative
Total NN without 2012 flathead catfish
West Fork
r
Middle Fork
r
p
p
r
p
0.57
0.22
0.72
0.55
0.42
0.24
0.68
0.10
0.26
0.41
0.77
0.51
0.14
0.38
—
0.13
0.38
0.81
0.52
—
0.24
0.78
0.30
0.34
—
0.70
0.12
0.63
0.56
—
0.11
0.12
0.01
0.83
0.82
0.99
0.05
0.75
0.87
0.94
0.14
0.02
0.38
0.88
0.08
0.52
0.05
0.90
0.02
0.44
0.53
0.53
0.70
0.61
0.53
0.12
0.01
0.69
0.79
0.38
0.28
0.28
0.04
0.07
0.10
0.51
0.97
0.04
0.76
0.09
0.33
0.70
—
0.15
0.79
—
0.76
—
0.13
0.89
0.59
0.19
—
0.52
0.04
—
0.06
—
0.11
0.22
0.48
0.62
0.17
—
—
0.12
0.92
—
0.86
0.72
0.41
0.27
0.49
—
—
0.57
0.01
—
An α of 0.10 was considered significant. Bold values indicate significant relationships.
succeeding year. This was coincident with higher and longer
duration spring discharge in 2010 than either 2009 or 2011
(discharge at time of sampling in 2010 was higher than most
years, but water was clear). Extended runoff may have delayed
spawning by native fishes and hence their not being
represented in the June catch. This might explain the decrease
in mass of small-bodied natives, of which age-0 individuals
typically comprise a larger portion of June mass than for
suckers. Comparatively low mass of juvenile suckers might
relate to effects of high spring discharge and delayed spawning
coupled with downstream displacement. Displacement by
elevated flows, however, does not seem a reasonable explanation for the reduction in mass of sub-adults and adults of both
sucker species in 2010. Nor did somewhat elevated flows
during sampling in 2010 preclude thorough sampling of all
habitats and thus reducing sampling efficiency. We have no
plausible explanation for the dramatic decline in mass of
native fishes, especially sub-adult and adult suckers, in 2010,
and subsequent rebound in 2011. Ash-laden flows associated
with the 2011 Miller Fire further caused mortality of fishes
in the removal reach as well as reference sites prior to sampling
in 2012 and thus likely diminished mass (from 2011 levels) of
all native fishes in both reference sites and removal reach.
These points illustrate how extrinsic factors confound removal
efficiency estimates or the statistical power to detect responses
in the native fish community.
Copyright © 2014 John Wiley & Sons, Ltd.
Because population turnover rates are higher for native
short-lived (≤3 years) species, we might expect a more rapid
response from them than for long-lived (≥3 years) species.
Indeed, our results provide some substantiation for this
premise in that mass of spikedace increased over time in
the removal reach, and there was no evident change in mass
of long-lived natives. Whether a native species benefitted
from nonnative predator removal might also be influenced by
the habitat it occupied (Stefferud et al., 2011). Thus, spikedace,
which is found in low to moderate-velocity habitats also occupied by nonnative predators, increased in mass, whereas no
change was noted for loach minnow, an obligate riffle-dweller
where nonnative predators were largely absent.
Removal of sufficient numbers of undesired fishes from an
open system to achieve desired effects is a challenge and is
dependent upon overcoming several obstacles. For example,
our inability to diminish smallmouth bass may have been that
individuals from contiguous stream reaches moved into the
removal reach, its low capture probability or there was a
compensatory response (Zipkin et al., 2009). Movement has
been implicated as the cause for the inability to control
mobile brook trout Salvelinus fontinalis; it quickly (<1 year)
recolonized areas where it was eradicated (Phinney, 1975;
Peterson et al., 2004). Because juvenile smallmouth bass
were infrequently captured during the 6-year effort, it seems
unlikely its persistence was a compensatory response. Nor
River Res. Applic. (2014)
DOI: 10.1002/rra
D. L. PROPST ET AL.
Figure 4. Proportion of each age class of large-bodied native (Sonora sucker and desert sucker) and nonnative (yellow bullhead, rainbow trout,
brown trout and smallmouth bass) fishes at West Fork Gila River Heart Bar removal reach, Catron County, New Mexico, 2007–2012
was relative capture efficiency likely a factor. All habitats,
including debris and root wad pools, occupied by smallmouth
bass in early June when discharge was at or near annual minimum were efficiently sampled with backpack electrofishers,
as indicated by our capture efficiency estimates. Thus, movement into the removal reach was the most likely reason mass
of this species was not diminished, rather than reproduction
and recruitment within the reach or poor capture probability.
Demonstrating the efficacy of removal efforts may
depend upon the initial conditions of the target system, as
it should be easier to demonstrate a positive effect if starting
under conditions where nonnative fishes vastly outnumber
natives. In the Colorado River below Glen Canyon Dam
where nonnative fish were numerically dominant, Coggins
et al. (2011) demonstrated a strong removal effect,
Copyright © 2014 John Wiley & Sons, Ltd.
decreasing the percentage of nonnatives in the assemblage
from 95% to 50% in 2.5 years. In comparison, nonnative
fish mass in the West Fork Gila River Heart Bar reach was
12% of total in 2007; and for 3 of the next 4 years, nonnative
mass was substantially less, indicating that removal efforts
were suppressing nonnative mass. But collection of 31 kg
(38% of total mass) of flathead catfish in 2012 dramatically
altered that perception.
Thus, the dilemma: With marginal support, was there
sufficient gain for native fishes to warrant continuation of
nonnative fishes removal from the West Fork Gila River?
Or from a broader perspective, do results of this effort
support application of this approach to other systems
supporting a native fish assemblage compromised by problem nonnative fishes? Several factors make a clear response
River Res. Applic. (2014)
DOI: 10.1002/rra
NONNATIVE PREDATOR REMOVAL
problematic. Certainly, the mixed results of other efforts
(e.g. Meronek et al., 1996; Meyer et al., 2006) provide a
cautionary backdrop. The inherent variability in species
abundances and environmental factors driving their
numbers coupled with a limited number of independent data
points (1 per year per species) reduces the power of any
analysis and thus the robustness of any interpretation.
Although present in 2009 and 2011, the dramatic increase in
flathead catfish mass (and abundance) in 2012 is puzzling.
Regardless of reason, if flathead catfish persists at these levels,
it imposes increased predation pressure on native fishes of the
upper Gila River.
A basic reason for an effort, such as that described herein,
was to remove or ameliorate documented threats to imperilled species. Of the eight native fishes occurring in the
study area, three (spikedace, loach minnow and Gila trout)
are federally protected, and a fourth (headwater chub) is
state protected. Each persists as several partially isolated
populations and loss of any would additionally compromise
their survival. Although restoration of imperiled organisms
to historical habitats is an essential element of many
conservation efforts, it is also critical that extant native
assemblages be maintained by viable, practical and
cost-effective measures (Clarkson et al., 2012). For each
of these species, opportunities to restore historical
populations are limited, thus making survival of extant
populations essential. Results of this study indicated that at
least one native fish responded positively to removal of
nonnative predators and that three problem nonnative
species declined. Importantly, this was accomplished with
a moderate expenditure (about $17 700/year) and within
what participating agencies could collectively afford.
Whether continued removal of nonnative predators will
yield more compelling results as to the efficacy of the effort,
only continued removal will tell. Regardless, options to
conserve native fish faunas are limited, and until or unless
nonnative predator suppression is clearly demonstrated to
have no positive and possible negative effects on native
fauna, the prudent course is to continue the effort.
ACKNOWLEDGEMENTS
Field work was accomplished with the participation of
individuals from cooperating agencies and universities.
Their assistance, especially that of Z. Law, N. Smith, S.
Carman, J. Zapata, S. Coleman, T. Archdeacon, C.
Roberts, D. Trujillo and B. Berger, is gratefully
acknowledged. Partial funding for this work was provided
to US Fish & Wildlife Service (NM Fish & Wildlife Conservation Office), US Forest Service (Gila National Forest),
N.M. Department of Game & Fish (Conservation Services
Division) and Kansas State University by grants from the
Gila River Basin Native Fishes Conservation Program.
Copyright © 2014 John Wiley & Sons, Ltd.
The comments and suggestions of an anonymous referee
greatly improved the manuscript and are much appreciated.
Scientific collecting permits were issued by US Fish &
Wildlife Service and N.M. Department of Game & Fish.
REFERENCES
Bestgen KR, Walford CD, Hill AA, Hawkins JA. 2007. Native fish
response to removal of non-native predator fish in the Yampa River,
Colorado. Final Report, Colorado River Recovery Implementation
Program Project Number 140. Larval Fish Laboratory Contribution
150. Colorado State University, Fort Collins.
Blinn DW, Poff NL. 2005. Colorado River Basin. Benke AC, Cushing CE
(eds). Rivers of North America. Elsevier Academic Press: Amsterdam;
483–538.
Coggins LG Jr., Yard MD, Pine WE III. 2011. Nonnative fish control in the
Colorado River in Grand Canyon, Arizona: an effective program or
serendipitous timing? Transactions of the American Fisheries Society
140: 456–470.
Clarkson RW, Marsh PC, Stefferud SE, Stefferud JA. 2005. Conflicts
between native fish and nonnative sport fish management in the southwestern United States. Fisheries 30: 20–27.
Clarkson RW, Marsh PC, Dowling TE. 2012. Population prioritization for
conservation of imperiled warmwater fishes in an arid-region drainage.
Aquatic Conservation: Marine and Freshwater Ecosystems 22: 498–510.
Cucherousset J, Olden JD. 2011. Ecological impacts of non-native freshwater
fishes. Fisheries 36: 215–230.
Dudgeon D, Arthington AH, Gessner MO, Kawabata Z-I, Knowler DJ,
Lévêque C, Naiman RJ, Prieur-Richard A-H, Soto D, Stiassny MLJ,
Sullivan CA. 2006. Freshwater biodiversity: importance, status, and
conservation challenges. Biological Reviews 81: 163–182.
Gido KB, Propst DL, Olden JD, Bestgen KR. 2013. Multidecadal responses
of native and introduced fishes to natural and altered flow regimes in the
American Southwest. Canadian Journal of Fisheries and Aquatic
Sciences 70: 554–564.
Gozlan RE, Britton JR, Cowx I, Copp GH. 2010. Current knowledge on
non-native freshwater fish introductions. Journal of Fish Biology 76:
751–786.
Fagan WF, Kennedy CM, Unmack PJ. 2005. Quantifying rarity, losses, and
risks for native fishes of the lower Colorado River basin: implications for
conservation listing. Conservation Biology 19: 1872–1882.
Fausch KD, Torgersen CE, Baxter CV, Li HW. 2002. Landscapes to
riverscapes: bridging the gap between research and conservation of
stream fishes. BioScience 52: 483–498.
Franssen NR, Davis JE, Ryden DW, Gido KB. 2014. Fish community
responses to mechanical removal of nonnative fishes in a large southwestern river. Fisheries.
Hendrickson DA, Minckley WL. 1984. Ciénegas—vanishing climax
communities of the American Southwest. Desert Plants 6: 130–175.
Hoagstrom CW, Brooks JE, Davenport SR. 2011. A large-scale conservation perspective considering endemic fishes of the North American
plains. Biological Conservation 144: 21–34.
Knapp RA, Boiano DM, Vrendenburg VT. 2007. Removal of nonnative
fish results in population expansion of a declining amphibian (mountain
yellow-legged frog, Rana mucosa). Biological Conservation 135: 11–20.
Kulp MA, Moore SE. 2000. Multiple electrofishing removals for eliminating rainbow trout in a small southern Appalachian stream. North
American Journal of Fisheries Management 20: 259–266.
Meronek TG, Bouchard PM, Buckner ER, Burri TM, Demmerly KK,
Hateli DC, Klumb RA, Schmidt SH, Coble DW. 1996. A review of fish
River Res. Applic. (2014)
DOI: 10.1002/rra
D. L. PROPST ET AL.
control projects. North American Journal of Fisheries Management 16:
63–74.
Meyer KA, Lamansky JA Jr., Schill DJ. 2006. Evaluation of an unsuccessful brook trout electrofishing removal project in a small Rocky Mountain
stream. North American Journal of Fisheries Management 26: 849–860.
Minckley WL, Marsh PC. 2009. Inland Fishes of the Greater Southwest:
Chronicle of a Vanishing Biota. University of Arizona Press: Tucson.
Mueller GA. 2005. Predatory fish removal and native fish recovery in the
Colorado River mainstem: what have we learned? Fisheries 30: 10–19.
Olden JD, Poff NL, Bestgen KR. 2008. Trait synergisms and the rarity,
extirpation, and extinction risk of desert fishes. Ecology 83: 847–856.
Peterson DP, Fausch KD, White GC. 2004. Population ecology of an
invasion: effects of brook trout on native cutthroat trout. Ecological
Applications 14: 754–772.
Phinney DE. 1975. Repopulation of an eradicated stream section by
brook trout. Transactions of the American Fisheries Society 104:
685–687.
Pilger TJ, Gido KB, Propst DL. 2010. Diet and trophic niche overlap of native and nonnative fishes in the Gila River, USA: implications for native
fish conservation. Ecology of Freshwater Fish 9: 300–321.
Pine WE III, Kwak TJ, Rice JA. 2007. Modeling management scenarios
and the effects of an introduced apex predator on a coastal riverine
fish community. Transactions of the American Fisheries Society
136: 105–120.
Copyright © 2014 John Wiley & Sons, Ltd.
Pool TK, Olden JD. 2014. Assessing long-term fish responses and shortterm solutions to flow regulation in a dryland river basin. Ecology of
Freshwater Fish. DOI: 10.1111/eff.12125
Propst DL, Gido KB, Stefferud JA. 2008. Natural flow regimes, nonnative
fishes, and native fish persistence in arid-land river systems. Ecological
Applications 18: 1236–1252.
Seager R, Ting M, Held I, Kushnir Y, Lu J, Vecchi G, Huang H-P, Harnik
N, Leetmaa A, Lau N-C, Li C, Velez J, Naik N. 2007. Model projections
of an imminent transition to a more arid climate in southwestern North
America. Science 316: 1181–1184.
Stefferud JA, Gido KB, Propst DL. 2011. Spatially variable response of
native fish assemblages to discharge, predators and habitat characteristics
in an arid-land river. Freshwater Biology 56: 1403–1416.
Tyus HM, Saunders JF III. 2000. Nonnative fish reference and endangered
fish recovery: lessons from the Colorado River. Fisheries 25: 17–24.
Vorosmarty CJ, McIntyre PB, Gessner MO, Dudgeon D, Prusevich A,
Green P, Bunn SE, Sullivan CA, Liermann CR, Davies PM. 2010.
Global threats to human water security and river biodiversity. Nature
467: 555–561.
Weedman DA, Sponholtz P, Hedwall S. 2005. Fossil Creek Native Fish
Restoration Project. Arizona Game and Fish Department: Phoenix.
Zipkin EF, Kraft CE, Cooch EG, Sullivan PJ. 2009. When can efforts to
control nuisance and invasive species backfire? Ecological Applications
19: 1585–1595.
River Res. Applic. (2014)
DOI: 10.1002/rra