Great Basin Naturalist
Volume 52 Number 1 Article 2
5-11-1992
Secondary production estimates of benthic insects in three cold desert streams
W. L. Gaines Central Washington University, Ellensburg, Washington
C. E. Cushing Pacific Northwest Laboratory, Richland, Washington
S. D. Smith Central Washington University, Ellensburg, Washington
Follow this and additional works at: https://scholarsarchive.byu.edu/gbn
Recommended Citation Gaines, W. L.; Cushing, C. E.; and Smith, S. D. (1992) "Secondary production estimates of benthic insects in three cold desert streams," Great Basin Naturalist: Vol. 52 : No. 1 , Article 2. Available at: https://scholarsarchive.byu.edu/gbn/vol52/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 52(l), 1992, pp. 11-24
SECONDARY PRODUCTION ESTIMATES OF BENTHIC INSECTS IN THREE COLD DESERT STREAMS
W L. Gaine/> 2, C. E. Cushing3, and S. D. Smith1
ABS111.ACf.-We studied aquatic insect production in three cold desert streams in southeastern Washington. The Size-Frequency (SF) and PIB methods were used to assess production, which is expressed by taxon, functional group, and trophic level. Dipterans (midges and black f1ies) were the most productive taxa, accounting for 40-70% ofthe total insect production. Production by collectors and detritivores was thegreatestofall functional groups and trophic levels, respectively, in all study streams. Insects with rapid development times and multiple cohorts are vmy important in cold desert streams; they were major L contributors to the total insect production. Total insect production rates in our study streams (14-23 g DW-m-z.yr- ) were greaterthan those found in Deep Creek, Idaho (1.2 g DW-m -2'yr-1), the only other cold desert stream for which production data are available. Our values also were generally greater than published datu for most cold/mesic (3-27 g DW"m"2'yr"1) and humid/mesic (3--2.5 g DW·m-2'yr-1) streams, but lower than in Sonoran Desert Stream:, (>120 g DW'm-2')'1"-1) or New Zealand streams (~40 g DW_m-2'yr"1). Our (lata support the contention of others that production, rather than density or biomass, is the most accurate and meaningful way to assess the role of these organisms in lotic ecosystems.
Key words: productivity, benthos, spring-stremTk~, cold desert, funcUunal groups, trophic levels, Diptem., Trichoptem, Coleoptera, f:phemeroptem, Odonata, Plecoptera.
Community-level production of insects has mass analysiS and overestimated by numerical been assessed in relatively few stream types, and analysis in a southeastern stream (Benke et a1. of all macroinvertebrates in even fewer_ Partic 1984). Waters (1977) states that production is ularly, little is known about secondary produc important to understanding ecosystem dynam tion in arid region streams, The only studies of ics because it is the means by which energy is secondary production in arid region streams made available to higher trophic levels. that we are aware ofare those ofMinshall et aI. While most secondary production studies (I973) in Deep Creek, Idaho, in the cold desert have focused on one or a few species in a stream province, and Fisher and Gray (1983) and Jack (Benke and Wallace 1980, Waters and son and Fisher (1986) in Sycamore Creek, Ari Hokenstrom 1980, O'Hop et a1. 1984), more zona, in the hot desert region, recent studies have estimated secondary pro Secondary production is the rate of animal duction of the entire macrobenthic fauna tissue elaboration over time regardless of the (Kmeger and Waters 1983, Benke et a1. 1984, fate (e.g., canlivory, emergence) ofthat produc Smock et .1. 1985, HUlyn and Wallace 1987). tion (Benke and Wallace 1980). Estimating sec Yet to be integrated into community-level anal ondary production in a stream provides one yses are the hyporheic fauna, protozoa, and assessment ofthe role ofanimals in the ecosys othermicroinvertebrates_ The community-level tem (Benke andWallace 1980) as well as insight approach prOvides a more integrated insight into ecosystem dynamiCS. Estimating only den into the ecology of stream ecosystems, sity and biomass, regardless of time, may not The purpose ofthis study was to measure the accurately describe the role oforganisms in the secondary productionofinsects in threestreams stream. For instance, the role of gathering-col located in the cold desert phYSiographic prov lectorinvertebrates was underestimatedhy bio- ince ofsoutheastern Washington, We emphasize
I D"partm"n! of Biological Sciellces, C"mtral Washington Univt 11 12 GREAT BASIN NATURALIST [Volume 52 T"HLE 1. Physical and chemil~1 characteristics of study rt:acht:s in Douglas Creek, Snively Springs. nnd Rattlesnake Springs, July 19..'3.5 to June:: 19S6. Average AVt':mge Avemge width delxh disch.'lrge Dissolved 02 3 Stream (m) (Ill) (1I1 /s) (mgtL) Douglas Creek 4.0 o.:n 0.6 9.6--14 Snively S~rings 1.3 0.10 0.04 8.6-12 Hl.lulesna c Springs 1.7 0.05 0.05 8.2-10 TABLE 2. Percent sllhstratull1 t)11es in study reaches of Dnllglas Creek, Snively Springs, and Rattlemake Springs, July HJ85 to June 1986. Substratum type Boulder Cobble Pebblt: Crovel SnnUisilt Stream (>256 mOl) (64-225 mm) (16-64 mm) (2-16 mm) «2 mrn; Douglas Creek 21 29 24 16 10 Snively Springs 7 20 25 11 37 Rattlesnake Springs 0 1 7 11 81 that the estimates published here are, in several the upper reaches where flow is pennanent and cases, based on Fjg. 1. Study reaches: A, Douglas Creek; B, Snively Springs; C, Rattlesnake Springs. 14 GREAT BASIN NATURALIST [Volume 52 20 ,..------",---c-..,.,,---,------, Douglas Creek o Daily High o Daily Low 10 5 2g ~==~='=='======: Snively Springs ~ ()o 15 ~ 2~~=='==~====:'===='===~ Rattlesnake Springs 20 15 10 J A SON D J FMAM J 1985 1986 Fig. 2. Annual water temperature regimes: Douglas Creek, Snively Springs, and Rattlesnake Springs, July 1985 to June 1986. Watercress is presently the dominant in-stream METHODS autotroph, although periphytou primary pro duction exceeded that ofwatercress in 1969-70 We sampled segments of each stream repre (Cushing and Wolf 1984). senting the various habitats that were present. 1992J INSECf PRODUCI1Vm IN SPRING-STREAMS 15 One study reach was sampled in SS and one in sources ofinfonnation to estimateCPls for each RS, and three reaches were sampled in the taxon in OUf study streams. First, we surveyed larger DC. Samples were taken to calculate an the available life-history data gathered from lit average standing stock for each stream to he erature reviews and estrapoJated the results to used to calculate production estimates. The apply to our situations. Second, we made field 'Mmpling scheme was not designed to allow observations to determine presence/absence of intrastream comparisons of production esti taxa and collected size-frequency information mates between different habitats, but rather to for each taxon to estimate larval development provide representative production estimates of times and CPls. Lastly, we conducted in situ the entire stream. growth studies for Baetis sp., Chemnatopsyche Samples were collected monthly from July sp., and Simulium sp. to aIlaw further refine 19&5 through June 1986. We collected three ment ofour CPI estimates. These growth stud samples during each visit. A Portable Inverte ies involved placing insects within growth 2 brate BOK Sampler (PIBS) (0.1 m , mesh size chambers in R5. Chamhers were consttucted 350 ~mJ was used in DC. A Surber sampler with mesh nelling on each end to allow water (0.09 m , mesh size 350 ~m) was used in SS and and food material to pass through. Measure RS hecause these spling-streams are too shal ments were taken and development times low for a PIBS. Samples were taken to a depth recorded to estimate CPls. Using the combina of 10 cm and preserved in 70% ethyl alcohol. tion of all these data sources, we feel confident Insects were separated from organic debris by that our CPI estimates are reasonable approxi sugar flotation (Andersan 1959) and sorted by mations. taxa. Insects were identified to the lowest taxo ProductianlBiomass (PIB) ratios (Waters nomic level possible and counted, and body 1977) were used to estimate secondary produc length was measured to the nearest 1 mm using tion for less-abundant taxa. These PIB ratios a microscope and ocular micrometer. The tro were either taxon-specific values derived from phic status of each taxon was determined by the study streams or an assumed cohort PIB eKamining gut contents (Gaines et aI. 1989) or value of 5 (Waters 1977, Benke et at. 1984). by reference to Merritt and Cummins (1984). These taxa were not present in sufficient num Biomass was determined as dry weight (DW) bers to provide an accurate SF distribution far all size classes after drying at 60 C for 24 h curve that is necessary to compute SF produc and weighing to the nearest 0.1 mg. tion estimates. The Size-Frequency(SF) methad(Hynes and Coleman 1968, Hamilton 1969, Hynes 1980, RESULTS Waters and Hokenstrom 1980) was used to estimate secondary production of the most Production calculations for DC, SS, and RS common taxa. An average SF distribution was are given in Tables 3, 4, and 5, respectively. The determined from monthly sample sets; these follOWing text describes some of the assump represent~d the survivorship curve ofan "aver· tions we used in our calculations, data support age cohort" (Hamilton 1969, Benke and Waide ing these assumptions. and other infonnation 1977); "zero" values were included when calcu relevant to the production calculations. All pro lating densities. Production was estimated by duction estimates, unless noted otherwise. are calculating the loss between successive size given in units of mg DW.m-2.yr-l classes and then multiplying the loss by the Douglas Creek number ofsize classes using the equation given by Hamilton (1969). Production estimates were EPHEMEROPTERA.-Mayflies typically exhibit refined by multiplying by365/CPI (Cobart Pro widely varied larval development times (Clif duction Interval; Benke 1979). ford 1982). ClifTord (1982) examined life-cycle We found that conducting growth studies for data of &5 species of Heptageniidae and found all taJ90% had at least one univoltine cycle. not practicable. To establish reasonable esti Field data for Baelis sp. in DC provided little mates oflarvaldevelopment times and CPls,we clarification of the CPI. Based upon field data followed the example of Benke et at. (1984), ofBaetis sp. from RS and S5, and a growth study who used available life-history data and field in RS, we estimated a CPI of 60 d. Similar data to estimate CPls. We used three major temperature regimes in DC and RS support this 16 GREAT BASIN NATURALI,'T [Volume 52 TAllL/\ 3, Annual pr Annual Calculation R production Annual 2 2 2 365/CPI'l method N/m SE CV (mg D\iVlm ) SE CV (rng DW/m ) PIB Ephemcroptera Baetis sp. (gc, D)I' 6' SF!; 2416 0.41 92.4 263.7 0.41 91.9 8320 31.5 l)araleptophlehul sp. (gc, D) 1" SF 225 0.35 78.5 48.1 0.38 85.4 249 5.2 [.>£'11Cmcllta sp. (g, I-I) .1 +,0 SF 160 0.47 104.0 51.4 0.51 1040 238 4.6 l'ricorytfwdes sp. (gc, D) 9' PHd 6 0.80 159.2 1.7 0.67 151.0 77 45.Oe TerrAL 2807 364.9 884 Odonata Argia tihi.alis (p, C) 1· PB 30 0.46 103.9 8.9 0.49 UO.3 44 5.0'" Plecoptera [soperla sp. (p, C) ]" SF 77 0.58 129.4 42.8 0.58 lJ3.9 183 4.3 Trichoptera Hydropsyche sr. (fe, D) 1"'0 SF 445 0.57 127.1 413.5 0.65 145.8 1700 4.1 Cheum(/topsyche sp. (Ii:, D) 2+ •' SF 156 0.53 US.3 84.1 0.60 1350 818 9.7 Lellcotricftia pictipes (g, H) 1+," SF 95 0.63 139,7 7.7 0.68 15.32 32 4.2 TOTAL 696 505.3 2550 Coleoptera Optioserow; sp. (g, H) I+ SF 4322 0..37 83.5 606.7 0.36 80.0 2160 3.6 Diptera ChironoJnl1s sp. (ge, D) 15' PB 753 0.71 152.3 60.7 0.69 153.8 4920 8Ur Simulium sp. (fe, Dj 12" PB 41 0.75 168.6 .31.2 0.72 136.1 1680 54.0f Paramctliocnemus sp. (ge, D) W SF 196 0.44 98.0 10.4 0.46 101.9 875 84.1 Chaet'ocladius sp. (gc. D) 15' SF 115 0.57 127.8 3..5 0.66 129.4 426 121.7 Heleniella sp. (ge, D) 15' SF 141 0.52 U6A 4.5. 0.54 116.5 423 94.0 e Tipulidae (s, D) P PB 37 0.37 82..5 82.1 0.48 103.1 4U 5.0 Plwerwspectnl sp. (g, Hj 9' PB 60 0.07 15.5 4.9 0.07 15.0 221 45.0" Polypedilum sp. (s, H) 18' SF 33 0.69 154.5 2.2 0.78 129.1 161 73.1 Tahanidae (p, C) I" PB 51 0.48 106.6 27.8 0.48 107.5 130 .'5.0" Thienemannirnyia sp. (p, C) 15' PB 11 0.81 180..5 0.9 0.83 185.4 75 834 Brilliajlavifrons (s, D) 1.5' PB 12 0.25 55.0 0.9 0.26 .57.4 68 75.0" Empididae (p, C) 1.5" PB 1 0.22 50.0 0.1 0.18 40.0 8 75.0" TerrAL 1451 229.2 9358 CHAND TOTAL 9383 1757.8 23219 ;'S"IIrL~' of (;1']'''('.1, • ~ ,lerived fWll1gm"1h ,tndie." + = Held dat" and SF di TABLE 4. Annual produt.'tion ofinsects from Snively Springs, July 1985 to June 1986. Annuul Calculation B production Annual II 2 3651CPI method N/m SE CV (mgOW/m') SE CV (mgDW/m') PIB Ephellleroptera Baetis sp. (gc, D)h 6'- SFc 1388 0.62 104.7 18S.4 0.55 96.3 7010 37.8 Parakptr'Phlehia sp. (go, 0) I' SF 54 0.27 47.5 15.5 0.28 48.2 67 4.3 TOTAL 1442 200.9 7077 Odonata d Mgia tibial;" (p, C) I' PB 22 0.61 106.6 27.8 0.68 118.6 139 .5.01' Trichoptera Cheumatr'Psyclw sp. (fe, 0) 2'- SF 433 0.41 83.0 200.9 0.51 86.9 1300 6.5 Diptera , Simulium '1" (fe, OJ 12+,• SF 276 0.70 121.3 34.3 0.82 142.6 1880 54.8 ChirorunntLS sp. (gc, D) J50 SF 412 0.54 93.2 17.1 0.58 99.8 1390 8Ll TIpulidae (s, 0) I' PB 25 0.60 103.8 219.2 0.50 87,4 1100 5.Oe HelenieUa '1" (go, D) 15' SF 381 0,40 69.2 9.2 0.37 64.7 550 60.3 Palypedilum'l" (s, H) 18" SF J23 0.56 96.2 3.2 0.52 89.1 220 68.6 Chaetoc1adius '1'. (gc, D) 15' SF 92 0.83 108.3 2.7 0.69 120.2 210 77.8 Ooodae (gc, D) 15' PB 21 0.55 95.9 1.3 0.65 111.5 98 75.0" Thienem "SllUTCt': uf CI'I \l~d: ... delivoed from gnrwtll stu<.lie<; ... Ii.:kl d"l;>. 'H,d SF' dislr;h\llil>rl.~ 0 .. Iitt'I~\l lire: - .. h:~d lIpon el'l for si milaf (.';1",1 i ~~ucU (.t' DlPTERA.-Simulium sp. were not present in Empididae grew to a maximum size similar to sufficient numbers in DC to calculate an SF many ofthe midges; tberefore, a CPI of25 d was production estimate. The PIB ratio was calcu used. lated by averaging the PIB ratios obtained for Snively Splings Sil'lWlium sp. in SS and RS by the SF method. Accurate CPl estimates for Chironomidae EPHEMEROPTERA.~Gray(1981) reported a could not be obtained from field obselVations or larval development time of 20 d for BacUs SF distribution. Therefore, wedelivedCPlesti quilleri in Sycamore Creek, Arizona. Becauseof mates, as did Benke et aJ. (1984), and used lower stream temperatures, however. Baetis sp. growth data from Mackey (1977). Mackey developed more slowly in all streams in this (1977) reported lalVal development times of21 study. We assumed a CPl of60d. Paraleptophle d for Chironomus sp., 13 d for Polypedilum hia sp. was present only dUling the summer; convidum, and 36 d for Phaenospedra flavipes thus, we used only summer data to calculate at 15 C. CPls were compensated for slightly production because annual P was essentially lower average temperatures in DC (13 C) and equal to summer P. environmental stress (e.g., food availability, OOONATA.-Al-gia tibialis was not present in competition, etc.). These PIB ratios seem high sufficient numbe" to make an SF production but are comparable to other data where short estimate. CPls were used to estimate PIB ratios (Benkeet TRICHOPTERA.-Field data and SF dataindi at 1984, Jackson and Fisher 1986). Tabanidae cated a bivoltine life cycle and a CPI of6 mo for and Tipulidae were assumed to be univoltine Cheumatopsyche sp., tbe only caddisfly in SS. with a development time of 1 yr (Krueger and DIPTERA.-Becker (1973) reported a lalVal Cook 1984). This is consistent with the estimate development time of 13 d for S. vittatum grown ofa 1-yr development time for Taban~ dorsifer in tbe laboratory at 17 C. A 3O-d CPl was esti in Sycamore Creek, Arizona (Gray 1981). mated considering lower stream temperatures 18 GREAT BASIN NATURALIST [Volume 52 TABLE 5. Annual production ofinsects from Rattlesnake Springs, July 198,5 to June 1986. Annual Calculation B production Annual 2 2 ,165/CPI" method N/m SE CV (mg DW/m ) SE CV (mg DW/m2) P/B Ephemcroptera Betetis sp. (gc, D)h 6~>+'~ SFc 1336 0.61 107.2 47.3 0.58 104.0 2540 53.8 d l'riconJthodes sp. (gc, D) 9" PE I 0.05 8.3 0 ..3 0.07 12.2 14 45,0" TOTAL 1337 47.6 25M Odonata Argia tihialis (p, C) 1· PB 67 0.72 124.1 74.3 0.78 134.9 372 "".Oe Trichoptera Chelunatopsyche sp. (fe, D) 2 c ,+.o SF 140 0.69 118.9 48.6 0.78 134.5 486 10.0 Parapsyche sp. (fe, D) I- PE 10 0.24 41.7 26.8 0.25 43.4 134 c.- Oe Urnru'Phill1s sp. (s, D) I- PE 52 0.45 76.9 22.0 0.38 66.3 115 5.0" TOTAL 202 97.4 735 Coleoptera Hydaticus sr. (p, C) I- PE 4 0.50 87.4 1.2 035 60.1 6 c.- Oe e Hydropbilidae (p, C) r PE I 0.27 47.6 0.3 0.25 4.3l 2 5.0 TOTAL 5 8 Diptera Simulium sp. (fe, D) 12°"'-'0 SF 1777 0.73 125.8 212.3 0.73 127.5 11,180 52.6 Chironomlls sp. (gc, Dj IS" SF 192 0.50 87.3 7.0 0.58 loo.S 489 69.9 Heleniella sr. (gc, D) 15° SF 352 0.51 89.0 .5.4 0.51 88.4 480 88.9 Thienenumnimyia sr. (p, C) 15" SF 114 0.55 94.9 3.3 0.55 95.2 279 83.6 Tabanidae (p, C) ro PE 34 0.51 85.6 15.9 0.64 111.0 80 5.De Misc. Chironomidac (gc, D) 15" PB 18 0.29 50.1 0.8 0.38 66.3 60 75.0e f Polypedilum sp. (s, H) 18" PB 13 0.62 108.2 0.6 0.46 78.9 41 68.6 Chaetocladius sp. (ge, Dj 15° SF 59 0.73 126.4 0.4 0.56 97.7 30 75.0 Empididae (p, C) 15- PE 8 0.39 68.3 0.4 0.23 39.8 30 75.0" Tipulid.ae (s, D) ro PB 3 0.21 .35.9 2.0 0.26 44.3 10 5.0e Dixidae (gc, D) 15' PE 2 0.28 64.7 0.1 0.29 50.0 8 7S.0e TOTAL 2572 248.2 12.687 GHAI\'D TOTAL 4183 469.0 16,356 'S TABLR 6. Annual production (P, mg DW.m-2·yr_l) and percent production ofinsect functional groups in Doughts Creek, Snively Springs, and Rattlesnake Springs, July 1985 to June 1986. Douglas Creek Snively Springs Rattlesnake Springs Functional group P % P % P % Grazer/scmper 2651 11.4 0 0.0 0 0.0 Collector Gatherer 15,282 65.8 9332 65.9 3621 22.2 Filterer 4198 18.! 3177 22.5 1l,800 72.1 (Total) (19,480) (83.9) (12,509) (88.4) (15,421) (94.31 Shredder 639 2.8 1316 9.3 166 1.0 Predator 449 1.9 329 2.3 769 4.7 GRAND TOTAL 23,219 100.0 14,154 100.0 16,356 100.0 , 2 TABLI'; 7. Annualjroduction (P, mg DW·m- 'yr.l) and percent production of insect trophic levels in Douglas Creek, Snively Springs. an Rattlesnake Springs, July 1985 to June 1986. Douglas Creek Snively Springs Rattlesnake Springs Trophic level p % P % P % Herbivore 2812 121 220 1.6 41 0.3 Detritivore 19,967 86.0 13.605 96.1 15,546 95.0 Carnivore 440 1.9 329 2.3 769 4.7 TOTAL 23,219 100.0 14,154 100.0 16,356 100.0 Functional Group Production DISCUSSION Productionby oollectors was greatest ofall func Interstream Comparisons tional groups in all study streams. Collector pro 2 duction was highest in DC, 19.5 g.m- .yr'1, DC was clearly the most productive of the accounting for 83.9% ofthe total annual produc. tlu'ee streams studied (Table 6), and this is prob tion ofinsects. In SS and RS, oolleclor production ably related to the variety ofsubstratum (Table was 12.5 gand 15.4 g, representing88.4and94.3% 2) and resulting increase in microhabitat diver of the total annual production, respectively. The sity. Minshall (1984) thoroughly reviewed the annual production ofall functional groups in each importance ofsubstratum heterogeneity and its study stream is shown in Table 6. influence on insect abundance and distribution. SS and RS were similar in size and had similar Trophic Level Production total productivity estimates (Table 6), although Herbivores and detoitivores are both second important differences existed among the biotic ary producers at the same trophic level; carni components. vores are tertiary producers. For this discussion, In tenus offunctional group productivity, col we address them separately. Detritivore pro lectors dominated in each ofthe streams. Gath duction was greatest ofall trophic levels in each erers were more important in DC and SS, and study stream. Inpel.detritivore productionwas filterers in RS. The greater filterer/gatherer about 20.0 g-m ·yr , accounting for 86.0% of ratio in RS is probably related to the shifting the totalannual insect production. InSS and RS, nature of the sandy substratum (Table 2) and detritivore production was 13.6 g and 15.5 g, resulting absence ofareas for detritus to collect representing 96.1 and 95.0% ofthe total annual and be hmvested. The Hltering simuliids insect production. Herbivores contributed occurred on the abundant watercress plants. 12.1% of the productivity in DG, but no other The scarcity of solid substratum for periphyton trophic level in any of the three streams was an development in RS also explains the absence of important contributor to secondary production. grazers in this stream. However. substratum The annual production of all trophic levels in composition does notexplain a lack ofgmzers in each stream is given in Table 7. SS, where solid substmtum is present (Table 2). 20 CHEAT BASIN NATURALIST [Volume 52 2 In 55, the dense riparian canopy almost com TABLE8. Comparative annual production (mg DW'm- 'yr_ pletely shaded and obscured the stream. This 1) oftllxa common to Douglas Creek, Snively Springs, and Rattlesnake Springs, July 1985 to June 1986. probably prevented thc development of a sub stantial periphytic food base lor grazers. In DC, Douglas Snively Hattlesnake which had both solid substratum and unshaded Taxon Creek SpIings Springs stream bottom, a significant grazer community Ephemeroptera was present (Table 6). BaeUs sr. 8:m 7012 2542 Comparing the productivity of taxa common Odonata to all three streams shows some differences that Argia ti};iali~ 44 [39 372 are difllcult to explain (Table 8). For example, Trichoptera Simulium sp. production was similar in DC and Cheumatopsyche sp. 818 [298 486 55, butwas an orderofmagnitude greaterin RS. Diptera This may indicate a richer source ofsuspended Simulium sp. 1680 1879 1l,l75 [386 f()od in RS; however, comparative measure Chirofl,o'/nus sp. 4920 489 Polypedilum sp. 161 220 41 ments ofthis resource were not made. Cushing Tabanidae 1.30 53 80 and Wolf (1982) report a value of 1513 TipuJidae 411 1096 [0 Kca].m'2y..-[ of suspended POM in RS, but comparable data arc not available for DC and SS. This value is mueh less than that reported severely limit the potential productivity of RS. by Minshall (1978) for Deep Creek, a small, cold It is notable that the dominant secondary pro desert stream in southeastern Idaho. Since ducers in RS are the black flies, organisms that Si1111Jlium sp. production far exceeded that of are found in abundance soon after discharge any other insect in RS (Table 5), competitive diminishes (Cushing and Gaines 1989). exclusion (Hemphill and Cooper 1983) may make it more successful in competing for the Intrastream Comparisons limited attachment sites. Cheumatopsyche sp. DOUGLAS CHEEK~Secondary production in and Parapsyche sp., two filtering TIichoptera in DCwas spread over a wider variety offunctional HS, had a combined production of 620 mg as groups (Table 6) and trophic levels (Table 7), compared with Sinmliurn sp, production of >[[,000 mg. This is a 20-lold difference for even though it was dominated by cletritus-feed organisms ofthe same functional group. Except ing collector-gatherers. Chironomus sp. and fc>r Simulium sp., dipteran productionwas high BaeUs sp. wercthedominant secondmy produc est in DC for Chirorwmus sp. and Tabanidae, ers in the stream. while in SS, production ofPolypedilurn sp. and SNIVELY SPRlNGS.-ln SS, about 50% of tbe Tipulidae was highest. Tipulidae production secondary production was due to Bactis sp., a increased by an order of magnitude from RS to detritus-feeding collector-gatherer; and, as DC to SS. This may be related to the relatively mentioned above, the grazing component was high amounts of particulate organic mattcr absent. Total dipteran production was of the (POM) found in the study section of SS (Cush same order of magnitude as that for Baetis sp. ing 1988). Production of Baetis sp. is three to hut was spread out among several organisms, four times lower in RS than in the other two notably Simulium sp., Chirorwmus sp., and streams (Table 8). Tipulidae (Table 4). A likely explanation for some of the differ RATTLESNAKE SPIUNGs.-Secondary pro, enccs shown in Table 8 is the winter spates that duction in RS was less diverse than in the other occur in RS, but not in SS or DC. These spates, study streams, \.vith over 68% ofthe production described by Cushing and Gaines (1989), scour due to the filtering dctritivore Simulium sp. The the entire streambed: flushing out accumulated second highest producer was BacUs sp., but POM and much ofthe fauna. They occur about production was far lower than the black flies every three years and act as a «reset" mecha (Tahle 5). The high production of simuliids in nism. Because they occur in winter when there RS can be attributed to the presence ofmultiple are no ovipositing adults, and because they cohorts with short development times. Gray scour and eliminate sources for both upstream (1981) suggested tbat rapid development may migration and downstream dlift, they must be advantageous in streams subject to spates. 1992] INSECT PRODUCTIVITY IN SPRING-STREAMS 21 TABU; 9. Comparative whole stream secondmy proouction of insects (P, g DW'm-2'yr-l), except as indicated, in five geoclimatic regions. Streams grouped by geographical region, not by temperature regimes. Stream P 5" Fc Gc Grise Fred Source Cold/mesic Unnamed, Quebec s.d' Harper 1978 FactoI)' Br., Maine 12.2 Neves 1979 Sand R.• Alberta O.8c Soluk 1985 Caribou R, Minnesota 3.54 0.83 0.62 1.36 0.14 0.59 Krueger and Waters 1983 Blackhoof R" Minnesota 7.13 1.()() 3.53 1.15 0.37 1.08 Krueger and Waters 1983 No. Branch Cr., Minnesota 13.23 0.73 5.33 9.43 1.()() 2.07 Kmeger and Waters 1983 Fort R., Massachusetts 3.3 Fisher 1977 Bear Br., Massachusetts 4.8 Fisher and Likens 1973 L'Ance du Nord, France 12.5 (Total detritiVO'B P = P- Pred.) 2.0 Maslin and Pattee 1981 Bisballe baek, Denmark 26.7 1.3 Mortensen and Simonsen 1983 HumidJmesic Satilla R. Georgiad 25.2 2.9 18.0 4.3 Benke et at 1984 Snag substratee 84.S 49.3 8.1 75 Sandy substrateI"' 21.0 0 17.9 3.1 Mud substratee 17.9 0.2 8.6 9.2 Cedar R., So. Carolina 3.0 0.1 1.0 1.3 0.02 0.6 Smock et a1. 1985 Lower Shope Fk., No. Carolina 1.4 Georgian and Wallace 1983 Upper Ball Cr., No. Carolina Huryn and Wallace 1987 Bedrock-outcrop 6.1 0.6 2.1 2.1 0.6 0.7 Riffle 5.6 1.4 0.3 1.8 1.0 1.1 Pool 7.8 2.4 0.03 3.0 0.3 1.9 Hot desert Sycamore Cr., Alizona 120.9 Jackson and Fisher 1986 New Zealand Hinau R. 38.2 Hopkins 1976 Horokiwi R. 41.5 Hopkins 1976 Cold desert Deep Cr., Stu. 1, Idaho 1.2 Minshall et al. 1973 Douglas Cr., Washington 23.2 0.6 4.2 15.3 2.7 0.4 This study Snively Spr., Washington 14.2 1.3 3.2 9.3 0 0.3 This study Rattlesnake Spr., Washington 16.4 0.2 3.6 IlS 0 0.8 This study 's • shrl"'tklef; Fo: .. Aitering-o:olledlll"; Go:' .. !:atheling-o:ollettor; Grlso: = grazer/ocmper; l'red ~ predators. ~Ernerger" only_ '"Only two spec.1es ofo:hi l\Jnomids. "Expre.\'sed per l1nit ,\rea oftotal stream bottOio. ('E~pres~ed per uni tare;' ofl",bitnt. Comparisons with Other Streams Total insect production rates in this study 2 ranged from 14 to 23 g DW.m- -yr-1 and are Annual PIB ratios ranged from 3.6 to 121.7 for insects from the studystreams. The high annual compared with values for other streams PIB ratios are attributed to insects with rapid grouped by geographical region (Table 9). Pro development and multiple cohorts (e.g., many duction rates in cold desert streams are well Chironomidae). The annual PIB ratios found in below the higher values found in New Zealand these cold desert spring-streams are generally streams, the richer areas (snags) ofhumidlmesic lowerthan those reportedbyJackson and Fisher streams in the southeastern United States, and (1986) for Sonoran Desert stream insects and by Sonoran hot desert streams. However, produc Benke et a1. (1984) for southeastern blackwater tion rates in cold desert streams are higher than stream insects. The Sonoran and blackwater those in streams in cold/mesic areas of the streams are warmer and insect development is United States. These rankings relate to the faster, resulting in a greater number ofcohorts. interaction among stream water temperature, Our annual PIB ratios were generally higher insect development, cohort production inter than reported for northern temperate streams vals, and other factors. However, it should be (Krueger and Waters 1983), where cooler kept in mind that other factors, e.g., geochem streams result in insect development at slower istry, maybeinfluential in governing production rates with fewer cohorts. as well as temperature. Production values in 22 GHEAT BASIN NATURALIST [Volume 52 Rattlesnake Springs, which has a sandy substra to stream communities in terms of production. tum, are comparable to the sandy areas of the In many streams, they contribute a large per Satilla River in Georgia (16.4 vs 13.1 g centage of the total community production DW.m-2-yr-\ respectively); production of eol because of their rapid development and high ICdor-gathcrers was identical. tumover rates. We found high PIB ratios for Benke et a!. (1984) stated that measurement simuliids and cruronomids, but other investiga of secondary productivity ofbenthic organisms tors have reported similar results (Fisher and provides a truer indication of their import,mee Gray 1983, Benke et al. 1984, Stites and Benke in lotic ecosystems than does measurement of 1989). This life-history strategy is particularly either density or biomass. This is intuitively advantageous for insects inhabiting the streams reasonable since measurement of P, a rate, that are subjected to severe spates. includes consideration ofboth biomass andden Detritus is the major food resource in these sity. Our results support the validity ofBenke et small streams; collector-gatherers predominate al.'s (1984) contention. Clearly, our data reveal where there is more substratum diversity (DC that collectors arc the dominant functional and SS), and fHterers in systems more prone to group, and detritivores the domin;mt trophic the effects of spates (RS). Grazer/scrapers are level in terms of the secondary productivity of present whenever suitable substratum and suf insect.s in these three streams (Tables 6 and 7). Hcient sunlight are available for development of If only biomass or density data are evaluated a periphyton crop. Shredders, surprisingly, arc from these streams (Tables 3, 4, and 5; Gaines not well represented in these small headwater et a!. 1989), anomalies become evident. Density streams. This may be related to the flushing of data in DC reveal that herbivores are equally as the systems hy the spates and/or the low nnmerous as detritivores, but biomass data amounts ofallochthonous detritus reaching the reveal that detritivores are about wo times streams (Cushing 1988). Secondary productiv greater than herbivores. Conversely, when the ity of these cold desert spring-streams was less insect.s are separated into functional groups, the than that ofstreams in hot deserts, but generally biomass ofgrazer/scrapers (herbivores) exceeds higher than that in most cold/mesic ,md that of collectors in DC by a factor of wo. humid/mesic streams. Finally, our results Further, collector-Hlterers in DC represent underscore the contentions ofBenke et al. (1984) 18% ofthe production and 30% ofthe hiomass, that measuring the secondary production of but only 7% of the density. In SS, trophic level insects in streams provides a better assessment of compmisons reveal that detritivores dominate their role than density or biomass, but the anom production, biomass, and density, but if func alies described above argue for care in applying tional groups are compared, biomass data would this generalization to all streams. overemphasize the importance of shredders (,30%), whieh form only 5% ofthe density and ACKNOWLEDGMENTS 9% oftotal production. In RS, the largest anom aly appears when comparing functional groups. This paper represents a portion of the thesis Although collector-filterers represent 72% of suhmitted by WLG to CentralWashington Uni the total production and 61 % of the biomass, versity for the M.S. degree. The researcb was their density is similarto the collector-gatherers. performed at Pacific Northwest Laboratory In c'oncll1sion, wehave found thattaxawith short dUling a Northwest College and University development times and multiple cohOltS, such as Association for Science (NORCUS) Fellowship midges and black flies, are important to cold (University of Washington) to WLG. It was desert spring-stream production. Previous studies funded under Contract DE-AM06-76 have addressed the difflculties in obtaining accu RL02225 and was supported by the U.S. rate Held estimates of Simuliidae (black fly) and Department of Energy (DOE) under Contract Cruronomidae (midge) laIvae CPIs, and thus pro DE-AC06-76RLO 1830 between DOE and duction estimates (Benke et al. 1984, Behmerand Battelle Memorial Institute. Hawkins 1986, Stites and Benke 1989). Their We would like to thank Dr. William Coffman small size, rapid turnover rate, high density, and for identifying the chironomids, and Dr. Pat diversity make accurate species-specific CPI esti Schefter for identifYing the caddisflies. The mates difficult. These same characteristics, how manusclipt was improved by comments from evel~ make midges and black flies very important threeanonymous reviewers; ourthanks to them. 1992] INSEC' PHODUGfIVITY IN SPHING-STHEAMS 23 LITEHATURE CITED GEOHGIAN, 1'., cmd J. B. WALLACE. 1983. Seasonal produc tion dynamics in a guild of petiphyton-grazing insects ANDEHSON, R. O. 1959, A modified notation technique for in a southern Appalachi _~;;-' 1984. Aquatic insec.:t-substratum relationships. SOLUK, D. A. 1985. Macroinvertebrate abundance and pro Pages .'358--400 in V II. Re-sh and D. M. Rosenberg, duction of psammophilous Chironomidae in shifting cds.) The cc·ology of aquatic inseds. Pmeger Publish sand areas of a lowland river. Canadian Journal of ers, New York. Fisheries and Aquatic Sciences 42: 1296-1302. MINSIlALL, G. W., D. A. ANDl\EWS. F. L Ros~~. D. W. SUAW. STITES, D. L., and A. C. BENKE. 1989. Rapid growth rates and H. L. NEWELL. 1973. Validation studies at Deep ofchironomids in three habitats ofa subtropical black Creek, CurlewValley. Idaho State University Research water river and their implications for P:B ratios. Lim Memorandum No. 73-48. nology and Oceanography 34: 1278-1289. MOll.TJ<;NSEN, E., and J. L. SIMONSEN. 1983. Production WATE BS, T. E 1977. Secondary production in inland waters. estimates of the benthic invertebrate community in a Advances in Ecological Research 10: 91-164. small Danish stream. Hydrobiologia 102: 155-162. WATEHS, T. F., and J. C. HOKENSTHOM, 1980. Annual pro NEVES. H. J. 1979. Secondary proo.ucti,on ofepilithic fauna duction and drift of the stream amphipod Gammams in a woodland stream. American Midland Naturalist pseudolimnaeus in Valley Creek, Minnesota. Limnol 102: 209-224. ogy and Oceanography 25: 700-710. O'IIol'. J., J. B. WALLACE, and J. D. HAEFNP;n, 1984. Production of a stream shredder, Peltoperla nutria Received 1June 1991 (Plecoptera; Peltoperlidae) in disturbed and undis Revised 1 D&Jemher 1991 turbed hardwood catchments. Freshwater Biology 14: Accepted 10January 1992 13-21. SMOCK, L. A., K GILlNSKY, andD. L. STONg]>UJ\NRH.1985. Macroinvcltebrate production in a southeastern United States blackwater stream. Ecology 66: 1491 1.50.,.