Great Basin Naturalist
Volume 53 Number 3 Article 8
9-27-1993
Prey selection and food habits of Burrowing Owls in Colorado
David L. Plumpton Texas Tech University, Lubbock, Texas
R. Scott Lutz Texas Tech University, Lubbock, Texas
Follow this and additional works at: https://scholarsarchive.byu.edu/gbn
Recommended Citation Plumpton, David L. and Lutz, R. Scott (1993) "Prey selection and food habits of Burrowing Owls in Colorado," Great Basin Naturalist: Vol. 53 : No. 3 , Article 8. Available at: https://scholarsarchive.byu.edu/gbn/vol53/iss3/8
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]. Creal Basin Naturalisl 53(3}, pp. 299-304
PREY SELECTION AND FOOD HABITS OF BURROWING OWLS IN COLORADO
David L. P)umpton l,2 and R. Scott Lutz1
AIJSTltAGr.-Food habits of Hurrowing Owls (Spcotyto cunic.;ularia) were studied during the hreeding seasons of 1990 llnd 1991 ill cf'..ntral Colorado. Concurrent insect availability studies were conducted to dett:l"mine selection for specific insect families. Analysis of 1445 castings imlic-.l.ted lise ofonly one insect family, the carrion beetles (Silphiclae), at a rote greater than expected based. on availability in one year. Castings and pre)' remains showed different dietary components. Pre)' remains showed greater use ofsmall mammals, moths, amphibians, and passcrincs, and castings indi cated greater use of mice and beetles. Mcthodolq,'Y in raptar fOOd habits stndil:S may therefore hias results.
Key words: Speotyto cunicularia, Burrowing OwlJood habits, casting. prey 1"emains, Colorado.
Much of the diet literature for Burrowing Northern Great Plains province in the North Owls (Speotyw cunicularia) consists solely of Temperate Grassland Biome, RMA has a studies based on casting analysis. Study of semiarid climate and features low humidity, Burrowing Owl food habits hased on prey light rainfall, and moderate to high winds. remains found near burrows has shown prey Average annual precipitation is about 38 em. items not detected from casting analysis Elevation ranges from 1564 m to 1625 m (Thomsen 1971, MacCraeken et al. 1985). above sea level (Environmental Science and Most food habits studies lack concurrent Engioeering 1989). quantitative studies of prey populations. This Vegetation is represented by five major deficiency is also apparent in raptor food communities: weedy forbs, cheatgrass (B1'o habits studies (Brown 1974, Olendorff and mus spp.)/weedy forbs, cheatgrass/perennial Stoddart 1974). grassland, native perennial grassland, and in 1987 the Rocky Mountain Arsenal crested whcatgrass (Agropyron cristatum). (RMA) was established as an Environmental Minor communities include sand sagebrush Protection Agency superfund site, and (Artemisia filifolill) shmbland, mbber rabbit cleanup operations were initiated. Cleanup of brush (Chrysothamnus nauseosus) shrl1bland, RMA may require alterations to existing habi yucca (Yucca spp.) grassland, cottonwood tats, possibly affecting the Burrowing Owl (Populus deltoides), and w;l]ow (Salix spp.) prey ba'e. The primary objective of this study (Environmental Science and Engineering was to determine the food habits of Burrow 1989). ing Owls nesting at RMA, and how use of insects by Burrowing Owls is related to avail MATERIALS AND METHODS ability. This will allow a hetter nnderstanding Food Habits of the effects of environmental cleanup at RMA on Burrowing Owls. Our second objec Food habits were studied hy analysis of tive was to investigate potential differences in regurgitated castings and prey remains. Cast diet from analyses of castings and prey ings were collected and prey remains record remams. ed at hurrows biweekly from 14 June to 9 August 1990 (n = 19 burrows), and from 5 STUDY AREA April to 26 July 1991 (n = 28 burrows). Cast ings were separated by date of L""Ollection and RMA is located in south central Adams by burrow. Castings were soaked overnight in County, Colorado, and encompasses 6900 ha. a 2 molar (8%) NaOH solution, leaving only As part of the High Plains district of the bone and chitin (De~n 1978). Contents were
1Dcp;lltlllenl of Illlugc aud wildli'" Man:>gctllcnL Texas Tr.ch \.>Ili~enily, L.uI.lI
299 300 GREAT BASIN NATURALIST [Volume 53 separated and enumerated to the lowest possi orders. In 1991, 680 castings contained 944 ble taxon, usually family. Burrowing owls nor prey individuals from 9 families in 3 orders. mally consumed only a portion of many prey Invertehrates made up 48.4% ofdiets in 1990, items. For this reason, the percentage of a and 61.2% in 1991 based on casting analysis given prey item in the diet was based on fre (Table I). Prey remains for 1990 and 1991 quency of occurrence, rather than an estimate included 187 individuals from 18 families in 6 ofpercent biomass. orders (Table 2). Data were tested within years fix among Use ofonly tiger beetles (Cicindelidae) dif burrows variation in use of each insect family fered (P ~ .OOl) among burrows in 1990. No fcmnd in castings. Wilcoxon ranked-sum tests differences (P > .05) wcre found among bur werc uscd (SAS Institute, Inc. 1988). rows in 1991. Insect Use and Availability Insect Use and Availability Pitfitll traps were used to determine rela In 1990, 705 insects from 10 families were tive abundances of insects near occupied bur captured in pitfall traps, and 7 families were rows (Wolda 1990). Traps consisted of six 10 represented in castings, including 1 family not oz dear plastic cups buried flush with the soil captured in pitfall traps. The X2 test indicated surface, placed at I-m intervals from all bur a difference between owl use of insects and rows (n ~ 47) in a randomly selected azimuth. pitfall-trapped insects (X 2 ~ II,963, P < This ensured that insect sampling occurred in .00Ol, 10 dl), implying that Burrowing Owls the birds' foraging microhabitat (Hutto 1990). selected insect families disproportionately to 'n-aps were set for one week and were Hlled to numbers captured in pitfall traps. Specifically, a depth of ahout 4 cm with a 25% solution of differences (P < .05) were found in use of ethylene glycol to act as a preservative. Two short-horned grasshoppers (Acrididae), trapping intervals were used, separated by ground beetles (Carabidae), camel crickcts one month. (Gryllacrididae), and carrion beetles (Silphi Data were tested for differences between dae) in 1990 (Table 3). trapping intervals and years using Kruskal/ In 1991,4692 insects from 13 families were Wallis and Wilcoxon 2-sample tests (SAS captured in pitfall traps, and 7 families were Iostitute, Inc. 1988). No differences (P < .05) represented in castings. As in 1990, the X2 were detected between trapping intervals, so test indicated a difference between owl use of data were pooled. A ycar effcct (P < .05) was insects and pitfall-trapped insects (X2 ~ 643, detected for hoth insect use and availahility, P < .0001, 12 dl), but we found no differences 2 so data wcre analyzed by year. AX goodness (P > .05) in use of specific insect families of-fit test was uscd to test for overall differ (Tahle 4). Although more than 20% of cate ences in use and availability. When significant gories contained less than five expected differences were detected, Bonferroni confi observations (Dixon and Masscy 1969:238), dence intervals were used to identify differ the average expected observation was well ences within individual insect family classes over six for all categories, whkh made the use (Neu et aI. 1974, Byers ct aI. 1984). of this approximation appropriate (Roscoe and Castings Versus Prey Remains Byars 1971). Frequency of occurrence of prey items in Castings Versus Prey Remains castings and as prey remains was tested for Differences were detected in castings and between-year differences. Yearly differences prey remains for both 1990 and 1991. During and differences in contents of castings and both years combined, spadefoot toads (Salien prey remains were tested using KruskalIVVallis tia) and thirteen-lined ground squirrels (Sper and Wilcoxon 2-sample tcsts (SAS Institute, mophilus tridecemlineatus) occurred exclu Inc. 1988). sively (P < .05) as prey remains, as did all moths (Lepidoptera) (P < .(lOOI). Passerine RESULTS remains were found once in castings, but more Food Habits often (P < .0001) as prey remains in both In 1990, 765 dissected castings contained years. During 1990 black-tailed prairie dog 759 prey individuals from II families in 4 (Cynornys ludovicianus) occurred exclusively 1993] BURROWING OWL FOOD HABITS 301
TABLE 1. Frequency and percent occurrence of prey items in castings of Burrowing Owls at the Rocky Mountain Arsenal, Colorado, 1990-91. 1990 1991 Total
Prey item N %N N %N N %N Vertebrates RODENTIA Cricetidae Oryptorn paroa 1 0.1 a 0.0 0 0.0 Microtus ochrogaster 40 5.3 66 6.9 106 62 Peromyscus rnaniculatus 272 35.8 192 20.3 464 27.2 Passeriformes 1 0.1 0 0.0 0 0.0
Subtotal 314 41.4 258 27.2 570 33.4
Invertebrates COLEOPTERA Carabidae I 01 20 2.1 21 1.2 Cicindelidae 54 71 41 4.3 95 5.6 Scarabeiclae 55 7.2 80 8.5 135 7.9 Silphidae 109 14.3 124 13.1 233 13.7 Tenebrionidae 84 ILl 273 28.9 357 21.0 ORTHOYrERA Acrididae 63 8.3 36 3.8 99 5.8 Gryllacrididae 2 0.3 5 0.5 7 0.4
Subtotal 368 48.4 579 61.2 947 55.6
Other Rocks 52 6.8 104 110 156 92 Glass 1 0.1 0 0.0 1 tl4l Vegetation 17 2.2 1 0.1 18 1.0 Eggshell 7 0.9 2 0.2 9 0.5
Subtotal 77 10.1 107 1L3 184 10.7
Total 759 99.9 944 99.7 1703 99.7
"tr - <0.1%. as prey remains (P < .005), while darkling cally by day and mammalian prey were hunt (Tenebrionidae), scarab (Scarabeidae), and ed exclusively at night. Haug and Oliphant tiger beetles occurred exclusively in castings (1990) never observed BUlTowing Owls forag (P < .05). Carrion beetles occurred in castings ing diurnally for mammals. Likewise, in over and prey remains, but in greater numbers (P 200 hours of observation, mammalian prey < .005) in castings. During 1991, Ord's kan deliveries to the burrow during daylight were garoo rat (Dipodomys ordii) was recorded witnessed only twice at RMA in 1990 and exclusively as prey remains (P < .05). Short 1991 (Plumpton 1992). Use of insect families horned grasshoppers occurred in castings and varied little between nesting burrows. Both prey remains, but to a greater extent (P < castings and prey remains revealed that deer .0001) as prey remains. Deer mice also mice were the single most important prey occurred in both, but to a greater extent (P < species on this site. .001) in castings. Insect Use and Availability
DISCUSSION Burrowing Owls at RMA took insects in the families Acrididae, Carabidae, and Gryl Food Habits lacrididae at a rate less than expected, while While Burrowing Owls took large numbers preying on the family Silpbidae more than of several families ofColeopterans, we believe expected based on pitfall trap results in 1990, that these prey items were taken opportunisti- In 1991 Burrowing Owls preyed on insect 302 GREAT BASIN NATURALIST [Volume 53
TABLE 2. Food habits of Burrowing Owls based on prey remains found during biweekly searches at the Rocky Moun- tain Arsenal, 14 Juoe-9 August 1990, and 5 April-26 July 1991. 1990 1991 Total
Prey item N %N N %N N %N Vertebrates HODENTJA Cricetidae Microtus ochrogaster 11 8.5 II 59 Perumyscus maniculatus 40 67.8 10 7.8 50 26.7 Sciuridae Cynornys ludovidanus I 1.7 I 0.5 SrJennophilus tridecemlineatus I 1.7 I 0.8 2 1.1 I-Ieteromyidae Dipodomys ordii I 0.8 I 0.5 PASSEHIFOHMES Alaudidae Eremophila alpestris 4 6.8 5 3.9 9 4.8 Emherizidac Sturnella neglecta 3 5.1 3 2.3 6 3.2 Icterus galbula I 0.8 I 0.5 Mllscicapidac Cathams spp. I 0.8 I 0.5 Unknown spp. I 1.7 8 6.2 9 4.8 SQUA\IATA Viperidae Crotalus viridL~ I 08 I 0.5 SALlENTJ.>\ Scaphiopus spp. I 1.7 2 1.5 3 1.6
Subtotal 51 86.5 44 34.2 95 50.6
Invertebrates COLEOl"rEHA Carabidae I 0.8 I 0.5 Cicindelidae 2 1.5 2 1.1 Scarabeidae 2 1.5 2 l.l Silphidae 2 3.3 8 6.3 to 5.3 Tcnebrionidae 2 3.3 25 19.5 27 14.4 LEPIDOPTERA Citheroniidae Hyalophora cecropia I 0.8 I 0.5 Satllrniidae Antheraea polyphemus 1 1.7 I 0.5 Sphingidac Cclerio lineata 8 6.3 8 4.3 Noctuidae 2 3.3 3 2.3 5 2.7 OJrrlJOfYrEHA Acrididae I 1.7 34 26.6 35 18.7
Subtotal 8 13.3 84 65.6 92 49.2
Total 59 99.8 128 99.8 187 99.7
families as expected based on pitfall trap bones may not have been ingested. \Ve sus results. pected that larger items, such as prairie dog and prairie rattlesnake (Crotalus viridis), were Castings Versus Prey Remains scavenged from roadsides. Other authors have Castings indicated greater use ofdeer mice recorded organisms from prey remains that aud beetles (Coleoptera), while prey remaius were not evident from castings (Thomsen indicated greater usc of mammals larger than 1971, MacCracken et al. 1985). Recordiug deer mice, toads, and passerines, from which prey remains while collecting castings 1993] BURROWING OWL FOOD HABITS 303
TABLE 3. Use ofinsect prey by Burrowing Owls at the Rocky Mountain Arsenal, Colorado, 1990. lnsect Obsetved proportion Expected proportion Bonferroni interval famjly in castings in castings on proportion
Acrididae .171 .235 .U5 s P:5 .22711 Carabjdae .002 .404 -.005 < P ~ .010" Cicindelidae .147 .0 .094S PS .199" Coccinelidae .0 .001 Os psO Curculionidae .0 .005 Os P~O Elateridae .0 .001 OS P~O GryIIacrididae .005 .167 -.005:5 P S .OlGa Lygaeidae .0 .001 aS PSQ Scarabeidae .149 .078 .096 S P S .202' Silphidae .295 .004 .228 S P S .364" Tenebrionidae .228 .101 .165 S P S .291n
'SigniGc:ant ilt P < .1)5,
TABLE 4. Use ofinsect prey by Burrowing Owls at the Rocky Mountain Arsenal Colorado, 1991. Insect Ohsetved proportion Expected proportion Bonferroni intclval family in castings in casti.ngs on proportion Acriclidae .062 .175 .130 s P S .222 Carabidae .034 .092 .0575 P S; .127 Cicindelidae .071 .075 .044 s P < .108 Cernmbycidae .000 .001 -.001 s P S .002 Coccinelidae .000 .001 -.003 S P S .004 Curculionidae .000 .001 -.003 S P S .005 Elateridae .000 .003 -.004 S P S .010 Gryllacrididae .008 .021 .004 S P S .039 Lygaeidae .000 .014 .010 S P S .029 Mantidae .000 .001 -.OOiS P ~ .002 Scarabeidae .138 .371 .313 S P S .429 Silphidae .214 .089 .055 S P ~ .124 Tenebrionidae .471 .154 .llOS P~ .197
requires little additional effort and may reveal LITERATURE CITED use of prey items beyond those found in cast L. mgs. BROWN, 1974. Data required for effective study of rap· tor populations. Pages 9-20 in F. E. Hamerstrom, Jr., B. E. Harrell, and R. R. Olendorff. eds., Man ACKNOWLEDGMENTS agement of raptors. Raptor Research Foundation, Vennillion. South Dakota. We thank D. J. Buford, L. Hahner, j. K. BYERS, C. R., R. K. STEINHORST, AND P. R. KRAusMAN. 1984. Clarification of a technique for analysis of llti Jones, Jr., K. Meaney, T. Nelson, L. Pezzolesi, lization·availability data. Journal of Wildlife Man C. Preston, and R. Sites for assistance in col agement 48: 1050-1053. lection and analysis of food habits and prey DECN. H. J. 1918. A new method of analyzing pellets availability. We also thank the u.s. Army and from owls. etc. Dansk Omithologisk Forenings Tids,krift no 143. the u.s. Fish and Wildlife Service for funding DIXON, W. J., AND F. 1- MASSEY. 1969. Introduction to the study. F. C. Bryant, S. Demarais, C. D. statistical analysis. McCraw-Hill, New York 370 pp. Marti, M. K. Rylander, R. C. Whitmore, and ENVI~ONMEN"AL SCIENCE AND ENCINEEJ\INC. 1989. two anonymous reviewers provided belpful Environmental setting, section 2.0. Rocky Mountain criticism of our manuscript. nus is conbibu Arsenal biota remedial investigation-dran final report, version 2.2. tion number T-9-627 of the College of Agri HAve, E, A., AND L. W. OUPHANT. 1990. Movements, cultural Sciences, Texas Tech University. activity patterns, and habitat use of Burrowing Owls 304 GREAT BASIN NATURAUST [Volume 53
in S'lskntt;l.eW;lJl. Journal of Wildlife Management tuin Arsenal, Colorado. Unpublished masler'.~ thesis, •>4. 27~'15. Tuxa.s Tech Univcrsity, Lubbock. 72 pp. IItJ11'O, It L. 1990. MeasurillK the availability of food Ho!)cm·:. J. T., AKD J. A, BYARS, 1971. An investigation of rCsourCtlS. Studies ill Avian Biology 13; 20-28. the restnlints wilh respect to sample size conllll(lIlly MACClv.<.:KEf". J. C., D. W. UnRsK. ANn It ,...1. HANSEN. imposed 011 lht: use of the chi-square statistic. JO\lr 1985. BurrowinK Owl fOod... in Conata Basin, South nal of the American Statistical Association 66: Oakol:t. Great Basin Naturalist 45: 287-290. 755-759. ~El:. C. W., C. H. Bn;RS. ,\NO J. :\i. P(,;EK. 1974. A teclJ SAS tt\STlTUTe, iNC. i988. SAS/STAT ust:r's guide, lliquc for analysis of utilization-availahility data. rdC'J.5c 6.03 edition. Cary, North Carolina. 1028 pp. Jnurnal ofWiJdlifc Management 38: 541-545. TUOMSEN, l.. 1971. Bchavior and ecology of Bunuwing OLENDOflff. R. R., AND J. W. ST(mOAnT. 1974. The Ow'Is on the Oakland Municipal Airport. Condor 73: potential for management of rnptor populations in 177-Hr2. western h"Tass!:mds. Pages 47-88 in F. . Hamer· WOLDA, H. 1990. Food availahility for an ins(,'Ctivore and strom, Jr., n. E. Harrell. and R. R Olendorff, eds., how to measure it Studies in Avian Bioloh'Y 13: Management of raptors. Raplor Research Founda 38-43. tion, V~nnillioll. South Dakota. PLUMPTON. D, L. 1992. Aspects of nest sitc selection and ReceiL'ed 9 November 1992 habitat usc I>y Burrowin~ Owls at the Rocky Moun- Accepted 11 March 1993