Oviposition of the Dobsonfly (Corydalus Cornutus, Megaloptera) on a Large River Author(S): Brian P

Am.Midl. Nat. 127:348-354

Oviposition of the Dobsonfly(Corydalus cornutus, Megaloptera) on a Large River

BRIAN P. MANGAN EcologyIII, Inc., R.R. 1, Berwick,Pennsylvania 18603

ABSTRACT.-Dobsonflyoviposition and theresults of twoannual surveys of oviposition sitesalong the Susquehanna River, near Berwick, Pennsylvania, in 1985-1986are described. Unlikeother megalopterans, none of the females prepared the substrate before egg deposition. Afterdeposition, females spread a clearfluid over the eggs,which dried to forma hard whitecoating. This coating appeared to protect the eggs from excessive heating and predation. Some4428 eggmasses were observed at 260 sitesin thetwo annual surveys. Trees were themost often used sites, but rocks and deadfallshad higheraverage numbers of egg masses on them.Aggregations of egg masseswere observed in surveyedareas and on individual sites(e.g., 339 on one tree),and Morisita'sDispersion Index showed that egg masses were contagiousin distribution.

INTRODUCTION The dobsonflyCorydalus cornutus (L.) is a widely distributedaquatic insectand an im- portantpredator in many lotic ecosystems.Larvae (hellgrammites)are prized as bait for game fishes.Although the insect is well known among anglers,it has receivedrelatively littlescientific attention. Descriptions of the life history,and particularlyoviposition, are sparse and incomplete.Riley (1879) was firstto characterizeoviposition, noting egg masses on objectsoverhanging water. Parfin (1952) studiedpupation, adult emergenceand copu- lationof the insect,and Baker and Neunzig (1968) describedthe egg masses and first-instar larvae. Brown and Fitzpatrick(1978) reportedlife history and populationenergetics of this species. This study investigateddobsonfly oviposition on the Susquehanna River, near Berwick,Pennsylvania, in 1985-1986. My objectiveswere to describedobsonfly oviposition and surveyoviposition sites along a large river.

STUDY AREA The studywas conductedalong a 6.4-kmsection of the Susquehanna River,approximately 11 km N of Berwick,Luzerne County,Pennsylvania. In the past,water quality and benthic macroinvertebrateswere depressedby coal mine drainage which enteredupriver from the studyarea (Deutsch, 1981). However, water quality has improvedsince 1976 (Soya et al., 1984) and benthicbiomass has increasedas much as fivefold(Zelenak et al., 1985). Mean riverwidth in this area was ca. 300 m and ranged from170-620 m. The river'sslope was about 0.3 m/km,and consequently,few natural rifflesoccurred in this area. Overhanging trees characterizedthe shoreline.Silver maple (Acer saccharinum)and riverbirch (Betula nigra) predominated;other species presentincluded sycamore(Platanus oxidentalis),red maple (Acer rubrum),weeping willow (Salix babylonica)and white oak (Quercus alba). Deadfalls and rocks (both on and offshore) were located intermittentlythrough the area.

METHODS AND MATERIALS During eveningsof June, July and August 1985, between 1800-2400 h, searcheswere made froma motorboatalong the rivershoreline for ovipositing females. A high intensity light was used for illumination.When ovipositingfemales were found,information con- cerningsubstrate preparation, egg depositionand postovipositionactivity was recorded. 348 1992 MANGAN:DOBSONFLY OVIPISITION 349

In Augustand September1985 and Septemberthrough November 1986, six 1.6-kmareas of rivershoreline were surveyedfor oviposition sites. These areas were randomlyselected fromtwelve 1.6-kmareas available fromthis sectionof river.The surveyswere conducted duringdaylight hours and the same six areas were surveyedboth years. Sites were observed with fieldglasses froma motorboator fromshore. An ovipositionsite was definedas any objectwith an egg mass laid on it. Substratetype, numberof egg masses presentand theirplacement relative to the water were recorded.In 1986, heightsof individual egg masses above the water were measured using a 7.6-m surveyingrod (standardizedto the riverlevel measured on 30 September 1986). Also in 1986, an attemptwas made to quantify"potential" ovipositionsites in each surveyarea. These potentialsites were the numbersof trees,rocks and deadfallsin the surveyareas and areas withoutegg masses thatsatisfied the criterionof locationnear the water. The purpose of this was to investigatea possible relationshipbetween observednumbers of oviposition sites and relativesite availability. Egg mass dispersionpatterns were testedstatistically for departure from randomness by chi-square analysis,and were furtherevaluated using Morisita's Index of Dispersion:

Id= n(2 x2- N)/N(N - 1) wheren is the numberof plots (actual and potentialoviposition sites), N is the totalnumber of individualscounted on all n plots,and sigma x2 is the square ofthe numberof individuals per plot summedover all plots (Id = 1.0 if dispersionis random,Id = 0 if perfectlyuniform, Id = n ifmaximally aggregated; Brower and Zar, 1984). The numbersof potential oviposition sites countedin 1986 were also used with 1985 surveydata so that the dispersionindex could be applied to thatyear. Change in potentialsites between these years was considered minimal.

RESULTS AND DISCUSSION Ovipositionbehavior.-Oviposition began in June and lasted throughSeptember in 1985 and 1986. The egg masses were composedof threelayers of eggs similarto the description given by Baker and Neunzig (1968). Average numberof ova per mass was 920 (range = 561-1491; SD = 268; n = 12), comparedto reportsof 3000 per mass in Missouri (Riley, 1879) and averages of 1080 in North Carolina (Baker and Neunzig, 1968) and 1309 in Texas (Brown and Fitzpatrick,1978). The average dimensionsof 68 egg masses laid on horizontaland diagonal surfacesof a boat dock in the riverwas 18.4 mm long and 15 mm wide, very similar to those reportedfrom Texas (Brown and Fitzpatrick,1978). The predominantshape was oval and all masses were white. Ovipositionbehavior of eightfemales was observedin the field.All ovipositionsoccurred aftersundown. None of the femalesprepared the substratebefore egg deposition.As individual eggs exited the abdomen,they were coated with a clear fluidwhich had adhesive properties.This adhesion assistedegg attachmentto the substrateand to othereggs in the mass. In contrastto Corydalus,another megalopteran, an alderfly,Sialis cornuta,was reported by Pritchardand Leischner (1973) to secretea stickysubstance from its abdomen onto the substratebefore egg deposition. The threelayers of eggs were depositedconcurrently; the bottomlayer containedthe most eggs, and the second and thirdlayers subsequentlyfewer. Eggs in the bottomlayer were laid parallel to the substrate,while those in the upper layers were angled away fromthe substrate(some almost perpendicular).Females stopped oftenduring depositionto probe the incompletemasses with the tips of theirabdomens, suggesting that a tactilestimulation 350 THE AMERICAN MIDLAND NATURALIST 127(2) is involvedin egg placement.The head and thorax remained motionlessas the abdomen was moved laterally to deposit eggs in rows. Egg depositionrequired about an hour to complete. When egg depositionwas completed,a clear fluidwas secretedfrom the abdomen and spreadover the eggs. The femalesbrushed the fluidover the eggswith the tip ofthe abdomen in a lateral motion,often extending the fluidbeyond the eggs and onto the substrate.This fluiddried to forma hard whitecoating on the egg mass. This is probablythe same material that coated individualeggs duringdeposition. I moved two femalesfrom their egg masses beforeapplication of the egg coating. Seeminglydisoriented, they moved around on the substratefor ca. 10 min,then applied the egg coatingdirectly to the substrate.Subsequently, the eggs of both uncoveredegg masses turnedbrown when exposed to directsunlight the followingday and the eggs did not hatch. Other egg masses laid on this same substrateand near these masses hatched successfully.This suggestedthe abilityof the white coatingto protectthe eggs fromexcessive heating, as proposed by Brown and Fitzpatrick(1978). Other observationssuggested that the coatingprotected the eggs frompredators as well. I foundonly fouregg masses that appeared damaged by predators,as indicatedby a hole in the protectivecoat. Eggs in the area of the hole were partiallyor wholly removed.The remainingundamaged eggs of these masses hatchedsuccessfully. The hard surfaceformed by this coatingis probablyimpenetrable to most potentialinsect predators. A Californian dobsonfly,Neohermes aridus (Maddux), lays eggs withoutprotective coats (Maddux, 1952), as do Sialis alderflies,and both were reportedto have large percentagesof theiregg masses parasitized(Azam and Anderson,1969; Pritchardand Leischner,1973). Brown and Fitz- patrick (1978) reportedthat only three Corydalusegg masses were attacked,of several thousandobserved in theirstudy. The females I observeddeposited only one egg mass beforemoving off the substrate. Brown and Fitzpatrick(1978) reportedthat Corydalus in Texas usually laid two egg masses each and sometimesa thirdsmaller one, and calculated an average of 2.3 egg masses per femalefrom numbers of ova per femaleprior to ovipositing.I collected10 femalesby light trap and counteda mean of 1293 ova, with a range of 633 to 1635 ova per female (SD = 298). I do notknow whether these females oviposited before collection. However, considering the range of ova per egg mass noted earlier,these femalescould lay one large egg mass or multiplesmaller ones. Ovipositionsites.-River surveysrevealed 3204 egg masses laid on 149 sites in 1985 and 1224 masses on 111 sites in 1986 (Table 1). Trees, rocks,deadfalls and a boat were used as sitesin thesurveyed areas. Otherobjects oviposited on outsidethe surveyed areas included: bridges,boat motors,an aluminum dock, cementand brickwalls and window glass. Gen- erally,the surveyedareas whichcontained many sites in 1985, also had a greaterproportion of sitesin 1986. However, the average numberof egg masses per site in the 1st year (21.5) was almosttwice that of the second (11). Certainsites identified in 1985 were again identified as sites the followingyear. These sites were oftenlarge rocks near shore and some were trees. Trees were the most frequentoviposition sites in the surveyedareas forboth years. Egg masses were laid on the trunks,limbs and leaves of trees,which usually occurredwithin 10 m of the rivershoreline (as measuredto the trunks).Egg masses that were observedon leaves and tree branches (less than one-thirdthe diameterof the dobsonfly'sabdomen), indicatedthe high degree of plasticityby femalesin exploitingoviposition sites. In 1985, about 75% of the siteswere trees,on which 73% of the egg masses were laid; in 1986, 93% were trees that held 80% of the egg masses. However, in areas where rock and deadfall sites occurred,the mean numbersof masses on these sites oftenexceeded the mean of tree 1992 MANGAN:DOBSONFLY OVIPISITION 351

TABLE 1.-Oviposition sites and egg masses of Corydaluscornutus

Type Number of sites Total egg masses Mean numberof egg masses (SD) of site 1985 1986 1985 1986 1985 1986 Tree 111 103 2327 982 21.0 (?44.1) 9.5 (?15.7) Rock 19 4 467 114 24.6 (?43.6) 28.5 (?21.8) Deadfall 18 4 409 128 22.7 (?39.2) 32.0 (?47.1) Boat 1 0 1 0 1.0 0 Total 149 i11 3204 1224

sites (Table 1). These rocks and deadfalls were probablypreferred habitat forovipositing due theirproximity to the water. There was a significantcorrelation (P < 0.001) betweenthe numbers of actual oviposition sitesand potentialsites in each of the surveyedareas (r = 0.97). Those areas thathad more potentialsites, generally had moreactual sites.However, the implicationof this relationship is uncertain,since those areas which had fewersites did not have greateraverages of egg masses per site, which should indicate limitedoviposition space. Conceivably,there were moreactual sitesin areas withmore potential sites, simply because ofgreater site availability. Most ovipositionsites and egg masses in bothyears were positionedover water described as pools (118 sites in 1985, 90 sites in 1986). Others were laid over riffles.In 1985, 313

350

300-

LLJ 250-

oD 200- 0 LiU

150 - o 0? LiJ

D100

50-

0 0-1.0 1.1-2.0 2.1-3.0 3.1-4.0 4.1-5.0 5.1-6.0 6.1-7.0 7.1-7.6 >7.6

HEIGHT (m) FIG. 1.-Height of egg masses above riverlevel in 1986 (all surveyareas combined) 352 THE AMERICAN MIDLAND NATURALIST 127(2)

?A CI CD cso o ~~~~~~~00o :S :e~~~~~~~~~~~~~~~~~I

6n N t~~c

O~~~~~LoO O N m .=H

Q~~~~~~~rO0 0 00

t %;)CA -Zml Q

=~~~~ c" .', o cs ;s) ttd L) n t,-

Q buN tt b e o~~~~~~~b = N~~~0 CC) o^

l~~~~~~~~0 00 bb Ln

bo b no 0no U > CZO OO O 1992 MANGAN:DOBSONFLY OVIPISITION 353

TABLE3.-Morisita's DispersionIndex from dobsonfly oviposition data

Index value Surveyarea Number of sites 1985 1986 1 38 17.3 19.2 2 28 9.0 6.8 3 139 7.4 6.8 4 77 24.0 10.6 5 66 8.4 7.3 6 50 7.0 15.4 egg masses were ovipositedover land, as were 86 masses the followingyear. The majority of these were due to early season ovipositionson sites near the edge of the water, which consequentlywere foundover land later in the season as riverlevel decreasedthroughout the summer.Four egg masses were observedon buildingsfrom 15-45 m fromwater. These were likelythe resultof phototaxisby femalesto nearbylight sources. Egg masses were ovipositedin a wide range of heightsabove the riverin 1986 (Fig. 1). Some masses on rock siteswere only a few centimetersabove riverlevel; otherswere in the tops of the highestshoreline trees (>40 m). About 30% of the egg masses were laid 0-2.0 m above riverlevel, indicatingthe use of lower sites such as rocksand deadfalls.Some egg masses on lowersites were destroyedwhen theriver level increased.Egg-mass height seemed most importantduring larval eclosion,since larvae essentiallyfree-fall from the masses to the water below. Larvae fromegg masses close to the water had less distanceto traveland a smallerprobability of missingthe water,whereas highermasses (the 26% laid above 7.6 m) run the greaterrisk of missingthe water completely. In bothyears, the majorityof egg masses were ovipositedat relativelyfew sites.In 1985, 60% of the masses were laid at 11% of the sites,and in 1986, 59% were laid at 14% of the sites (Table 2). In bothyears, 51-59% of the sites (in the six surveyedareas combined)had 1-5 egg masses laid on them,but these sites containedonly 5-13% of the total numberof egg masses in each year. I observed11% (339) of the total numberof egg masses surveyed in 1985, at one tree site. The followingyear 10% of the masses (119) were laid at the same site. In 1985, 16 sites held more than 50 egg masses per site, and fourof these had more than 150 masses ovipositedper site. Chi-square analyses of egg mass distributionrevealed thatthe patternswere highlysignificant (P < 0.005) in departurefrom randomness in each area for both years. Morisita's Dispersion Index reinforcedthis conclusion,indicating contagiousdistributions in all six areas forboth years,but also that masses were one-half or less theirmaximum aggregation values (Table 3). The distributionsof the egg masses on the siteswere also contagious.On many sitesthe masseswere grouped, and oftenthese eggs were only separated by the width of their protective coats.Multiple egg masseson singleleaves were observedrepeatedly. A red oak leaf collected in 1986 had 13 egg masses on one side and fourmore on the other.Riley (1879) observed a maple leaf with 20 egg masses on it. However, despite such crowding,egg masses were not laid on top of each other.This contagiousnesssuggests the possibilityof an oviposition pheromone.Such chemical communicationwas demonstratedwith pheromonesassociated with mosquito eggs and the egg-podsof locusts (Osgood, 1971; Norris, 1970). This com- municationcould increase exploitationof ovipositionspace, althoughspace did not appear to be a limitingfactor in thisstudy area. Dobsonflyoviposition along habitatssuch as smaller streamsmight better highlight such an adaptive feature.Brown and Fitzpatrick(1978) 354 THE AMERICAN MIDLAND NATURALIST 127(2) reportedthat 95% of dobsonflyegg masses were on several logs in theirstudy area, which was a rivermuch smallerthan the Susquehanna. This studyhas revealed some of the plasticityof Corydalusoviposition behavior, while raising questions for furtherresearch. How do the majorityof female dobsonflieslocate ovipositionsites over water? Is therean adaptive significanceto the reportedclustering of egg masses? What are the thermaland predatorprotection ranges offeredby the protective coat of the egg mass? Does ovipositionbehavior as reportedhere change along small stream habitats?Perhaps futurestudy will reveal more of the biologyof these interestinginsects. Acknowledgments.-I thank Professor Robert Sagar and Dr. FrederickHill fordirection in this study.I am gratefulto TheodoreV. Jacobsen,Dr. WilliamG. Deutschand an anonymousreviewer fortheir comments on themanuscript. This studywas supportedby the PennsylvaniaPower and LightCompany.

LITERATURE CITED AZAM,K. M. ANDN. H. ANDERSON. 1969. Life historyand habitsof Sialis rotundaand S. californica in westernOregon. Ann. Entomol.Soc. Am., 62:549-558. BAKER,J. R. ANDH. H. NEUNZIG. 1968. The egg masses,eggs, and first-instarlarvae of the eastern North American Corydalidae. Ann. Entomol.Soc. Am., 61:1181-1187. BROWER,J. E. AND J. H. ZAR. 1984. Field and laboratorymethods for general ecology,2nd ed. Wm. C..Brown Publishers,Dubuque, Iowa. 226 p. BROWN,A. V. ANDL. C. FITZPATRICK.1978. Life historyand populationenergetics of the dobsonfly, Corydaluscornutus. Ecology, 59:1091-1108. DEUTSCH,W. G. 1981. Suppressionof macrobenthosin an iron-pollutedstretch of the Susquehanna River (Pennsylvania). Proc. Pa. Acad. Sci., 55:37-42. MADDUX, D. E. 1952. A study of the dobsonflies(Megaloptera) of the Chico, California area. Unpubl. Masters Thesis, Chico State College, Chico, California. 45 p. NORRIS,M. J. 1970. Aggregationresponse in ovipositingfemales of the desertlocust, with special referenceto the chemical factor.J. InsectPhysiol., 16:1493-1515. OSGOOD,C. E. 1971. An ovipositionpheromone associated with the egg raftsof Culex tarsalis.J. Econ. Entomol.,64:1038-1041. PARFIN,S. 1952. Notes on the bionomics of Corydaluscornutus (Linne), Chauliodesrastricornis Rambur, C. pectinicornis(Linne) and Neohermessp. Am. Midl. Nat., 47:426-434. PRITCHARD,G. AND T. G. LEISCHNER. 1973. The life historyand feedinghabits of Sialis cornuta Ross in a series of abandoned beaver ponds (Insecta: Megaloptera). Can. J. Zool., 51:121- 131. RILEY,C. V. 1879. On the larval characteristicsof Corydalusand Chauliodesand on the development of Corydaluscornutus. Can. Entomol.,11:96-98. SOYA,W. J., B. P. MANGANAND T. V. JACOBSEN.1984. Physicochemicalanalysis, p. 7-57. In: T. V. Jacobsen (ed.). Ecological studies of the Susquehanna River in the vicinityof the Sus- quehanna Steam Electric Station (1983 annual report). IchthyologicalAssociates, Inc., Ber- wick, Pennsylvania. ZELENAK,L. S., W. G. DEUTSCH ANDW. F. GALE. 1985. Benthicmacroinvertebrates, p. 77-111. In: T. V. Jacobsen (ed.). Ecological studies of the Susquehanna River in the vicinityof the Susquehanna Steam Electric Station (1984 annual report). IchthyologicalAssociates, Inc., Berwick,Pennsylvania.

SUBMITTED 8 APRIL 1991 ACCEPTED 23 AUGUST 1991 The University of Notre Dame

Oviposition of the Dobsonfly (Corydalus cornutus, Megaloptera) on a Large River Author(s): Brian P. Mangan Source: American Midland Naturalist, Vol. 127, No. 2 (Apr., 1992), pp. 348-354 Published by: The University of Notre Dame Stable URL: http://www.jstor.org/stable/2426541 . Accessed: 08/05/2011 11:16

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=notredame. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

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 [email protected].

The University of Notre Dame is collaborating with JSTOR to digitize, preserve and extend access to American Midland Naturalist.

http://www.jstor.org