DIET OF CETTI'S CElTIA CElTI (TEMMINK, 1820) IN A LOCALITY OF SOUTHERN SPAIN

LA ALIMENTACION DEL RUISENOR BASTARDO CETTIA CEITI (TEMMINK, 1820) EN UNA LOCALIDAD DEL SUR DE ESPAÑA

Javier MOLINA*,Josd A. H~DAR*' e Ismael CAMACHO*

Diet is one of the less-known biological fea- minguensis and Iris pseudacorus, with pro- Nres in a high number of small insectivorous nounced variations in water level due to the bis. This is due mainly to the difficulty in re- imgation schedules for nearby croplands. ln cognising their prey, usually tiny soft-bodied the periphery. stands a hedgerow with Tamarix arthropods that decompose quickly during di- sp., willows Salix fragilis, castor-oil plants Ri- gestion. In spite of this, intensive studies of cinus communis, and red Phragmites commu- many of northem or central-European passeri- nis and Arundo donax. Crops in the sumun- nes has provided substantial information on the ding area are mainly sugar cane and orchards of diet in these ; however, data remains chirimoya and avocado, although some pea- scanty for Mediterranean counüies, where die- ches, grapes and plums can be found. tary information is completely unavailable for The study was based exclusively on faecal some species. analysis. We obtained the samples by catching One such species is the Cetti's Cet- the biuds in mis1 nets, set twice monthly from tia cerri, a svictly linked to ponds, March 1995 to October 1996. Mist nets were lakes, mmhes and rivers, whose altitudinal dis- placed for 48 h each time. in the periphery of üibution in southern Spain ranges from O to the marsh. Trapped biids were kept in a uans- 1000 m a.s.1. (Jourdain, 1936, 1937; Plegue- pirable cellophane bag in a dark, quiet site until zuelos, 1992). Only scant dietary information is the deposition of excremenis. and aftenvards available from some localities within its distri- the birds were ringed and released. Ringing bution area, mainly England and France in allowed us to avoid collecting more than one Western Europe (Ghroudet, 1963; Hollyer, excrement per month from the same , thus 1975; Harvey, 1977; De Lust, 1979; Bibby, ensuring the statistical independence of 1982; Bibby & Green, 1983) and the republics monthly diet samples. of the former Soviet Union (see references in Faeces were dispersed in water and exami- Cramp, 1992). In most of these cases, data are ned under a 10x-40x binocular microscope sparse and fragmentary, with small sample si- equipped with a micrometer. The remains were zes. The aim of the present work is to increase separated into vegetal and fractions, and knowledge conceming the diet of this bird, their relative volumes estimated. Animal prey with results from a locality of southem Spain were usually detennined to the lowest taxono- throughout a seasonal cycle. mic level possible. Measurements of characte- The field work was camed out at the site ristic prey parts remaining in the faeces allowed known as Charca de Suárez, near the mouth of an estimate of the prey's body size, by means the Guadalfeo river (Motril. Granada province, of a series of regression equations previously Spain). This area is siNated 1 m a.s.1.. less than developed (Hódar. 1997). Dueto the low num- 200 m fmm the seashore, and covers ca. 8 ha. ber of items, we did not anempt to estimate the The central part is water-logged and almost biomass provided by each group. We obtained completely covered by reedmace Typha do- size estimates for 24 prey belonging to eight of

* Depaitamento de Biología Animal y Ecología. Facultad de Ciencias. Universidad de Granada. E-1 8071 Granada, España ' Author for correspondence. E-mil: jhodarC3goliat.ugr.e~. 218 MOLINA. J.: HODAR. J. A. & CAMACHO, 1. the taxonomic groups distinguished in the diet Coleoptera others. Aphididae with Homoptera (Araneae, Opiiio~da,Homoptera othen, Hete- others) or, when this was not possible. by crea- roptera, Neuroptera. Carabidae, Coleoptera ot- ting a new row for miscellaneous arthropods hers and Formicidae). These eight groups re- (Araneae, Opilionida. Lepidoptera larvae and oresent IWO chisof the total prev found CTable Neuroptera). Non-parameuic tests were applied i),so that the sample could'be-considered as because of the low sample sizes (Zar, 1996). representative in spite of its low size. The analysis of 41 excrements gave a total The differential digestibility of soft- versus of 114 animal prey, as well as many vegetal re- hard-bodied prey hampers the identification of mains. Only one piece of grit was found. Animal the former by faecal analysis, and rnay produce prey were predonunantly Hexapoda (73% in fre- severe biases in the estimates of the overall im- quency), whereas Cmstaceae (13.5%). Arach- portance of these two types of prey in diets nida (7.2%) and Gastropoda (6.3%) were less (se e.g. Custer & Pitelka, 1975; Jenni et al., numerous. Among insects, the most prevalent 1990; Pulido & Diaz, 1994). as well as in the groups were Coleoptera, Homoptem, Heteropte- estimation of prey lengths, since hard-bodied ra and Formicidae. lt is remarkable that some of prey leave more measurable remains. Rather the most abundan1 groups in the diet, such as surprisingly, rnost work on bird-diet analysis crustaceans and beetles, are hardly sclerotysed, ignore these digestibility effects. This may be which favours theu identification in faeces. Co- because the bi&es caused by these differential rrection factors applied to ants. aphids and spi- rates of aassaee and dieestibilitv remain scan- ders (from 1.5 to 2.5 following F'ulido & Día, tily doc;meGed and fhus po;ly understood 1994 for Blue Tits Parus caerdeus) would have (Rosenberg & Cooper, 1990), and therefore given a very different picture, with much more authors prefer to offer cmde data instead of co- weight of these groups in the Cetti's Warbler rrected data in which they are no1 confident. diet. Nevenheless, it is difficult to assess to what In the present work, we have not applied co- extent these correction factors could be applica- rrection factors for differential passage rales ble from one bird species to another. and digestibilities extracted from the literature, The prey composition of the diet was quite since such facton appear lo be species-specific, constant over time, without signiiicant diffe nor have we computed such factors for the Cet- rences among the three seasons considered ti's Warbler because of logistic constraints; ho- (G,, = 11.6, P = 0.64). Similarly, no differences wever, we comment on the possible effecis of were discerned in prey size (H, = 1.3, P = such corrections on our results. 0.52). which proved very small, 3.0I0.4 mm of Dietary data were computed on a seasonal body length on average (n = 24). However, this basis (breeding, March-July; post-breeding, Au- lack of significan1differences may be the con- gust-October, and winter, November-February). sequence of low sample sizes. Furthermore, both as the percentage of occurrence (percenta- since most of the prey for which length was ge of faeces in which an item appeared) and as estimated were hardbodied insects, like beetles ¡he numenc frequency (pcentage of items be- and anu, the actual average body lengih would loneine to a orev class with reswct lo the total be different if more aphids or spiders were in- of Gey-itern< ~ósenber~& ~&per, 1990). Alt- cluded. There was a^notewonhy presence of hough some of the samples for each season may Amphipoda, a cmstacean group linked to coas- have belonged lo the same individual, implying tal habitats; in this case, the proximity to the some degree of pseudoreplication, no attempt sea and the salinity of the soil favoured the pre- was made to completely eliminate this problem, sence of this group in the study area. Both prey since this would have suooosed too low samole size and dieiatv comr>ositioncorresponded with sizes (but see above). ~&~arisonsof diet &m- the feeding p&tern;of the bird, which is pn- positions were made with G tests, whereas prey marily a ground-gleaner (Geroudet, 1963: No- sizes were com~aredwith the Kmskal-Wallis val. 1975; Hollyer, 1975; but see Hill, 1993). test. To apply the G test, we used a pling pro- A major pan of the diet was comprised of cedure in order to avoid empty cells in the vegetal remains. While Cetti's Warbler has anaiysis. This was done by combining rows ha- been considered up to now lo be almost exclu- ving empty cells with the most taxonomically sively insectivorous, we found vegetal food as proximate row (Carabidae and Scarabeidae with a basic dietary component for this bird. Even DlET OF CETn'S WARBLERS CET7IA CEi7IflEMMINK. 18201

Diet composition of he Cetti's Warbler in the Charca dc Suárez (Motril, SE Spain). Diet composition is de- tailed, by seasons, as percentage of occurrence (% P) and nurneric frequency (% F).. [Composición de la dieta del Ruiseñor Bastardo eri ia Charca do Su'rez (Motril. SE.Espnña). La coniposició~i de la dieta se expresa, derallada por esraciunes, conlo porcenroje de presencia í%PI y porcentaje de fre- cuencia (% F).]

Breeding Post-breeding Winter lReprodrccci

%P %F %P %F 90 P % F Arachnida Araneae Opiliotlidn

Hexapoda Ephetneroptera Homoptera Aphididae Homoptera ot hers Hereroprero Diptera hpidopiera larvae Neuroptercr Coleoptera Carabidae Chrysomelidae Scara beidac Coleoptera others Hynienupreru Furmicidue

Castempoda Seeds

Vegetal, % presence 30.0 81.3 60.0

Vegetal, % volume (X SE) 19 i10 63 * 9 39 I9 Prey lengih (mm,W I SE, n) [Longitud de lar presas] 2.9 í 1.3, 6 2.9 r 0.6,7 3.1 I 0.6, 11 Number of iaeces INiímero de e.rcremenlos] Number of prey [Nilmero de presas) 30 32 52 DETOF CEíII'S WARBLERS CI during the breeding period, 30% of samples the Ceni's Warbler. Cetfia cem'. Ibis, 124: 288- contained some vegetal remains, and especially 301. in the post-repductive season and winter, this CRAMP,S. @d.) 1992. The birds of fhe wesfem Pa- component reached considerable pmportions. learclic. Vol. VI. Oxford University Press. Ox- This may imply lower importante of animal ford. a Cusm, T. W. & F'RELKA, F. A. 1975. Comtion food, which was appreciable in the low number factors for digestion rates for preys taken by Snow of prey per excrement (x = 2.8, range O-@, in Buntings (Plecfrophenax nivalis). Candor. 77: contrast with the results found by other aut- -?Ih?l? .- - .- . hors. For instance, Bibby & Green (1983) DE LUST,R. 1979. A antekeningen bij het broedge found averages of 12.6 and 9.1 animal prey per drae van Cettia cetri. Wielewaal. 45: 47-51. excrement in two different zones of France. GEROÜDET, P. 1963. Les Passereaur 11. Des mésan- Although these data correspond to birds fatte- ges aru fawenes. Delachaux et NiesUt. Neuch2tel. ning for migration. and thus probably inges- Suisse. ting food at higher rates than the birds studied HAFMY, W. G. 1977. CeWs Warblers in east Kent in 1975. Brirish Birds. 70: 89-96. by us, some of the birds captured by us in the HILL. B. J. 1993. Ceni's warbler flvcatchine.- British post-breeding period would also have been fat- Birds, 86: 11. tening. The fact that most of the vegetal re- H~DAR,J. A. 1997. The use and usefulness of re- mains were unidentifiable, the scarcity of gression equations for estimation of prey length fleshy-fmiting wild plants in the study site (see and biomass in diet sNdies of insectivore verte- before), the scant number of seeds found in fa- brales. Miscel.lania Zuoldgica, 20: 1-10. eces and the maximum values reached in au- HOLLYER.J. N. 1975. The warblers in Kent. Keni tumn (August-October. Table 1). suggest that, Birds RepoH, 22: 84-95. Jmi, L., REU~MAN,P. & JWM-EIERMANN,S. 1990. in conmt with other similar buds that feed on Recognizability of different food types in faeces wild bemes and other fleshy fmits during au- and in alimentq flushes of Sylvia warblers. Ibis. hnnn and winter, a good part of the fmits con- 132: 445-453. sumed by Cetti's Warbler in our arca came JOURDAIN,F. C. T. 1936. The birds of southern from crops, such as peaches, grapes or plums, Spain. Ibis, 1936: 725-763. available near the study area. JOUROALN,F. C. T. 1937. The birds of southern Spain. Ibis, 1937: 110-152. ACKN0WLEDGEMENTS.-The members 0f the NOVAL,A. 1975. El Libro de la Fauna Iblrica. vol. group Buxus (especially M. Arellano. F. Tanagona 4. Ed. Naranco. Oviedo: and S. Valverde) helped us in the field work. M. PLEGUUUELOS,J. M. 1992. Avifauna nidificanfe de Díaz, F. Pulido and an anonymous reviewer made las Sierras BCricos Orientales Y de~resionesde valuable commens on an earlier version of the ma- Guadix, Baza y Granada. Su cÜrlo8'rafiado. Uni- nusnipt. David Nesbin revised the English version. versidad de Granada. Granada. This work was partially f nanced by a grant (lo Pmw, F. J. & Dkz. M. 1994. Diet and prey type J.A.H.) whithin the pmject AMB-0479 financed by selection bv adult and voune blue tits P&& &e- Ihe Spanish CICYT. ruleus: theéffect of cok¡& for prey digestibi- lity. Ardeola, 41: 151-159. ROSMBERO,K. V. & COOPER.R. J. 1990. Appma- ches to avian diet analysis. Srudies in Avian Bio- FmFmNCEs iogy, 13: 80-90. ZAR,J. H. 1996. BiosfatiFlical analysir, 3rd ed. Pren- BKIBY,C. J. & OREEN,R. E. 1983. Focd and fatte- tice Hall. Englewood Cliffs. New Jersey. ning of migrating warblers in some French marsh- lands. Ringing and Migration, 4: 175-184. [Recibido: 22-4-98] BDBY,C. J. 1982. Polygyny and breeding ecology of [Aceptado: 3-9-981