The Influence of Periphyton Abundance on Baetis bicaudatus Distribution and Colonization in a Small Stream Author(s): Carl Richards and

G. Wayne Minshall Reviewed work(s): Source: Journal of the North American Benthological Society, Vol. 7, No. 2 (Jun., 1988), pp. 77-86 Published by: Society for Freshwater Science Stable URL: http://www.jstor.org/stable/1467913 . Accessed: 06/03/2013 19:49
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Society for Freshwater Science is collaborating with JSTOR to digitize, preserve and extend access to Journal of the North American Benthological Society.

http://www.jstor.org

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

J. N. Am. Benthol. Soc., 1988, 7(2):77-86 ? 1988 by The North American Benthological Society

The influence of periphyton abundance on Baetis bicaudatus distribution and colonization in a small stream
CARL RICHARDS1 AND G. WAYNE MINSHALL Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209 USA Abstract. Colonization, distribution, and immigration of Baetisbicaudatus on rock surfaces was observedin a small Idaho(USA)stream.Baetis densities on rockswere assessedby directobservation through clear plastic viewing boxes. Baetisdistributions among regions on individual rocks was related to periphyton abundance.Densities of Baetison periphyton patches reflected the relative abundanceof periphyton. Among rocks, distributionsalso were most strongly influenced by pesize did not greatlyinfluencedistributionpatterns. riphytonabundance.Watervelocity and substrate Chlorophyll a was the best indicator of periphyton abundance. Immigrationrates, determined through time-lapsephotography,were as high as 5.8 individuals/100 cm2/hr.Over 50%of nymphs arrivedfrom the drift. At such immigrationrates, the proportionof immigrantsthat emigrate per unit time would have to be high to maintain densities observed on rocks after 48 hr. The results of this study show that Baetisbicaudatus abundance on rock surfaces is an interplay between immigrationrate and nymphal response to food abundance. microdistribution,immigration,emigration,colonization, periphyKeywords: Baetisbicaudatus, ton, grazer,streamecology. The distribution of lotic invertebrates is influenced by a variety of physical and biotic factors. Water velocity (Ciborowski and Clifford 1984, Corkum et al. 1977, Edington 1968) and substrate (Minshall 1984, Minshall and Minshall 1977, Reice 1980) often are cited as physical factors influencing local distributions of stream invertebrates. Biotic factors such as predation (Oberndorfer et al. 1984, Peckarsky 1980, Peckarsky and Dodson 1980) and competition (Hart 1983, 1985, Hemphill and Cooper 1983, McAuliffe 1983, 1984a, 1984b, Wiley 1981, Wiley and Kohler 1981) also affect local distributions. The relative contributions of biotic and physical influences on invertebrate distributions is a topic of continuing interest (see Minshall and Petersen 1985, Reice 1985, Robinson and Minshall 1986). Recent studies in streams suggest that food may be limiting to many stream invertebrates, particularly within the grazing community (Hart 1985, Hill and Knight 1987, Mayer and Likens 1987, McAuliffe 1983, 1984a, 1984b, Vaughn 1986). The periphyton community on the surfaces of stony substrates in streams represents a high-quality nutritional resource for stream invertebrates. Carbon/nitrogen ratios typically are lower for periphyton than for other foods within streams (Gregory 1983). If grazers are limited by food, their local distributions should reflect the abundance of this resource. Periphyton abundance varies over at least two spatial scales within streams: among rocks and on individual rock surfaces. Grazer foraging behavior often is concentrated in high resource areas over an individual substrate surface (Hart 1981, Hart and Resh 1980, Kohler 1984, Wiley and Kohler 1984). This type of foraging behavior should result in clumped distributions in areas of high periphyton abundance on individual substrates, although this has not been documented. Furthermore, if food is limiting and grazers have sufficient mobility to move among rocks, inter-rock distributions should be positively related to periphyton abundance. The interplay between the response to local (on a rock) and regional (among rocks) periphyton abundance should result in grazer distributions that reflect the abundance of food. Although some investigators have found that grazer distributions are related to periphyton abundance among rocks (Douglas 1958, Hawkins et al. 1982, Lamberti and Resh 1983, McAuliffe 1984b, 1Present address: 1986), few studies have examined the of ShoVaughn Department Fisheries, shone-BannockTribes,P.O. Box 306, FortHall, Idaho dynamics of this relationship. Our investigation was conducted to examine 83203 USA. 77

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

78

C. RICHARDSAND G. W. MINSHALL

[Volume 7

the influence of periphyton abundance on the nanno- and micro-distribution and short-term colonization of rocks by a mayfly, Baetis bicaudatus(Dodds), in a small Rocky Mountain stream. Baetis bicaudatusis widely distributed in small cool streams of western North America (Jensen 1966). Members of this genus are rapid colonizers (Ciborowski and Clifford 1984, McArthur and Barnes 1985, Shaw and Minshall 1980, Sheldon 1977) and often occur in large numbers. Densities are often greatest in stream reaches with high algal production (Behmer and Hawkins 1985, Wallace and Gurtz 1986). Drift behavior can be strongly related to food abundance (Bohle 1978, Kohler 1985). By drifting, Baetis can potentially move quickly among stream reaches or habitat patches. Several studies show that Baetis nymphs spend more time in high periphyton abundance areas when foraging on individual stony substrates (Hart 1985, Kohler 1984, McAuliffe 1984, Richards 1986), which indicates that crawling movements also may result in concentrations of Baetis in high food resource areas. Our study tested the hyboth pothesis that distribution of Baetisbicaudatus among rocks and on individual rocks is directly related to periphyton abundance. Methods Study site Experiments were conducted in Spring Run, a first order, spring-fed stream in the Targhee National Forest, southeastern Idaho (43?16'0"N, 111?04'30"W).Spring Run is a high gradient (410%) stream characterized by turbulent flow. Channel width ranged from 0.4 to 1.2 m and water depths ranged from 0.1 to 0.5 m. Mean summer flow was approximately 0.1 m3/s. Substrate was primarily gravel and cobble embedded in a marl matrix. Summer water temperatures ranged from 10 to 15?C. The study took place where the stream flowed through a 100-m-long meadow. Stream-side vegetation was predominantly grasses and forbs. A canopy of riparian vegetation was not present here, although one was present both upstream and downstream of the meadow. Periphyton communities on stony substrates in Spring Run were composed primarily of diatoms. Mid-summer periphyton abundances as indexed by chlorophyll a (chl a) were similar or slightly lower

than those reported for streams of similar size and geographic location (Andrews and Minshall 1978, Bott et al. 1985, Cushing et al. 1983). Baetisbicaudatuswas the only species within the genus found in Spring Run where it attained abundances similar to those reported in other Rocky Mountain streams (Andrews and Minshall 1978, Sheldon 1977).

Generalprocedures In all distribution experiments, the number of Baetisbicaudatuspresent on rock surfaces was observed directly by eye with the aid of plastic viewing boxes. The open, clear plastic (2.5-mm thickness) viewing boxes (40 x 27 x 16 cm) were wedged between two lengths of angle iron placed across the stream so that the open side of the box faced directly up. The bottom of each box was held approximately 1 cm below the airwater interface. Rocks below the viewing boxes were easily visible as were Baetis bicaudatus nymphs on their surfaces. The distance between the upper rock surfaces and the plastic viewing box bottom ranged from 7 to 15 cm. Nymphs could move freely on and off the surface. Placement of the viewing boxes did not noticeably alter water velocity and water-column turbulence patterns. Viewing boxes were set up at 7-10 small riffle/runs within the meadow area. Boxes were placed where they could be adjusted and maintained easily and thereby minimize observer disturbance. All rocks used in the experiment were cobble from within the wetted perimeter of the stream chosen for similarity in size (5-15 cm, maximum diameter) and surface complexity (smooth). All rocks and rounded edges, few or no angular irregularities, and flattened upper surfaces. An observation consisted of looking through a viewing box and counting the number of nymphs present on the upper flattened surfaces of the rock only. Less than a minute was required to count the individuals on any one rock. All individuals were counted regardless of size, although during all experiments the majority on rock surfaces were mature nymphs (3-5 mm). Periphyton was sampled by scrubbing a known area of rock surface using a technique modified from Stockner and Armstrong (1971) as described by Robinson and Minshall (1986).

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

1988]

BAETIS DISTRIBUTION AND COLONIZATION

79

A 3-cm2 sample was removed by placing a 35ml syringe tube which had been cut in half and had a neoprene gasket glued to the outer edge of the flared plunger end. The flared end was pressed against the rock surface and the enclosed periphyton was scrubbed into a slurry with a coarse brush. The slurry was then removed from the tube with a pipette and vacuum filtered through 2.4-cm Whatman GF/C filters (pore size 0.45 ium). Filters were immediately frozen on dry ice before transportation to the laboratory. Chlorophyll a was extracted by grinding a previously frozen filter in 5 ml or reagent grade acetone (100%), transferring the extractant to a centrifuge tube, and refrigerating the mixture at 4?C for 24 hr. Chlorophyll a and phaeopigment concentrations were determined with a Turner Model 111 fluorometer by multiplying the fluorescence readings by a calibration factor derived with a Beckman Instruments Model-DB spectrophotometer (APHA 1980). Since so little solution was required for fluorometer readings (<1% of volume), the remaining solution from each sample was then used without adjustment for measurement of total biomass (ash-free dry weight). Samples were oven-dried to constant weight at 60?C, cooled to room temperature, weighed, ashed at 550?C for 2 hr, rehydrated, redried at 60?, and reweighed. Experimentson individual rocks To examine whether Baetis bicaudatus distribution on individual rock surfaces is influenced by periphyton abundance, a series of rocks having patchily distributed periphyton was created. The purpose of this experiment was to determine whether distributions are influenced by relatively large differences in periphyton abundance within a range of patch sizes. Rocks were selected from the stream as previously described and an acetate sheet, on which a checkerboard pattern of square holes had been cut, was pressed against the rock surface and open squares were scraped with a stiff brush. Only one size of square was scraped on any one rock. Rocks with 0.5-, 1.0-, and 2.5-cm-wide squares were created. The width of the smallest patches corresponded roughly to 1 body length of a mature Baetis bicaudatus nymph. The pattern covered most or all of the upper surface of each rock so that approximately 50% of the available

surface area was low resource (scraped) and 50% was high resource (unscraped). The total number of squares on each rock varied depending on the upper rock surface area and size of squares created. Total patterned area on rocks ranged from 35 to 150 cm2. Only one rock was removed from the stream at a time for pattern creation. This process took only a few minutes for each rock. In experiment A (June 1984), five viewing boxes were placed in the stream. One rock from each of the three patch size categories was placed beneath each viewing box for a total of 15 rocks. Rocks were in place by late afternoon and observations of Baetis bicaudatuson high and low resource squares on each rock were conducted at 0800, 1200, and 1700 hr the next day. The position of an individual's head was used to determine whether the nymph was on a high or low resource square. In the time required to count the number of nymphs on high and low resource squares, little or no movement occurred among squares on the rock. The total number of Baetis on a particular resource level of a rock was divided by the total area of those squares to obtain a density value. The average number of Baetis over the three observation periods on high and low resource squares for each rock was used in statistical analysis. One periphyton sample was taken from one high and low resource patch from each of the 2.5-cmpatch width rocks at the end of an experiment to obtain an estimate of periphyton abundance. Since the diameter of the periphyton sampling tube precluded sampling smaller patches, we used the content of the larger patches to give an estimate of the magnitude of difference between all high and low resource patches. In experiment B, the rocks from experiment A were allowed to remain in the stream for 14 days. We expected that during this 14-d period, periphyton would accumulate on the scraped patches so that the difference in periphyton abundance between high and low patches would be decreased, thereby allowing determination of whether Baetisbicaudatus distributions reflected the decreased difference between high and low patches. All invertebrates were removed from the rocks, rocks were placed beneath viewing boxes, counts of Baetis were made, and periphyton samples were taken as in experiment A. In experiment C (August 1985), seven view-

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

80

C. RICHARDSAND G. W. MINSHALL

[Volume 7

ing boxes were placed in the stream and all patch creation, observations and sampling were repeated as in experiment A with a total of 21 rocks. Group comparisons were made by two-way ANOVA after comparing variances of treatment means. Experimentsamong rocks In experiment D (June 1983), a series of rocks was chosen from a 1-km section of the stream that included the meadow. Rocks were picked from the stream, and if they met physical criteria stated above, were assigned by direct visual estimation to one of two periphyton abundance categories (high or low). Rocks that were difficult to assign to an abundance category were not used. This method tended to maximize the difference between low and high periphyton abundance rocks as determined by visual estimation only. Rocks were submerged in water in a bucket for a few minutes while being transported to a viewing box. Before being placed beneath a viewing box, the rocks were carefully examined and remaining invertebrates washed from the surfaces or removed with forceps. Rocks were exposed to the air for only a few minutes while removing invertebrates or assessing periphyton abundance. Five high resource and eight low resource rocks were collected. Rocks were placed beneath viewing boxes by late afternoon and allowed to become colonized beneath the viewing boxes for 24 hr before the number of nymphs was counted on their surfaces. Following the observation period, three periphyton samples were taken from each rock surface. Group means were compared with a t-test. In experiments E (June 1984), F (July 1984), and G (August 1984), a series of rocks was placed beneath the viewing boxes after invertebrates had been removed as previously described. An effort was made to represent the full range of periphyton densities but no attempt was made to divide rocks into low and high resource categories. Rocks were set in place by late afternoon or early evening and one observation of Baetisnumbers was made a few hours after rock placement if light permitted, and at 0600, 1200, and 1700 the next day, and at 0600 and 1200 the following day. The mean of the last three ob-

servations was used in statistical analysis. Water velocity a few centimeters above the surface of each rock was assessed with a small Ott C-1 current meter. Rocks were then placed on the streambank one by one and two periphyton samples were taken from their surface as previously described. Since diatom communities on smooth rock surfaces in Spring Run appeared relatively homogenous, we felt that the average of two samples would adequately describe periphyton abundance. Upper surface area was then measured before removing the next rock from the stream. The relationship between all variables and Baetis bicaudatus density was examined with a Pearson correlation analysis. Data were first compared with a normal probability plot to determine whether transformations were necessary. Immigrationrates Immigration rates were estimated directly by counting the number of individuals which arrived on the surfaces of solid bricks (10 x 10 x 6 cm) in 1-2-hr periods at the same time as experiments E, F, and G. Bricks had been placed in the stream at least 1 mo before experiments to become colonized with periphyton. At the start of an observation period, all invertebrates were removed from a brick and the brick was placed beneath a viewing box as previously described. Bricks were then photographed (see Wiley and Kohler 1981) at 5-s intervals using an 8-mm movie camera (Minolta XL601) positioned above the viewing box with a tripod. At the end of 1-2 hr, invertebrates were again removed from the brick and a new filming period begun. At the end of filming, one periphyton sample was taken from each brick. Filming took place between 1000 and 1600 hr because of lighting limitations. The number of Baetis arriving on the upper brick surface in the film period was determined by examining individual frames of the films sequentially with a Copal film editor. Since Baetis bicaudatusnymphs typically move <8 mm/min in Spring Run (Richards 1986), nymphs that appeared on the brick surfaces farther than 10 mm from an edge between 5-s exposures were judged to have drifted to the brick surface. The remainder were assumed to have crawled to the surface.

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

1988]

BAETIS DISTRIBUTION AND COLONIZATION

81

Results During all experiments, Baetis bicaudatuswas the most abundant invertebrate on rock surfaces. Neothremma (Trichoptera:Limnephilidae) and Drunella coloradensis (Ephemeroptera: Ephemerellidae) also were present but in much smaller numbers. Examination of all numeric data showed that transformations were not required to meet assumptions of the statistical test. Distributionamong patches on a rock Significant differences in Baetis bicaudatus density (p < 0.05) were noted between high resource (unscraped) and low resource (scraped) patches in all patchy rock experiments (Fig. 1). In all cases, the greatest Baetisdensities were on high resource patches. The magnitude of the difference between densities on high and low resource patches was greatest in experiments A and C when the net difference in chl a density was greatest. The ratio of chl a density between high and low resource patches in these experiments was 4.5 and 5.0, respectively. In experiment B, the chl a density ratio was lower (1.7) and the difference in Baetis density was not as pronounced (Fig. 1). No significant differences (p > 0.05) in Baetis bicaudatusdensity among the three patch sizes were noted in experiments A and C, however, a significant difference (p < 0.05) was noted in experiment B, where densities were greatest on 1.0-cm-width patches (Fig. 1). These results suggest that patch size had limited effects on Baetis bicaudatusdistributions. Distributionamong rocks In experiment D, high resource rocks had a significantly (p < 0.05) greater amount of chl a (mean ? 1 SE = 1.39 ? 0.11 /g/cm2, n = 5) than low resource rocks (0.33 ? 0.03 /tg/cm2, n = 8). Baetis bicaudatusdistribution reflected this difference. Baetisbicaudatusdensities (individuals/ cm2) were significantly (p < 0.05) greater on high resource rocks (0.29 + 0.04, n = 5) than on low resource rocks (0.19 ? 0.02, n = 8). Indicators of algal abundance also had the highest correlations with Baetis bicaudatusdensity in the other among-rock experiments (Table 1). The strongest correlations with Baetis bi-

E 0.100.00
_ __

6
'1

0.300.20-

B n-5 Exp.
n5

0.00

0.300.200.100.00 7 1t I Exp. C n=7 7

1.0 2.5 PATCH WIDTH (cm) FIG.1. Baetis bicaudatus density on high (open box) 0.5
and low (cross-hatched box) periphyton abundance patches on rock surfaces. Bars represent one standard error of the mean. n = number of replicate rocks with a patch width size. Periphyton abundances were (mean + 1 SE, Mg/cm2): exp. A high = 0.18 + 0.02, low = 0.04 ? 0.01; exp. B high = 0.20 + 0.3, low = 0.11 + 0.2; exp. C high = 0.34 + 0.04, low = 0.07 + 0.01. F values for treatment effects: exp. A periphyton = 92.38 (p < 0.001), width = 1.70 (p > 0.2); exp. B periphyton = 8.33 (p < 0.01), width = 6.89 (p < 0.01); exp. C periphyton = 62.79 (p < 0.001), width = 0.62 (p > 0.5).

caudatus density in all three among-rock experiments (experiments E, F, and G) were with chl a density. Lower correlations were observed with ash free dry weight of algae in experiment E and G. Other indicators of food abundance or quality had no significant relationships with Baetis bicaudatusdensity (Table 1). Of two physical variables measured, only velocity had a significant correlation with Baetis density (experiment G). No significant relationship was observed with surface area. These results suggest that chl a was a stronger predictor of Baetis bicaudatusdensity than ashfree dry weight, chl a: pheophytin ratio, velocity, or rock surface area. However, the physical variables examined had relatively small ranges which may have downplayed the importance of these variables (Table 1). The chl a-Baetis relationship was described by regression analysis for each among-rock experiment (Fig. 2).

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

82
Y=0.073+0.22X r=0.63 p=0.001
o 0 0 0 _-Q
o
o 0

C. RICHARDS AND G. W. MINSHALL

[Volume 7

0.24 0.16 0.08 0.00

p Exp.
0 Co

c^ m in 10
dt b cs N

r-

0 O

dc5 dc 6c c5 aC5I I

%
I

o
I I I

(d
(0 CM 0)
r tL Lr 0 O'-4 C) d O \t r

C'4 C 0 r
'-4

"0.24 d z0.16 j! 0.08

Y=0.01 4+0.24X r=0.64 p=0.01

Exp.F
0

0
CZ

0
0

Co00
0) :-

C666CoCo

0~ CM4 0~ 4

?
o

o
0
II

*a a.

;j4
c-l

OCO
C o

N O
o C4

No

z 0.00

O
0

0o-0

0.24 0.160.080.00

Y=0.01 2+0.21 X r=0.75 p=0.002

Exp. G

0 3) 0) (0 0) x (t Z4
'.1

.:

CNI 00

x a

0o0o0 o

Co6 0

0Oo
o

o o
0

If

C
'i-

0 CN o gC) NC C 0 N C0 CN I CC N C\l .:o< 0 CI t1 CM 0 00

0. 0

0.2 0.4 0.6 0.8 a ( ug / cm 2) CHLOROPHYLL

aI
c0 0)

1-4

Baetis bicaudatus density on rock surfaces versus chl a abundance in experiments E (June), F (July), and G (August).

FIG. 2.

N Co Cf

CM

'-~ 1f N

rs
01

04

Owing to the lack of intermediate chl a values in experiments E and F, these regressions should be treated with some caution. Regressions were compared by analysis of covariance (Sokal and Rohlf 1981). This procedure showed that there was no significant difference among the slopes of the three regression lines (F = 0.043, p > 0.50). There was, however, a significant difference among the origins of the regressions (F = 11.58, p < 0.001). The common slope of the regressions was 0.2174. These data show that the functional relationship between Baetis bicaudatusdensity and chl a was consistent through the summer in spite of variations in absolute densities. Immigration Immigration rate (number/100 cm2/hr) was highest in experiment E and lowest in experiment G (Table 2). Over 50% of the immigrants

Cl

\ c) aO C C 0\ ~N1 M oN dJ CO (N

0 II

I
Q) 4

'a

00

o Co

0 oo00 cr oo O'
LC r-o

o6
N

cu
*i

r-i C c5

O -CoO C

C0

(OINc

CU

&< P?e

ON 0 00 N1

Nt, O 0 00 Co
10

Co o o
C4

\6d tN
00 \0

b0

(C

ON

'-4 rJ

a'
In

0)
1-4

'
m

a
c;
u b^
> o^

6 V
uj

Eu
X go

0.. CO CU
C)-

in each experiment were from the drift. The

43 0) =

>

~ 04'a.,: ? ?> 0~
g4. 14

a1

d:

UUk3~~~~~0 U)

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

1988]

BAETIS DISTRIBUTION AND COLONIZATION

83

among-rock experiments could develop rapidly. Assuming no emigration from the rocks, < 12 hr would be required to reach even the highest observed densities. Discussion

TABLE2. Immigration rate of Baetis bicaudatus (number/100cm2/hr),percentageof total immigrants that arrivedfrom the drift, and chl a abundance(A/g/ cm2)on brick surfacesduring experiments E, F, and G. n = number of observationsof a brick surface. %

Chl a Experiment n Periphyton abundance had a significant effect on Baetis bicaudatusdistribution in Spring E (June) 8 5.84 + 0.78 52 0.305 Run when observed on two spatial scales. GreatF (July) 5 4.00 + 0.55 62 0.156 er densities of B. bicaudatusaccumulated on subG (August) 5 2.31 + 0.31 59 0.545 strates with high periphyton densities than on substrates with lower periphyton densities. These relationships developed within a short Dodds and Baetisbicaudatus.Once on a substrate, period of time. This information suggests that nymphs have the ability to assess food abundance with mouthparts and other sensory strucnewly arriving nymphs recognize relative peabundance and their behavior tures and can respond accordingly. In high reriphyton modify to increase the amount of time spent in high source patches, movement velocity decreases food resource areas. Clearly, in Spring Run unwhile rate of turning increases, effectively conder the conditions tested, periphyton abuncentrating foraging in high resource areas. This dance had more influence on B. bicaudatusdis- behavior should be effective among patches of tribution than did patch size or water velocity. any size. However, when patches are large and Differential densities along periphyton abunsearch time (and presumably energy expendidance gradients likely are due to a combination ture) within and among patches is greater, other of events that occur after a nymph arrives on behaviors such as active drift may be invoked. substrate and not before contact with the sur- Since a large portion of immigrants to vacant face. Although the swimming ability of this gebricks in this study were judged to have crawled nus is well noted (Ciborowski 1983, Elliott 1971a, onto the bricks from adjacent areas, it can be Minshall and Winger 1968), the effectiveness of assumed that area restricted search does have a this ability is greatest in slow water regions near role in distribution when the observation scale the shores (Ciborowski 1987). All experiments is rocks within a small section of stream. Benthic in this study were conducted in high velocity movements of several meters per day have been water where the possibility of continuously documented with Baetisrhodani(Pictet) and other mayflies (Elliott 1971b). Movements of such monitoring substrate and food abundance conditions while drifting in the water column are scale would be sufficient to search a number of reduced. There was little difference in substrate rock surfaces within a day for resource rich areas. The short amount of time required for B. biroughness or texture among rocks that might give microhydraulic clues to food abundance or caudatus to adjust its abundance relative to peprovide enhanced opportunity for settling. Im- riphyton abundance gradients and the high rates migration from the drift to substrate patches in of immigration observed indicate that emigrahigh velocity areas is most likely a random pro- tion from habitat patches must be directly recess in which the rate of immigration is pro- lated to periphyton abundance. This relationportional to the abundance of nymphs in the ship probably resembles the common slope drift. Although differential immigration from observed in the regression analyses. A large the drift to high resource areas was not exam- proportion of newly arriving immigrants would ined in this study, there is no published evihave to emigrate through either crawling or dence to suggest it exists. drift in order to maintain the densities observed After arriving on a rock, area restricted search in this study even within a few hours of rock (Krebs 1978) is a sufficient mechanism to pro- placement. duce greater densities in patches of high food Actual immigration to and emigration from abundance. Kohler (1984, 1985) and Richards individual substrates has rarely been assessed in streams owing to difficulties inherent in ob(1986) have described this form of food-related directional movement with Baetis tricaudatus serving stream substrates. Immigration rates for

from Mean + 1 SE Drift

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

84

C. RICHARDSAND G. W. MINSHALL 1=5.84

[Volume 7

E30.
0

1=4.00 =

/
/,

--------------

1=2.31

C) COr v,
LUi

12

18

24

30

36

TIME (hours)
FIG.3. Theoretical colonization rates from the rate, m = proportion emigrating per unit time, t = time, and e = base of Napierianlogarithms(Sheldon 1984). Immigrationrates are those observed in immigrationexperiments. Baetis obtained in this investigation are considerably higher than those reported in other studies, although there are few studies with which to compare. Wiley and Kohler (1981) reported an immigration rate of 26 Baetis per 100 cm2 in a 48-hr period. The turnover (proportion emigrating) in this period was 84%.Resource abundance was not reported. Their estimates were also based upon time-lapse photography, but may have been somewhat low due to the time between film exposures (30-60 s). The highest emigration values reported from sequential removal studies are those of Ciborowski and Clifford (1984) who calculated values of 22-69% leaving per 24 hr. Estimates of immigration based upon those values and observed densities would range from 14 to 44 individuals per 100 cm2 per 24 hr. The immigration rates observed in our study (55-140 per 100 cm2 per day) are considerably higher than either of these other studies. If the number of B. bicaudatuson a rock at the end of a 24-hr period had been used as an actual immigration rate similar to that used in a sequential removal study, values at least an order of magnitude lower would have been derived. Comparisons with existing colonization studies are further complicated by potential differences in the turnover rates on single large substrates and the baskets or trays that are typically used in such studies. These trays contain numerous small particles. Small individual particles may exhibit greater turnover rates because of their greater relative border area. Such an effect may influence observed turnover rate. Even these estimates of immigration and emigration may be low when compared with those
equation N = I/m(l-e-mt), where I = immigration

that occur after sunset. Baetis typically exhibits drift peaks during evening hours (Hynes 1970, Waters 1965). Assuming that drift is related to immigration on rock surfaces, the numbers of individuals arriving on a rock after dark would be greater than during daylight hours. The effect of this on densities, however, might be negated by increased emigration rates. At least one investigator (Kohler 1984) has observed greater movement velocities on substrates at night. The net effect of these two phenomena might be to increase the turnover of individuals after dark. The differences in Baetisbicaudatusdensity observed at similar periphyton densities among months may be caused by differing immigration rates among the months. Relatively small differences in immigration rates can cause significant changes in what would be observed as equilibrium density, assuming emigration remained constant (Sheldon 1984). This situation is illustrated in Figure 3. The elevation of regression lines observed in this study is in general agreement with this prediction. Baetis bicaudatus densities were greatest when immigration rates were highest. Although field experiments in our study were not designed to assess equilibrium density, significant differences can be observed long before equilibrium is achieved. The results of this study show that colonization studies in streams must consider the potential immigration rates of species and the rate at which reaction to environmental conditions occurs in the selection of census intervals. For species such as highly mobile grazers, careful attention also must be paid to algal abundance since emigration rates are strongly influenced by this factor. If substrates used in colonization studies are initially sterile, the equilibrium density achievable at any point in time may be driven more by the growth and accumulation of food resources than by immigration rate. The relatively long periods implied in the colonization curves depicted for Baetis by Ciborowski and Clifford (1984) and Sheldon (1977) may have been influenced by this factor. This study strengthens the view that grazers in streams are acutely aware of and respond rapidly to algal abundance. Furthermore, responses can be observed at food abundance levels that occur naturally within a stream. The relationship between grazers and their food re-

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

1988]

BAETIS DISTRIBUTION AND COLONIZATION

85

source may be even more dynamic than previously thought. Acknowledgements

of mayflies (Ephemeroptera). Canadian Journal of Zoology 55:1970-1977.

R. L. VANNOTE.1983. Periphyton, chlorophyll a, and diatoms of the Middle Fork of the Salmon River, Idaho. Holarctic Ecology 6:221-227. J. Anderson, J. R. Barnes, J. Griffith, C. P. DOUGLAS, B. 1958. The ecology of the attached diaHawkins, O. H. Ore, and P. L. Lawson made toms and other algae in a small stony stream. many helpful suggestions on earlier versions of Journal of Ecology 46:295-355. this manuscript. J. D. Allen, J. J. H. Ciborowski, EDINGTON, J. M. 1968. Habitat preferences in netS. L. Kohler, and R. J. Mackay also provided spinning caddis larvae with special reference to critical improvements. C. T. Robinson assisted the influence of water velocity. Journal of Animal in collection of field data. This research was Ecology 37:675-692. ELLIOTT, J. M. 1971a. The distances traveled by drifta Xi of resupported by Sigma grant-in-aid ing invertebrates in a Lake District stream. Oecosearch, the Idaho State University Research logia 6:191-220. Committee, and the Idaho State University DeELLIOTT, J. M. 1971b. Upstream movements of benpartment of Biological Sciences. thic invertebrates in a Lake District stream. Journal of Animal Ecology 40:235-252. S. V. 1983. Plant-herbivore interactions GREGORY, Literature Cited in stream systems. Pages 157-190 in J. R. Barnes D. D., AND G. W. MINSHALL.1978. Lonand G. W. Minshall (editors). Stream ecology: apANDREWS, gitudinal and seasonal distribution of benthic inplication and testing of general ecological theory. vertebrates in the Little Lost River, Idaho. AmerPlenum Press, New York. ican Midland Naturalist 102:225-236. D. D. 1981. Foraging and resource patchiness: HART, APHA. 1980. Standard methods for the examination field experiments with a grazing stream insect. of water and wastewater. 15th edition. American Oikos 37:46-52. Public Health Association, Washington, D.C. HART,D. D. 1983. The importance of competitive interactions in stream systems. Pages 99-136 in J. BEHMER,D. J., AND C. P. HAWKINS. 1986. Effects of overhead canopy on macroinvertebrate producR. Barnes and G. W. Minshall (editors). Stream tion in a Utah stream. Freshwater Biology 16:287ecology: application and testing of general eco300. logical theory. Plenum Press, New York. H. W. 1978. Beziehungen zwischen dem BOHLE, HART,D. D. 1985. Causes and consequences of terNahrungsangebot, der Nahrungsangebot, der ritoriality in a grazing stream insect. Ecology 66: Drift und der raumlichen Verteilung bei Larven 404-414. von Baitis rhodani (Ephemeroptera:Baetidae). ArHART, D. D., AND V. H. RESH. 1980. Movement patchiv fur Hydrobiologie 84:500-525. terns and foraging ecology of a stream caddisfly T. L., J. T. BROCK,C. S. DUNN, R. J. NAIMAN, R. larva. Canadian Journal of Zoology 58:1174-1185. BOTT, W. OVINK,AND R. C. PETERSEN. 1985. Benthic C. P., M. L. MURPHY, ANDN. H. ANDERSON. HAWKINS, 1982. Effects of canopy, substrate composition, community metabolism in four temperate stream and gradient on the structure of macroinvertesystems: an interbiome comparison and evaluation of the river continuum concept. Hydrobiolbrate communities in Cascade Range streams of ogia 123:3-45. Oregon. Ecology 63:1840-1856. CIBOROWSKI, J. J. H. 1983. Influence of current veHEMPHILL, N., ANDS. D. COOPER.1983. The effect of locity, density, and detritus on drift of two mayfly physical disturbance on the relative abundances of two filter-feeding insects in a small stream. species (Ephemeroptera). Canadian Journal of Zoology 61:119-125. Oecologia 58:378-382. CIBOROWSKI, J. J. H. 1987. Dynamics of drift and HILL,W. R., ANDA. W. KNIGHT.1987. Experimental microdistribution of two mayfly populations: a analysis of the grazing interaction between a predictive model. Canadian Journal of Fisheries mayfly and stream algae. Ecology 68:1955-1965. and Aquatic Sciences 44:832-845. H. B. N. 1970. The ecology of running waters. HYNES, CIBOROWSKI, Univ. Toronto Press, Toronto. J. J. H., AND H. F. CLIFFORD. 1984. Shortterm colonization patterns of lotic macroinverS. L. 1966. The mayflies of Idaho. M.S. TheJENSEN, tebrates. Canadian Journal of Fisheries and sis, University of Utah, Salt Lake City. S. L. 1984. Search mechanisms of a stream Aquatic Sciences 41:1626-1633. KOHLER, CORKUM, L. D., P. J. POINTING, AND J. J. CIBOROWSKI. grazer in patchy environments: the role of food 1977. The influence of current velocity and subabundance. Oecologia 62:209-218. strate on the distrubution and drift of two species S. L. 1985. Identification of stream drift KOHLER,

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions

86

C. RICHARDSAND G. W. MINSHALL

[Volume 7

mechanisms: an experimental and observational approach. Ecology 66:1749-1761. KREBS, J. R. 1978. Optimal foraging: decision rules for predators. Pages 23-63 in J. R. Krebs and N. B. Davies (editors). Behavioral ecology: an evolutionary approach. Blackwell Scientific Publications, Oxford. G. A., AND V. H. RESH. 1983. Stream peLAMBERTI, riphyton and insect herbivores: an experimental study of grazing by a caddisfly population. Ecology 64:1124-1135. M. S., AND G. E. LIKENS.1987. The imporMAYER, tance of algae in a shaded headwater stream as food for an abundant caddisfly (Trichoptera). Journal of the North American Benthological Society 6:262-269. MCARTHUR, J. V., AND J. R. BARNES.1985. Patterns of macroinvertebrate colonization in an intermittent Rocky Mountain stream in Utah. Great Basin Naturalist 45:117-123. MCAULIFFE, J. R. 1983. Competition, colonization patterns and disturbance in stream benthic communities. Pages 137-156 in J. R. Barnes and G. W. Minshall (editors). Stream ecology: application and testing of general ecological theory. Plenum Press, New York. MCAULIFFE, J. R. 1984a. Competition for space, disturbance, and the structure of a benthic stream community. Ecology 65:894-908. MCAULIFFE, J. R. 1984b. Resource depression by a stream herbivore: effects on distribution and abundance of other grazers. Oikos 42:322-334. G. W. 1984. Aquatic insect-substratum MINSHALL, relationships. Pages 358-400 in V. H. Resh and D. M. Rosenburg (editors). The ecology of aquatic insects. Praeger Publishers, New York. G. W., AND J. N. MINSHALL. 1977. MicroMINSHALL, distribution of benthic invertebrates in a Rocky Mountain (U.S.A.) stream. Hydrobiologia 55:231249. 1985. Towards G. W., ANDR. C. PETERSEN. MINSHALL, a theory of macroinvertebrate community structure in stream ecosystems. Archiv fiir Hydrobiologie 104:49-76. G. W., ANDP. V. WINGER.1968. The effect MINSHALL, of reduction in stream flow on invertebrate drift. Ecology 49:580-582. R. Y., J. V. MCARTHUR, OBERNDORFER, J. R. BARNES, AND J. DIXON. 1984. The effect of invertebrate predators on leaf litter processing in an alpine stream. Ecology 65:1325-1331. B. L. 1980. Predator-prey interactions PECKARSKY, between stoneflies and mayflies; behavioral observations. Ecology 61:932-943. B. L., AND S. I. DODSON. 1980. An exPECKARSKY, perimental analysis of biological factors contributing to stream community structure. Ecology 61: 1283-1290.

S. R. 1980. The role of substratum in benthic REICE, macroinvertebrate microdistribution and litter decomposition in a woodland stream. Ecology 61: 580-590. S. R. 1985. Experimental disturbance and the REICE, maintenance of species diversity in a stream community. Oecologia 67:90-97. C. 1986. Distribution and foraging beRICHARDS, havior of a grazing stream mayfly (Baetis bicaudatus). Ph.D. Dissertation, Idaho State University, Pocatello. 1986. Effects C. T., ANDG. W. MINSHALL. ROBINSON, of disturbance frequency on stream benthic community structure in relation to canopy cover and season. Journal of the North American Benthological Society 5:237-248. SHAW,D. W., AND G. W. MINSHALL.1980. Colonization of an introduced substrate by stream macroinvertebrates. Oikos 34:259-271. A. L. 1977. Colonization curves: applicaSHELDON, tion to stream insects on semi-natural substrates. Oikos 28:256-261. A. L. 1984. Colonization dynamics of SHELDON, aquatic insects. Pages 401-429 in V. H. Resh and D. M. Rosenberg (editors). The ecology of aquatic insects. Praegar Publishers, New York. R. R., AND F. J. ROHLF. 1981. Biometry. 2nd SOKAL, edition. W. H. Freeman and Company, San Francisco. 1971. PeSTOCKNER, J. G., AND F. A. J. ARMSTRONG. riphyton of the experimental lakes area, northwestern Ontario. Journal of the Fisheries Research Board of Canada 28:215-229. C. C. 1986. The role of periphyton abunVAUGHN, dance and quality in the microdistribution of a stream grazer, Helicopsyche borealis (Trichoptera: Helicopsychidae). Freshwater Biology 16:485-493. WALLACE,J. B., AND M. E. GURTZ.1986. Response of Baetis mayflies (Ephemeroptera) to catchment logging. American Midland Naturalist 115:2541. T. F. 1965. Interpretation of invertebrate WATERS, drift in streams. Ecology 46:327-334. M. J. 1981. Interacting influences of density WILEY, and preference on the emigration rates of some lotic chironomid larvae (Diptera:Chironomidae). Ecology 62:426-438. 1981. An assessment M. J., AND S. L. KOHLER. WILEY, of biological interactions in an epilithic stream community using time-lapse cinematography. Hydrobiologia 78:183-188. M. J., AND S. L. KOHLER.1984. Behavioral WILEY, adaptations of aquatic insects. Pages 101-133 in V. H. Resh and D. M. Rosenberg (editors). The ecology of aquatic insects. Praeger Publishers, New York. Received: 8 October 1986 Accepted: 26 April 1988

This content downloaded on Wed, 6 Mar 2013 19:49:02 PM All use subject to JSTOR Terms and Conditions