RaptorRes. 22(3):81-84 1988 The Raptor ResearchFoundation, Inc.

ROAD MORTALITY OF THE LITTLE OWL (Athene noctua) IN SPAIN

MAURO HERNANDEZ

ABSTRACT.--Westudied the causesof road mortality of the Little Owl (Athenenoctua) in Spain based on 418 road casualtiesand their relation to road characteristicsand type of roadsidecover. Road mortality presentsa clearseasonal pattern and a summerpeak (August) was found to beassociated with postfledging dispersion.Road illuminationdecreases road mortality.Roadside cover has an importantinfluence upon road casualties,possibly as a consequenceof perching.Higher numbersof deathsoccur in shrub type vegetationof < 1 m height. These resultsand thoseobtained by other authorsare discussed.

The Little Owl (Athenenoctua) is the most gen- Central Spain.A total of 418 collisionswith carsin Spain eraliststrigiforme in the Iberian Peninsula,and the were studied. Position, date of death and, when known, age of each owl casualtywere recorded.Features of the most widely distributed and most common owl in road and the type of roadsidecover were also recorded, Spain. In spite of this, many aspectsof the biology includingthe presenceor absenceof illumination,the pres- of this speciesare unknown. enceor absenceof road margin vegetationand its features, The Little Owl waspoorly documented (Valverde and the type road. Road margin vegetationwas studiedonly in non-illu- 1967; Agelet 1979; Hoz 1982) until the studiesof minated roads in an area of 20 m width from the road Mafiez (1983a and 1983b). No information is avail- border and 10 m length on each sidefrom the locationof able on the species'status in Spain, and mortality the carcass.Vegetation y 3 m in height was consideredas has neverbeen studied in Spain (Clark et al. 1978). woodland. If no homogeneousvegetation was found, the Mortality of strigiformesin Spainwas poorly treated presenceof only I tree was deemedsufficient for classi- fication as woodland. The other class of vegetation was by Garzon (1977) and, particularly on Bubo bubo shrub vegetationwhich was ':3 m height. (see Gonzalez et al. 1980). Westernhagen(1962) In many casesno information was available on the reportedon numerousroad casualtiesof kites (Mil- distance from where a bird was found to the next curve vus sp.) in Andalusia (southernSpain) and Hass in the road, and so,the typesof roadshave been classified (1978) includedSpanish roads in his study. by calculatingthe numberof curvesin 10 km of road. On this basis, three road classesresulted as follows: 1) roads In Europe, particularly in Great Britain, Ger- with no more than 10 curves/10 km; 2) roads having many, the Netherlandsand Denmark, road mortal- between 10 and 25 curves/10 km; and 3) roads with >25 ity amongbirds is well documented.Hodson (1962) curves/10 km. studiedthe varioustypes of roadsidecover and its RESULTS influence upon bird casualty numbers. Dunthorn Of 418 total collisionsrecorded during the study, and Errington (1964) made an attempt to measure 40 were excludeddue to insufficientdata being avail- the effectsof such deaths on the bird population. able. Mortality causedby car collisionsrepresents Hansen (1969) divided roads into categoriesusing 82% of non-natural deaths in Athene noctua, al- trafficdensity as a normand, lastly, Brautigam (1978) discussed the seasonal distribution of road deaths. thoughin somecases this value may be considerably higher. The remaining 18% were attributed as fol- Nevertheless,the Little Owl appears only occa- lows: shot (4.3%); caught in nest (3.1%); trapped sionally in thosesamples, and for this reasonthe (0.9%); preyedupon (3.5%); unknown (1.9%) and resultsof theseobservers are dubiousfor application othercauses (4.3%), includingfalling from nest,being to the speciesand to Spain.The purposeof thisstudy waterlogged,cannibalism and starvation. was to establishthe causesof road mortality of the Resultsconcerning the seasonalvariation in mor- Little Owl in Spain and to advanceknowledge on tality are given in Figure 1. The time of greatest non-naturalmortality of this species. mortality was recordedin summer.Between i Au- MATERIALS AND METHODS gustand 15 Augustwe found63.2% of the 378 deaths Data were compiledin a 3-yr period (1984-1986) in studied; 17.1% between 15 July and 31 July, and

81 82 MAVROHERNANDEZ VOL. 22, NO. 3

70

63,2

N = 378

17,1

'-71____1I 0,20,2I I 0I 0I 0I i'-71•I [ 0I 0,7ø•I I 0,7• [ I [ I I I [ 0I 0,,__I 0 I '•'•'• f 1 2 1 2 I 2 1 2 I 2 1 2 1 2 1 2 1 2 1 2 1 2 I 2 JAN FEB MAR APR MAY JUN JUl. AUG SEP OCT NOV DEC

MONTH Figure 1. Histogramsshowing the pattern of seasonalmortality by half-monthperiods.

7.1% between1 July and 15 July. During the re- of vegetationdefined were alternativelydistributed mainderof the year (21 2-wk periods)were found in which casecarcasses were foundonly in areasof the remaining 12.6% (47 deaths). shrub vegetation. With respectto age of Little Owls found, among 353 (93.3%) foundbetween 1 July and 1 November, DISCUSSION only 25 (7.0%) were adults, the remainder being According to the available data, road casualties juvenilebirds identified by their characteristicplum- are the most important causeof non-natural mor- age. tality of the Little Owl in Spain. Nevertheless,it is All car collisionsoccurred during the night. On possiblethat road deathsmay have been over rep- illuminated roads only 9 Little Owl deaths(2.3%, resentedsince birds killed on the roadare morelikely N = 378) were recorded,all of them being young to be found than those killed in most other ways owls; 369 (97.6%) were found on non-illuminated (Hodsonand Snow 1965). In Europedeaths on roads roads. are not as frequentbut representan important frac- Power transmissionor telephonelines parallel to tion of owls killed by human influence(Glue 1971; roads were not associated with car collisions and are Weir 1971; Glue and Scott 1980; Exo and Hennes generally placed <20 m from the road. 1980). Of the 369 deathsfound on non-illuminatedroads, Collisionswith carscan be producedwhen bright in 75 cases(20.3%) there was woodlandvegetation lights at night causetempor/try blindnessof owls in the road margin area and in the remainder (294 (Labisky 1960, in litt.) and asa resultof owlsswoop- cases,79.6%) the vegetationwas of the shrub type. ing towards a moving light (Hodson 1962). Sutton We found 184 deaths(62.5%) occurredon roadswith (1927) suggestedthis as a causeof ScreechOwls vegetationof no more than 0.5 m height and 110 (Otusado) being killed by cars. For thesereasons deaths(37.4%) occurredon roadswith a vegetation and perhapsothers, all road casualtiesoccurred at of 0.5-1 m height in road margins.On roadswith night (M. Mafiez, pets. comm.) and are also the a vegetationheight of 1-3 m no deathswere found. main differences in road deaths between diurnal and Often in the marginof severalroads the two types nocturnalbirds. Along illuminatedroads, mortality AUTUMN 1988 LITTLE OWL MORTALITY IN SPAIN 83

is less(only 2.3%), and possiblyLittle Owls are not Table 1. Number and percentageof deathsof the Little blinded in illuminated areas or are so with less crit- Owl in Spain associatedwith variousroadway ical intensity. conditions. Hansen (1969) suggestedthat bird behaviourde- termineswhether a speciesis liable to be killed in ROAD CONDITION N % traffic or not, and perhapsthis can explain why the Road Characteristic Little Owl is more frequently killed by cars than Illuminated 9 2.3 other owls. Road mortality presentsa clear seasonal Non-illuminated 369 97.6 pattern. In the first half of Augustwe found a very Total 378 99.9 conspicuousincrease in mortality (Fig. 1) producing Roadside Cover 63.2% of the annual deathsin only 15 d approxi- Woodland 75 20.3 mately.Summer mortality has alsobeen reported by Shrub 294 79.6 Glue (1973, in litt.) and Exo and Hennes (1980). In Total 369 99.9 our study,however, the summerpeak occurredlate in August and was not associatedwith stressof the Shrub Height 0.0-0.5 m 184 62.5 breedingseason as suggestedby Exo and Hennes 0.5-1.0 m 110 37.4 (1980) due to a greaterproportion of youngin our 1.0-3.0 m 0 0 sample(92.9% betweeni July and 1 November), Total 294 99.9 but to postfledgingdispersion occurring during this time (Glue and Scott 1980). Glue and Scott (1980) Road Classification Class 1 32 8.4 found that between1 July and 1 September40% of Class 2 281 74.4 youngdied, mostof them on roads.Dunthorn and Class 3 65 17.1 Errington (1964), Haas (1964), Hodsonand Snow Total 378 99.9 (1965) and Hansen (1969) have also reported an increasein road casualtiesduring this time frame and note the influence of a greater proportion of inexperiencedyoung on the observedincrease. Speed and density of traffic have an important The trend of the Little Owl to frequent openareas influenceupon bird road casualties(Dunthorn and influencesthe relationshipbetween number of deaths Errington 1964) and the greatestmortality is reg- and roadside cover. Nevertheless, we think that, in istered on roads where traffic is dense and fast mov- this study,the influenceis very small due to the ing (Hansen 1969). The number of casualtieson small amountof openregions of the studyarea. The Class 2 roads (Table 1) is higher than on those of influenceof road margin vegetationon casualtiesis Class3 and Class1. Nevertheless,the speedof traffic a consequenceof perchavailability because owls are is the higheston Class I roadsand is the loweston rarely killed while crossingroads. Vegetation of < 1 Class3. This may be explainedby the fact that on m is classifiedas shruband is not adequatefor perch- Class2 roadsthere are enoughcurves to preventthe ing and causesLittle Owls to use the pavementas owls from seeingcars far away, and also preventing a perch.Hodson (1962) establishedthat muchhigher carsfrom reachinghigh speed,while on Class3 roads number of deathsoccur opposite gaps and openings the speedshould be low and on Class I roads cars than along stretchesof roadway with uniform bor- are seenby the owls with time enoughto leavethe der, becausemost birds are killed while attempting road. to fly from one sideof the road to the other through Finally, as a practical measureto avoid car col- thesegaps (Dunthorn and Errington 1964). lisionsof Little Owls, it is advisableto place trees The presenceof treesin the proximity of a road or shrubsnear roads,or if it is possible,2 m height (<20 m) was associatedwith decreasedroad mor- perchesshould be placed 5 m from the road border tality (20.3%). However, the most adequatevege- and 10 to 15 m apart. tation, accordingto the informationobtained is that ACKNOWLEDGMENTS between I and 3 m height and located at <20 m I thank Eduardo Castrillo, Miguel A. Carro and Al- from the roadand not woodlandas found by Hodson berto Fernandez-Arias for critically reading an earlier (1962) for other birds.In many casesthe perchwas versionof the manuscript,Manuel Manez for his com- not a tree or a shrub but a wall or a barrier. ments and suggestionsand Juan M. Hernandez for the 84 MAUROHERNANDEZ VOL. 22, No. 3

English translation. I acknowledgethe support for this HAAS,W. 1964. Verlustevon Vogeln und Saugernauf study,provided by the Asociacionpara la Defensade las Autostraben. Orn. Mitt. 16:245-250. EstrigiformesIbericas--ADES. I also appreciatethe as- HANSEN,L. 1969. Trafikdodeni dendanske dyreverden. sistanceof ADENA 2000 (Barcelona); ANAT-LANE Dansk orn. Foren. Tisskr. 63:81-92. (Navarra); ANES (Toledo); M. A. Carro; S. Cerradelo; HODSON, N. L. 1962. Some notes on the causesof bird J. C. del Olmo; C. Fernandez;A. Fernandez-Arias;J. L. road casualties.Bird Study9(3):168-172. Garcia;J. Izquierdo;V. Machuca;O. Maroto; J. I. Mar- -- AND D. W. SNOW. 1965. The road deaths en- tin; A. Melero; C. Palacin; D. Saavedra and I. Sanchez. quiry, 1960-61. Bird Study12(2):90-99. LITERATURE CITED Hoz, M. DE LA. 1982. Algunosdatos sobre la alimen- AGELET, A. 1979. Notas sobre la alimentacion de Mo- tacion de Athene noctua en Asturias. Bol. C. Nat. IDEA chuelo comun--Athene noctua(Aves, Strigidae). Misc. 29:113-120. Zool. 5:186-188. LABISKY,R. F. 1960. A techniquefor capturing birds BRAUTIGAM,H. 1978. Vogelverlusteauf einer fernver- and mammals.Biol. NotesNo. dO. Nat. Hist. Survey kehrstrabevon 1974 bis 1977 in den kreisenAtenburg Div., Univ. Illinois. und Geithain. Orn. Mitt. 30:147-149. MANEZ, M. 1983a. Espectroalimenticio de Athenenoc- CLARK,R. J., D. G. SMITH AND L. H. KELSO. 1978. tua en Espafia.Alytes 1:275-290. Working bibliographyof owlsof the world. Nat'l. Wildl. 1983b. Variacionesgeograficas y estacionales Fed. Sci. iv Tech.Series, No. 1. Washington, DC. dela dietade Athene noctua en Espafia.Actas XV Cong. CRAMP, S. A. (ED). 1985. The birds of the Western Fauna Silv. Trujillo 1981. Pp. 617-634. Palaearctic.Vol. IV, pp. 514-525. Oxford Univ. Press, SPENCER,K.G. 1965. Avian casualtieson railways.Bird Oxford. Study12(3):257. DUNTHORN, A. A. AND F. P. ERRINGTON. 1964. Cas- SUTTON,G. M. 1927. Mortality among ScreechOwls ualties amongbirds along a selectedroad in Wiltshire. (Otusasio) of Pennsylvania.Auk 44:563-564. Bird Study11(3):168-182. VALVERDE,J. A. 1967. Estructura de una comunidad Exo, K. M. AND R. HENNES. 1980. Beitrag zur Po- de vertebradosterrestres. Consejo Superior de Inves- pulationsokologiedes Steinkauzes (Athene noctua). Die tigacionesCientificas, CSIC, Madrid. Vogelwarte30:162-179. WEIR, D. N. 1971. Mortality of hawks and owls in GARZON,J. 1977. Birds of prey in Spain: the present Speyside.Bird Study18(3):147-154. situation. Pages 159-170. In R. D. Chancelor, ED. WESTERNHAGEN,W. 1962. Greifvogel in Andalusien. World Conferenceon birds of prey. Report Proceed- Orn. Mitt. 14(3):162-169. ings, ICBP. GLUE, D. E. 1971. Ringing recoverycircumstances of Asociacionpara la Defensa de las EstrigiformesIbe- somesmall birds of prey. Bird Study18(3):137-146. ricas--ADES, P. Conde de los Gaitanes, 145, La -- AND D. SCOTT. 1980. Breeding biology of the Moraleja, 28109 Madrid, SPAIN. Little Owl. Brit. Birds 73:167-180. GONZALEZ,J. L., J. LOBON,L. M. GONZALEZY F. PA- Received1 February 1988; accepted15 July 1988 LACIOS. 1980. Datos sobre la evolucion no natural de la mortalidadde Bubobubo en Espafiadurante el pe- riodo 1972-1980. Bol. Est. C. Ecol. 9:63-66.

Raptor Protectionand Research--AlbemarleFarms of Charlottesville,Virginia, in conjunctionwith Virginia Poly- technicInstitute and StateUniversity (VPISU) in Blacksburg,is establishinga preserveto be usedas a training facility for internsin wildlife managementat VPISU. Raptor researchand educationfoundations, along with organizations concernedwith the protectionof hawksand othermigratory birds are invitedto makeproposals and initiate discussions for cooperativeefforts aimed at the developmentof further techniquesfor protectingpredatory birds. For more information,contact John Williamson, Albemarle Farms, Route 6, Box 69, Charlottesville,Virginia 22901 (804)293- 3978 or Fred Garcia (212)557-0100.