Great Basin Naturalist Volume 54 Number 1 Article 2 2-25-1994 Effects of cobble embeddedness on the microdistribution of the sculpin Cottus beldingi and its stonefly prey Roger J. Haro College of Agriculture, University of Idaho, Moscow, Idaho Merlyn A. Brusven College of Agriculture, University of Idaho, Moscow, Idaho Follow this and additional works at: https://scholarsarchive.byu.edu/gbn Recommended Citation Haro, Roger J. and Brusven, Merlyn A. (1994) "Effects of cobble embeddedness on the microdistribution of the sculpin Cottus beldingi and its stonefly prey," Great Basin Naturalist: Vol. 54 : No. 1 , Article 2. Available at: https://scholarsarchive.byu.edu/gbn/vol54/iss1/2 This Article is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Great Basin Naturalist by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. Great Basin Naturalist 54(1), © 1994, pp. 64-70 EFFECTS OF COBBLE EMBEDDEDNESS ON THE MICRODISTRIBUTION OF THE SCULPIN COTTUS BELDINGI AND ITS STONEFLY PREY Roger J. Haro1,2 and Merlyn A. Brusven1,3 AlwrnACT.-Laboratory experiments were undertaken to assess the effects of three levels of cobble embeddedness on the microdistribution of the sculpin Cottus beldingi and its stoncfly prey, Skwala americana. Experiments were con­ ducted separately and together as predator and prey in temperature- and flow-controlled artificial streams. Whcn tested either separately or together, both the predator sculpin and its stoneHy prey occurred in significantly greater number,~ on substrata having unemhedded cobbles than substrata having half- or completely cmbedded cobbles. Sloncfly densi­ ties were greater in substrata having unembedded cobbles even though predator densities within the more embedded cobble patches were significantly lowcr. These findings support the hypothesis that higher predator densities influence prey den"itie" less than the structural habitat quality oflInembedded-cobhle patches. Key words: predator-prey, cobble embeddedness, nonletfwl effects, stonejlies, sculpins, nonpoint source sedimenta­ tum, COUllS beldingi. Reduced summer flows and increased sedi­ that, under laboratory conditions, Paragnetina mentation in many western North American media (Plecoptera, Perlidae) selected larger streams may significantly diminish the size substrata over smaller ones in the presence of and availability of adequate microhabitat rainbow trout. patches for benthic fish and insects. Sedimen­ Sculpins hold a significant position in the tation from agricultural sources has been food web of Pacific Northwest stream commu­ linked to pronounced changes in the trophic nities and have been shown to reduce food structure of lotie fish assemblages (Berkman resources, food consumption, and the produc­ aud Rabeni 1987) and may affect macroinvcr­ tion of trout (Brocksen et aJ. 1968). Cottus tebrate community structure, further altering beldingi, the Paiute sculpin, is the most abun­ trophic relations within the lotic food web. dant fish species in Lapwai Creek (Kucera et Such trophic changes, in part, may result aJ. 1983), the stream investigated in this study. from alterations in prey rcfugia brought about This ambush predator feeds almost exclusive­ by the embeddedncss of cobble snbstrata. lyon benthic macroinvertebrates (Johnson Brusvcn and Rose (1981) found that cobble 1985). Finger (1982) reported that adult C. embeddedness significantly influeuccd the beldingi preferred coarse-grained substrata in vulnerability of two insect predators, Hesper­ an Oregon stream. In California, Card and operla pacifica (Plccoptcra, Pcdidae) aud Rhy­ Flittner (1974) found the highest densities of acophila vaccua (Trichoptera, Rhyacophili­ C. beldingi in rubble or gravel substrata. dae), to predation by Coitus rhotheus. They Skwala americana (Plecoptera, Perlodidae) is suggested high sculpin predation success in a common lotic stonefly found throughout west­ the embedded substrata was due to the loss of ern North America (Baumann et aJ. 1977). macroinvertebrate refugia under cobbles. Nymphs are important prey of C. beldingi Microhabitat shifts by macroinvertebrate (Johnson 1985) and are normally found in rel­ prey in response to vertebrate and macroin­ atively unsedimented, unembedded-cobble vertebrate predators have been reported by riffles (Short and Ward 1980). Tbey commonly several workers (Stein and Magnuson 1976, feed on small mayflies and midges (Fnller and Stein 1977, Peckarsky and Dodson 1980, Stewart 1977, Richardson and Gaufiu 1971). Peckarsky 1983). Feltmate et al. (1986) found Skwala americana is univoltine, with adult j Department ofPlnnt, Soil a",1 Entomologieal Scien,"",,,, Collogo ofAgl'icnltun,. Univon;;ly ofIdaho, Mom:Jw. Idaho 1'3844-2.3.19. 21'n,,'<ont addro",", :':d... ~,l of Nalllral HC"OUl\.-'C', UIlivcr"ily of Michig~Jl, Ann Arho,·, Michi"an -IH} 09. ·'S(,nd oomrllullieati",,-, and !'Oprint 1"0(11l05t, 10 thi" autl... ,r. 64 1994] PREDATOR AVOIDANCE AND STREAM SEDIMENTATION 65 emergence occurring from February through oratory. The streams were partitioned into July (Baumann et al. 1977). three 0.19-m2 sections bordered by a 0.04-m2 The purpose ofthis study was to determine sand area at each end (Fig. 1). the microdistribution of C. beldingi and UnchJorinated tapwater was recirculated Skwala americana, both separately and through tbe channels by electrical pumps together as predator and prey, when given (mean velocity, 2.44 cm s·I). Water tempera­ choices among three levels of cobble embed­ ture was maintained at 13.9°C with thermo­ dedness in an artificial stream. static refrigeration units placed within recir­ culating sumps. Water depth was held at 13.0 MATERIALS AND METHODS cm above the base substrate, thus assuring that all cobbles were fully submersed. Field Collection Substrate Treatments Lapwai Creek (46' 18'N, 116'43'W) drains Natural stream cohbles (62.0-127.0 mm, agricultural land in the Columhia River Basin principal axis) were collected from Lapwai 20 km east of Lewiston, Idaho. Field esti­ Creek. Cobhles were scrubbed with a hrush mates of sculpin density were made in Sep­ under hot tap water, dried, and labeled wi!b tember 1987 from riffles in Lapwai Creek and an identification number and orientation its tributaries. Sculpins were frightened into a arrow to enable replication of their spatial drift net attached to two wire-mesh side­ arrangement in either an unembedded or wings that sampled a rectangular 0.44-m2 half-embedded condition. Fifteen cobbles area. Adult sculpins (66.1 ± 9.4 mm) used in were randomly positioned in each of !be test the experiment were collected during Sep­ sections of the stream except for the simulat­ tember and October 1987. ed, fully embedded condition, which had no Late-instar stonefly nymphs (Skwala amer­ visible surface presence of cobbles. The cob­ icana) were collected with kick nets from !be bles covered ca. 60% of the two-dimensional same riffles from which sculpins were collect­ area of the unembedded and 50%-embedded ed. The nymphs were transported to the labo­ sections. ratory. To simulate a 50%-embedded condition, Preconditioning of Test Organisms balf ofeacb cobble's principal axis was c-ast in plaster of paris. The casts were later Riled Sculpins and stoneflies were acclimated in with a mixture of concrete and natural stream laboratory streams like those described by sand. After drying, the "half-casts were tex­ Brusven (1973). Ccbbles were placed in the tured with a wire brush and assigned an iden­ streams to provide cover, All organisms were tification number and directional arrow that acclimated for at least 48 h in holding streams corresponded to their natural-rock counter­ before each experiment. Sculpins were ran­ parts. domly selected 24 h before each expeliment, Two centimeters of washed stream sand isolated, and starved to better elicit hunger was spread over the bottom of eacb artiRcial and foraging behavior. stream to serve as a base substratum. Three cobble-embeddedness conditions were ran­ Experimental Stream Channels domly assigned among three stream sections. Two Plexiglas® streams (3.35 X 0.25 x In one section cobbles were placed on top of 0.20 m) were arranged side by side in the lab- the sand. In another section the "balf-casts" 1111 • SlIIIII 11.1 • SIIIIII 1.­ Ill­ Ilt. ''''' Fig. 1. Schematic diagram of an artificial channel .showing the stream sections and one cobble-cmbeddedness llrrangement: (U) = unemhedded cobbles. (II) = 5Q%.emhedded cobhles, (S) = lOO%·embedded cobbles. and (B) = sand buffer zooes. 66 GREAT BASIN NATURALIST [Volume 54 were positioned in the identical arrangement (i.e., upstream, midstream, and downstream). and orientation as their natural unembedded Each possible cobble arrangement was repli­ counterparts. They were placed directly on cated randomly in time (three times) for a the sand, thereby limiting access to their total of nine trials for each organism. A non­ under surfaces by the organisms studied. The parametric Kruskal-Wallis test with a posteri­ remaining section was left as a 2.0~cm layer of ori pairwise comparisons (Conover 1980) was sand and simulated a 100% cobble-embedded used to detect significant differences among condition with no cobbles evident on the sur­ mean-ranked numbers of organisms recov­ face. ered from three cobble-embeddedness condi­ Experimental Trials tions (Conover 1980). PREDATOR AND PREY-INTERACTIVE.-This