Oecologa (Berlin) (1983) 60:328-332

____

OewIogia

< Springer-Verlag 198.7

Algae-grazing minnows (Campostoma anomalum), piscivorous
bass (Mcropterus spp.), and the distribution of attached
algae in a small prairie-margin stream

Mary E. Power and William J. Matthews
Division of Environmental Studies University of California, Davis, CA 95616, USA
University of Oklahoma Biological Station Star Route B. Kingston, OK 73439. USA

Summary. Campostoma anomalum is an algae-grazing min-
now, abundant in many streams of the central and eastern
United States. In a small stream in south-central Oklahoma.
Carnposroma has a marked impact on standing crops of
attached algae. Pools with schools of Campostoma are bar-
ren, while pools in which Camposroma are apparently ex-
cluded by bass (Microprerus salmoides or M. puncrularus)
support large standing crops of filamentous green algae
(predominantly Spirogj.ra sp. and Rhi:oclonium sp.). Cam-
posromu grazed actively on algae-covered cobbles trans-
ferred into their pools, and visibly reduced standing crops
within one hour. After 24 h of exposure to Campostoma.
standing crops of attached algae on cobbles were reduced
from 22.0 to 6.3 mg ash-free dry weight an-'. When a
largemouth bass was tethered in a pool with Campostoma,
the minnows did not graze on algae-covered cobbles within
30-50 cm of the bass. but fed actively on cobbles that were
more than 1.3 m away. These results indicate that interac-
tions of Campostoma and their predators may be an impor-
tant factor contributing to pool-to-pool variation in at-
tached algae in small streams of the central and eastern
United States.

Introduction

At low flow, Brier Creek, a prairie-margin stream in south-
central Oklahoma, has two distinctly different types of
pools. Some are barren, with bedrock and gravel substrates
bare or thinly mantled with detritus. Other pools have
luxuriant standing crops of filamentous green algae. This
algae offers cover for a variety of fishes and invertebrates.
For example, we observed snails (Physa sp.) on floating
tops of attached green algal clumps, microhabitats which
afforded them food as well as protection from predatory
fishes.
Barren pools contain schools of Campostoma anomalum
(Rafinesque), a minnow that is widespread and abundant
in streams throughout the central and eastern United States
(Burr 1980; Pflieger 1975). Camposroma is almost complete-
ly algivorous. with diatoms and filamentous algae compris-
ing over 95% of the gut contents (Kraau 1923). These
minnows are eaten by bass (Microprerus spp.) (Scale1 1977;
Funk and Fleener 1974; Lennon and Parker 1960; Gunning

Offprinr requests io: M.E. Power

and Lewis 1956). Here, we present data suggesting that
the interactions of bass and Campostoma have major im-
pacts on standing crops of attached algae in a small stream
during periods of low flow. and may thereby exert import
direct and indirect effects on other stream biota.

Study site

Brier Creek. in Marshall County. Oklahoma USA (31"O'N.
97'2'W). is a clear stream which flows over limestone bed-
rock and gravel. in a tree-lined channel cutting through
pastureland. This stream has been described by Smith and
Powell (1 970). Our research was conducted in a 1 km reach
which includes fourteen pools and twelve riffles (Fig. l),
and corresponds to Section 5 of Smith and Powell (1971).
The study was carried out during November 1982. a period
of low stream discharge (0.01 2 m3 sec- '). All pools in our
study reach remained connected at this flow. This reach
of Brier Creek has remained free-flowing throughout the
year during all but one of the last twelve years (Smith and
Powell 1970; Ross, Matthews and Echelle, in preparation).
Minimum depths of thalwegs (the deepest part of the chan-
nel cross-section) of rimes between pools are shown in
Fig. 1. and listed in Table 1 along with other physical attri-
butes of the reach.

Field observn tions

Methods

We mapped the 1 km reach, measuring the maximum depth
at five to ten positions in each pool and at two in each
rime. On 8 November and 19 November. we performed
censuses in all pools. 'One observer snorkelled slowly up-
stream, spending 10-20min underwater in each pool. re-
cording the species, numbers and approximate sizes of all
fish encountered. A second observer on the bank watched
during the census to count any fish that escaped detection
by the underwater observer. Pool were narrow and clear.
SO that few fish. even small individuals. escaped observa-
tion. On only one occasion did the observer on the bank
note a fish (a bass) that the underwater observer failed
to see.
We described algal standing crops in pools qualitatively
and quantitatively in three ways. In all fourteen pools. u.e
measured the height of benthic algae at sampling sites which
were located at 1.0 or 0.5-m intervals along five cross-

329

+++ tmr
4

25

0 20 40

COntwr mlevols 25cm
Oad IOOcm nlerrols dorkenea

Fig. 1. Map, drawn to scale, of 1 km study reach in Brier Creek

stream transects spaced evenly along each pool. We noted
the predominant material at each site (usually Chara, fila-
mentous green algae (predominantly Spirogyra and Rhizo-
clonium), or loose deposits of detritus), and scored condi-
tion of the algae from 0 to 5 (O=very senescent, 5=very
fresh). In six pools, we collected samples by scraping 2cm2
areas of substrate with a scalpel and aspirating detached
material underwater. Samples were collected in small plastic
containers and refrigerated for 3-5 days. Subsequently,
samples were shaken, and drops were collected for micro-
scopic examination. Sample composition was quantified
under 400 x , using a Whipple grid and the method of Jones
(1968). Four fields were counted per sample, and 100 points
(grid intersections) were examined per field. When the grid

Table 1. Physical charactenstics of fourteen pools In a 1 km reach
of Bner Creek. Marshall County. Oklahoma

Pool                     R~me

Length Max. Max.       Length Minimum
(m) width depth       (m) thalweg

(m) (m) depth

(m)

100
49
49
39
57.6
59.1
58
88
22
21
74.9
43
6.4
120

8
8
6
8
6.5
7
9
6.5
5.5
4.5

10
9
2
9

1.50
0.70
0.45
1.25
0.50
1.25
1 .oo
0.60
0.80
0.50
0.60
0.75
0.45
1 .oo

1-2
2-3
3-4
4-5
s6
67
7-8
8-9
%lo

10-1 1
11-12
12-13
13-14

43
20
12.4
17.2
39
2
26
9
2b
29
11.3
15
5.3

0.12
0.05
0.06
0.05
0.06
0.03
0.05
0.06
0.02
0.04
0.04
0.12
0.07

' Pool 1 is the most downstream (Fig. 1)

Pools 9 and 10 are also separated by a 38 em high natural water-
fall

intersection touched an alga or an animal, it was scored.
Mineral particles. detritus and empty space were not scored.
The following eight categories suficed to describe all other
material encounterd: full and empty diatoms. full and
empty green algal filaments, full and empty blue-green al-
gae, full unicellular green algae and animals. After examina-
tion. material on slides was rinsed back into appropriate
containers with distilled water, and the entire sample was
filtered onto tared Whatmen No. 1 filter paper. dried 24 h
at 70" C and combusted for 3-3.5 h at 510" C for determi-
nation of ash-free dry weight.

Results

Distributions of Camposroma and large (> 70 mm SL) bass
among pools showed significant complementarity (P -=
0.025 and P< 0.05 (Fisher's exact test) on 8 November and
19 November, respectively). Counts of Cumposroma in giv-
en pools were similar on the two census dates (r2 = 0.91.
P<O.001). Numbers of bass per pool varied more between
the two dates (r'r0.50, P<O.lO) but in only one case
(pool 5) was a pool found with bass on one date but not
the other (Table 2). ,
Pools in which bass longer than 70mm SL occurred
and Camposroma were absent ("bass pools") had lush
standing crops of filamentous green algae (predominantly
Spirogyra sp. and Rhizocloniwn sp.) and beds of Chara spp.
heavily covered with epiphytes. In pools where Cumposroma
schools occumd and bass were absent ("Cumpostonta
pools") filamentous green algae were visible only along the
margins, in depths of no more than 15 and usually less
than 3 cm. Chara, where it occurred in Camposroma pools.
was stripped of epiphytes and many of its branches. The
bedrock and gravel substrates in Camposroma pools were
bare or thinly mantled with fine detritus. Figure 2 shows
the thickness of algal mats or detritus deposits in nine bass
and four Cumposroma pools at various depths. Table 3 sum-

#

330

Table 2. Numbers of bass longer than 70 mm SL and of Campos-
roma counted in fourteen pools of Brier Creek. and the abundance
of filamentous green algae (- =scant or absent, + =abundant)

Pool 8 November 1982 19 November 1982

Bass Cam- Fila- Bass Cam- Fila-

       algae algae

postoma mentous postoma mentous

   0     0 +
.3     0 +
31 0    400 -

        0 +
4) 3
5) 0 300
6) 4 0 +
7) 8 200 +/-
8) 3 0 +

        0 +
9) 4
10) 0 340
11) 7 0 +
12) 3 0 +
13) 0 70 -

14) 5 O -I-

-

-

4 J +
4 0 +

0 300     -
6 0     +
5 200     +
7 0     +
7 200    +I-

2 0 +
4 0 +
0 350 -

7 0 +
omitted due to turbidity
0 50 -

3 0 +

-

0
01

m

1

o Campostorno pwls

n Boss DOO~S

IS I

Depth (cm)

Fig. 2. Height of attached algae (ji+ 1 S.E.) at sites along cross-
stream transects in four Camposroma and nine bass pools. Data
are grouped into depth intervals. Heights of attached algae are
significantly greater in bass pools than Cumpostornu pools at depths
of 11-20cm, 2140cm and 414m (P's from Mann-Whtney
U tests<0.002, 0.01 and 0.006. respectively)

7 *wr 3 -`:rant of sites' sampled in fourteen pools where the fol-
: ~ .+ rr?g predominated

Bare Detri- Chara fila-
rock tus mentous

algae

Bass pools 0 9 30 61
(1. 2.4. 6, 7, 8. 9, 11, 12, 14)
Camposroma pools 11 56 9 23
(3. 5, 10, 13)

a Number of sites in Bass Pools: 273; in Cmposroma Pools: 87

marizes data on the macroscopic appearance of algae de-
scribed at sites along 60 cross-stream transects in fourteen
pools. There was a higher proportion of bare rock and
detritus in Camposroma pools. while Chara and filamentous
green algae were proportionately more abundant in bass

Table 4. Composition of algal samples collected in three bass and
three Camposioma pools

Mean proportion (k S.E.)

Bass pools Campostorno pools

Diatoms
Full 0.42 f 0.12 0.14 + 0.05
Empty 0.10 * 0.02 0.37 2 0.06

Filamentous green algae
Full 0.39 k 0.14 0.10 & 0.07
Empty 0.01 k 0.01 0.20 f 0.08

Blue-green algae
Full 0.05 f 0.02 0.09 f 0.07
Empty 0.03 f 0.01 0.10+0.06

pools (P< 0.001 from t-tests on arcsin-transformed data
(Sokal and Rohlf 1969) for all four comparisons).
Results from microscopic examinations of samples from
three bass and three Camposroma pools are summarized
in Table 4. Diatoms and filamentous green algae with intact
cell contents occurred in significantly higher proportions
in bass pools, while empty cells of both types were propor-
tionately more abundant in Campostornu pools (P< 0.001
for all four comparisons from t-tests on arcsin-transformed
proportions). Blue-green algae (Oscillaroria sp.) made up
a smaller proportion of algae in both types of pools (Ta-
ble 4). and both intact and empty blue-green filaments were
proportionately more abundant in Camposroma pools (P <
0.01 and P< 0.001, respectively). Unicellular green algae
and microscopic animals each comprised less than 1% of
539 and 806 points scored in bass and Camposloma pools.
respectively.

Experimental studies of bass-Campostoma-algae
interactions

To test the hypotheses that grazing by Campostonla affected
pool-to-pool occurrence of filamentous green algal mats.
and that bass influenced Cumposroma grazing on these al-
gae. we conducted two experiments. In the first. nine cob-
bles and one log, covered with thick mats of filamentous
green algae, were transferred from Pool 9 (a bass pool) to
the next pool upstream, Pool 10 (a Camposroma pool). The
two pools were separated by a rock ledge 38 cm high. The
minimum thalweg depth in the rime between Pools 9 and
10 was only 2 cm. over the lip of this ledge. Before cobbles
and the log from Pool 9 were placed in Pool 10, their stand-
ing crops of attached algae were sampled by scraping and
aspirating material from 1-2 cm2 areas on each. Two bare
cobbles were also moved within Pool 10 to serve as controls
for "novelty effects" that might attract minnows to any
displaced rock. The log and eleven cobbles were placed
in a single row. parallel to the banks of the pool. The two
control cobbles were in the first and fourth positions from
the downstream end. All cobbles were located within an
area where Cumposroma had formerly been seen. in water
2&50 cm deep. For 30min after the cobbles were posi-
tioned, two observers on the bank scanned the downstream
seven cobbles at 2-min intervals, recording the number of
Camposroma actively feeding on each.

331

Fig. 3. Attached filamentous algae on a submerged log before and
1 h after it was transferred into a pool with Cmposromo

Table 5. Average numbers of Cmposromo feeding on transferred

cobbles and controls (cobbles 1 and 4). N = 24 scans for all cobbles

Cobble 1 2 3 4 5 6 7

ir 0.0 7.5 7.3 0.0 5.4 6.4 14.5

S.E. 0.0 0.6 0.4 0.0 0.2 0.5 0.8

Cumposromu fed actively on algae on the transferred
cobbles and log within 10 min after they had been placed
in Pool 10. Control cobbles were virtually never visited (Ta-
ble 5). Figure 3 shows the transferred log just before and
1 h after it was exposed to Camposromu grazing. Standing
crops of attached algae (Xf S.E.) decreased from 22.050.3
to 6.3 f2.7 mg ash-free dry weight cm-' after 24 h of Cam-
posrornu grazing. Standing crops of algae on control cobbles
were not measurable before or after the experiments.
In a second experiment, conducted 7 days after the first,
cobbles with attached algae were again transferred from
Pool 9 to Pool 10, but placed in two groups of four. Each
cobble was 20-30 cm from the others in its group, and the
groups were 1.3 m apart. The eight cobbles were scanned
by one observer at 54x1 intervals for 25 min. Then a large-
mouth bass, 300 mm SL, was introduced into Pool 10. The
bass was tethered to a cobble on the streambed by a 0.8 m
nylon line fastened through the fish's lower jaw. This cobble
was placed in the center of the upstream group of trans-
ferred cobbles. The bass lunged once upon being place in
the pool, and afterwards, hovered quietly in mid-water.
After the bass was in position, the numbers of minnows
grazing on the eight cobbles was again recorded during
5 scan samples made at 5-min intervals. The bass and the
eight cobbles were left in the pool for 72 h, after which
the bass was removed and standing crops of attached algae
were sampled from the eight cobbles.
Composioma fed actively on attached algae on the eight
transferred cobbles before the introduction of the bass.
More fish visited the upstream group than the downstream

Table 6. Numbers of Cnmposromo feeding on algae-covered cobbles
transferred into their pool, before and after the introduction of
a tethered bass

Time Cobble

12345678

16:OO  cobbles 1-8 introduced into pool
16:15 1 1 2 3 2
16:20 1 1 2 3 4
16:25 2 1 4 3 4
16:30 3 2 2 6 8
16:35 5 2 1 6 6

662
112
245
175
265

16:37  bass tethered near cobbles 5-8
16:40 4 5 2 4 0
16:45 5 3 4 2 0
16:N 5 4 6 5 0
16:55 5 3 4 6 0
17:OO 7 5 6 7 0

000
000
000
000
000

group (Table 6). When we introduced the bass. all Carnpos-
romu initially stopped grazing. Within three min of the in-
troduction. however, Cumpostornu resumed grazing the
downstream group of cobbles. Upstream cobbles were no
longer visited by Cumpostornu after the bass was tethered
near the center of this group (Table 6). Cumposromo some-
times swam toward the "guarded" cobbles, but veered
sharply away when within 50 cm of the bass.
By the third day, the bass had assumed an acentric posi-
tion near three of the transferred cobbles, but more than
50 cm away from the fourth. We observed sporadic grazing
by Cuniposromu on this fourth cobble. Individuals would
dart in for several bites and then leave. Standing crops
of algae on the four guarded cobbles were
15.2k3.3 mg cm-'. On the three guarded, ungrazed cob-
bles they averaged 17.9k2.6 mg cm-', while standing
crops on cobbles 1.3 m away were 5.2 5 0.7, not significant-
ly different from cobbles exposed to 24 h of grazing in the
first experiment. The difference in standing crops of the
four guarded and four unguarded cobbles after 72 h was
significant (P= 0.03, Mann-Whitney U test).

Discussion

Certain grazers of attached algae exert strong effects on
aquatic communities (Paine 1980; Lubchenko and Gaines
1981; Vine 1974; Hay 1981; Kesler 1981; Hunter 1980).
When densities of such grazers change, changes ensue not
only in the food available to other grazers, but also in the
physical structure of the environment. For example, if sea
urchin densities fluctuate due to experimental manipula-
tions or changes in the abundances of their predators.
strong effects on kelp bed communities follow (Paine and
Vadas 1969; Mann and Breen 1972; Estes and Palmisano
1974; Duggins 1980).
Similarly, Curnposrorna schools can quickly reduce
standing crops of filamentous green algae. In Brier Creek.
this algae provides much of the cover in stream pools, a
factor of great importance to fish and invertebrates in small
freshwater habitats (Crowder and Cooper 1979; Werner
et al. 1977). At low flow in Brier Creek, Cumpostornu was
usually not present in pools where bass longer than 70 mm

I

332

SL occurred. (Largemouth bass become totally piscivorous
when they are 8&100 mm long (Keast 1970; Clady 1974;
Kramer and Smith 1962). Where large bass and Campos-
toma did co-occur in Brier Creek (Pool 7 and 8 November
and Pools 5 and 7 on 19 November), they were spatially
segregated, with bass in deeper areas and Campostoma in
shallower areas of the pool. Barren areas and standing
crops of filamentous green algae were correspondingly lo-
cated in shallower and deeper regions of the pool, respec-
tively.
Two different processes may account for the comple-
mentary distributions of Campostorno and bass in Brier
Creek. Bass predation may eliminate these minnows from
pools in which both species originally co-occur. Alternative-
ly, Camposroma may actively avoid bass by selecting pools
in which bass are absent. Fraser and Cerri (1982) found
that ai oidance of piscivorous fish by the minnows Semoti-
ltcs and Rhinichthys affected their distribution among com-
partments in an artifical stream. Our results from the exper-
imental introduction of a tethered bass indicate that Cam-
posroma avoid bass over short distances, but do not indicate
whether this behavior plays a role in lhgir habitat selection.
However the bass-Carnposrorna corsr;:ementarity arises.
it explains a conspicuous pattern of pooi-to-pool variation
ir, the quantity and quality of attached algae in Brier Creek.
:. ,i~:m~roma are widespread throughout the central and
easirrn United States (Burr 1980) and in some streams.
outweigh and outnumber all other species of fishes com-
bined (Lennon and Parker 1960; Beets 1979). Therefore,
the interaction described here may be an important factor
producing spatial variation in attached algae in many North
American streams.

Acknoicledgmenrs. We would like to thank Art Stewart for discus-
sion and help with algal identifications. and Ray Drenner and
Bill Dietrich for comments on the manuscript. The University of
Oklahoma Biological Station and the Universlr> of Oklahoma De-
partment of Zoology Visiting Faculty Prograrr: provided logistical
and financial support for this research.

References

Beets JP (1979) Population dynamics of the stoneroller minnow.
`Camposrorna anomalum (Rafinesque). in streams of a five-
county area in upper East Tennessee. MS thesis, University
  :,' 'I ennessee
Brr* BM (1980) Camposroma anomalum (Rafinesque), Stoneroller.
In: Atlas of North American fishes Lee DS et al. (eds) Raleigh,
North Carolina: Nonh Carolina State Museum of Natural His-
tory pp 143-144
Clady MD (1974) Food habits of yellnu nc"i:: smallmouth bass
and largemouth bass in two unproduciivt iakes in northern
Michigan. Am Midl Nat 91 :453459
Crowder LB, Cooper WE (1979) Structural complexity and fish-
prey interactions in ponds: a point of view. In: Response of
fish to habitat structure in standing water. Johnson DL, Stein
RA (eds) North central Div. Am Fish Soc Spec Pub 6

Duggins DO (1980) Kelp beds and sea otters: an experimental
approach. Ecology 61 :447-453
Estes JA, Palmisano JF (1974) Sea otters: their role in structuring
nearshore communities. Science 185: 1058-1060

Fraser DF. Cerri RD (1982) Experimental evaluation of predator-
prey relationships in a patchy environment : consequences for
habitat use patterns in minnows. Ecology 63 : 307-31 3
Funk JL, Fleener GG (1974) The fishery of a Missouri USA Ozark
stream, Big Piney River, and the efftcts of stocking fingerling
smallmouth bass. Trans Am Fish Soc 103 : 757-771
Gunning GE, Lewis WM (1956) Age and growth of two important
bait species in a coldwater stream in southern Illinois. Am Midl
  Nat 55:118-120
Hay ME (1981) Herbivory, algal distribution and the maintenance
of between habitat diversity on a tropical fringing reef. Amer
Natur 1 18 : 520-540

Hunter RD (1980) Effects of gradng on the quantity and quality
of freshwater aufwuchs. Hydrobiologia 69: 251-259
Jones RS (1968) A suggested method for quantifying gut contents
in herbivorous fishes. Micronesica 4: 369-371

Keast A (1970) Food specialization and bioenergetic interrelation-
ships in the fish fauna of some small Ontario waterways. In:
Marine food chains Steeie JH (ed). London. Oliver and Boyd.
pp 377-41 1

Kesler DH (1981) Periphyton grazing by Amnicola limnosa: An
enclosure-exclosure experiment. J Freshwater Ecology 1 : 51-59
Kraatz WC (1923) A study of the food of the minnow. Camposronia

anomalum. Ohio J Science 23:265-283
Kramer RH. Smith LL (1962) Formation of year classes in large-
mouth bass. Trans Am Fish Soc 91 : 29-41

Lennon RE. Parker PS (1960) The stoneroller. Camposrorna ono-
malum (Rafinesque) in the Great Smoky Mountains National - -
Park. Trans Am Fish SOC 89:263-270

Mann KH, Breen PA (1972) The relation between lobster abun-
dance and kelp forests. J Fish Res Bd Can 29 : 603-605
Paine RT (1980) Food webs: linkage, interaction strength and com-
munity infrastructure. J Anim Ecol 49:667-685

Paine RT, Vadas RL 11969) The effects of granng by sea urchins.
Srrongylocenrrorus spp.. on benthic algal populations. Limnol
Oceanogr 14:710-719
Pflieger WL (1975) The fishes of Missouri. Jefferson City. Mo,
  Western
Scalet CG (1977) Summer food habits of sympatric stream popula-
tions of spotted bass Microprerus puncrularus. and largemouth
bass Microprerus salmoides. (Osteichthyes. Centrarchidae ).
  Southwestern Nat 21 :49?-501
Smith CL, Powell CR (1971) The summer fish communities of
Brier Creek. Marshall County, Oklahoma. Am Mus Novitates
No. 2458, p 30

I
i
t

-

Sokal RR, Rohlf FJ (1969) Biometry, San Francisco. Freeman

Vine PJ (1974) ElTects of algal grazing and aggressive behavior
of the fishes Pomacenrrus lividus and Acanrhurus sohal on coral
reef ecology. Marine Biology 24: 131-136
WernerEE, Hall DJ, Laughlin DR, Wagner DJ, Wilsmann LA,
Funk FC (1977) Habitat partitioning in a freshwater commu-
nity. J Fish Res Bd Can 34:360-370

Received March 22, 1983