TheCondor96:170-177 8 The Cooper Ornithological Society 1994 CLUTCH SIZE IN PASSERINES OF SOUTHERN SOUTH AMERICA’ YORAM YOM-TOV Department of Zoology, Tel Aviv University,Tel Aviv 69978. Israel MIGUEL I. CHRISTIE AND GUSTAVO J. IGLESIAS Villegas369, 8400 Bariloche,Rio Negro,Argentina Abstract. Clutch size of the passerinesof southern South America was studied by com- paring data taken from the literature for 33 1 species(177 genera), which breed in Chile, Argentina, Uruguay, and southernBrazil. Mean overall clutch size was 2.98 eggs,appreciably smaller than clutch size of passerinesat similar latitudes in the northern hemisphere. There was a highly significantdifference between mean clutch size of the two sub-orders representedin the study area, with the Deutero-Oscineshaving on averagesmaller clutches (2.73) than the Oscines(3.32). This difference between the two sub-orderswas maintained in each of the five zoogeographicalregions of the study area. There were also significant differencesin clutch size between the five main zoogeographicalregions: overall mean clutch size was smallest in the Amazonianregion (2.67) and largestin the Chaco(3.29). Deutero- Oscineseggs were proportionally larger than thoseof Oscines(14% and 11%of bodymass, respectively). Therewas no correlationbetween clutch size and body mass, nor was clutch size related to either nest type or migration. There is no evidence for latitudinal gradient in clutch size in the area studied. Key words: Clutch size; Passeriformes;South America. INTRODUCTION clutch size increasesby as much as one egg per Clutch size varies both between and within spe- 6-10”latitude. Information about the occurrence cies, and the reasons for these variations have of this phenomenon in the southern hemisphere been the subject of many studies. Clutch size is is scarce,but Moreau (1944) showed that clutch constrained by several factors, which can be sizes of populations in South Africa average less groupedinto two main categories-ecologicaland than one egg more than those of populations in phylogenetic or taxonomic (Klomp 1970; Lack equatorial Africa, about 30” to the north. Simi- 1947, 1968; Perrins and Birkhead 1983). The iarly, Yom-Tov (1987) has shown that a very most commonly discussed factors in the first cat- weak but significant trend exists in eastern Aus- egory are latitude and habitat, while relatively tralia, where mean clutch size of passerinesin- little has been done concerning the connection creasesby about 0.8 eggover 30”. Hence, it seems between phylogeny and clutch size. that although a latitudinal trend in clutch size The trend of increase in clutch size of passer- exists in both northern and southern hemi- ines with latitude is one ofthe most striking vari- spheres,it is much more marked in the northern ations between and within speciesof birds, and hemisphere,where clutch size increasesat a much is especially marked in multibrooded passerines steeperrate. Clutch size is also related to habitat: (Perrins and Birkhead 1983, Murray 1985). within the tropics, clutches tend to be larger in Clutches of tropical passerinestend to have two savanna and grassland than in the rain forest to three eggs(Moreau 1944, Skutch 1985), those (Lack and Moreau 1965), and in temperate breeding in middle latitudes laying four or five regions clutches of great tits are larger in decid- eggs,and those in the Arctic five to seven (Rick- uouswoodland than in pine forests(Klomp 1970), lefs 1969). This trend was studied extensively in and blackbirds lay larger clutches in rural than the northern hemisphere, where it was shown to in urban sites (Snow 1958). Other ecologicalfac- occurin both the Old (Lack 1968) and New (Cody tors such as population density (Perrins 1965) 197 1) worlds and Klomp (1970) claimed that and type of nest (Klomp 1970) have also been shown to influence clutch size. Several hypotheseshave been suggestedto ex- I Received24 May 1993. Accepted 28 September plain the latitudinal trend and habitat type in- 1993. fluence (herein termed as “ecology”) on clutch [I701 CLUTCH SIZE OF SOUTH AMERICAN PASSERINES 171 size. Murray (1985) distinguished between two phylogeny and ecology. However, becausethere alternative general hypothesesregarding the evo- are almost no published data on clutch size in lution of clutch size, per se. The view developed relation to latitude in this vast region, we studied by Lack (1947, 1968) and his followers (Ashmole clutch size in relation to habitat type, which 196 1, Cody 1966, Ricklefs 1980) statesthat birds probably has a major effect on the ecologyof the reproduce up to the maximum energy that the birds. Clutch size is known to be affectedby type parents can devote to laying eggsand rearing the of nest (Perrins and Birkhead 1983), and we ex- young. Excessreproduction is eliminated by sub- amined this as well as our hypothesis that mi- sequent density-dependent mortality. A second gratory passerines should have larger clutches group of views comprising the ideas often re- than residents. ferred to as “adjusted reproduction,” according More than 460 species(about 240 genera) of to which the clutch size is adjusted to the mor- passerinebirds belonging to 25 families breed in tality rate of the species,with those having long southern South America. Eight of these (Den- life expectanciesalso having small clutches and drocolaptidae, Furnariidae, Formicariidae, vice versa. There is, however, little agreement Rhinocryptidae, Cotingidae, Pipridae, Tyranni- on how this adjustment evolves (Stresemann dae and Phytotomidae) belong to the infraorder 1927-1934, Rensch 1938, Skutch 1969, Wynne- Tyranni (Sibley and Ahlquist 1985) of the sub- Edwards 1962, Murray 1985). order Deutero-Oscines (Voous 1977), and the Although the single most important factor pro- rest belong to the suborder Oscines (Passeres). ducing variation in clutch size between species The Tyranni evolved in South America after it is their ecology (i.e., food availability and pre- was separated from the Old World portions of dation; Perrins and Birkhead 1983) there are Gondwanaland during the Cretaceous period, major differences in clutch size between taxo- about 80 million years ago. The Oscines evolved nomic groups. Lack (1968) noted that some or- in parallel. Some groups (i.e., Emberizidae, Ic- ders have a typical clutch size which is common teridae, Parulidae) in the New World, either in to most of their species.For example, Procellar- South America (i.e., Thraupidae, Icteridae, Vire- iformes lay only one egg,Columbiformes mostly onidae) or North America (i.e., Emberizidae), two. The songbirds (Passeriformes)have a vari- while others (i.e., Hirundinidae, Corvidae, Cin- able clutch size, and relatively little researchhas clidae, Turdidae) evolved mainly in the Old been carried out on the effect of phylogeny on World and dispersedto North America and then clutch size in this group. However, even within South America (Sibley and Ahlquist 1990). this order some families have larger clutchesthan Our study area includes Chile except for the others. For example, among the Australian pas- Tarapaca region in the tropical north, Argentina, serines, the “old endemic? families (i.e., Meli- Uruguay, and Rio Grande de Sul, the southeast phagidae, Maluridae and Acanthizidae) have province of Brazil, and excluding all oceanic is- clutches of 2-3 eggs,while the “new invaders” lands. This region (about 3.95 million sq. km) (i.e., Hirundinidae, Ploceidae and Corvidae) lay stretchesabout 4,000 kms between 22”S-56”s. 3-6 eggs(Yom-Tov 1987). Hence, phylogenetic We have divided it into five main zoogeograph- differencesmight also be important in determin- ical regions, namely the Amazonian, Chaco, An- ing clutch size. dean, Patagonian, and Araucanian (Cabrera and Most studies on variation in clutch size have Willink 1980, Narosky and Yzurieta 1989), which been carried out in the Northern Hemisphere (for occupy approximately 7%, 54%, 14%, 14% and review, see Klomp 1970) and comparisons have 11% of the total area, respectively. The Ama- been carried out between these regions and the zonian region occupiesthe northern half of Rio tropics (Skutch 1985). Little has been done on Grande de Sul and two small areasof north east- this subject in the Southern Hemisphere (but see em and north western Argentina and comprises Moreau 1944 and Yom-Tov 1987) and there is subtropical forest; the Chaco region comprises no study of clutch size in mid-latitudes in South subtropical deciduous xerophytic woodland, America. We studied the clutch size of passerines scruband grasslandscovering central Chile, most breeding in the southern cone of South America, of lowland north and central Argentina, all of to determine whether the clutch size of these Uruguay and the southern half of Rio Grande birds was small, like those in other two southern de Sul; the Andean region covers the Andes of continents. We try to relate our results to both Argentina and Chile from latitude 36”s to the 172 Y. YOM-TOV, M. I. CHRISTIE AND G. J. IGLESIAS north, including the Puna and the lowlands of approximation of both eggvolume and eggmass northern Chile, mostly subtropical to temperate (van Noordwijk et al. 198 1). deserts,xeric scrub and montane grasslands.The TYPE OF NEST Patagonian region comprises the cold scrub and grass steppes of southern Argentina and Chile. All nests were divided into four types: cup nests The Araucanian region occupies the temperate made of vegetational matter; nests in a hole in and cold forests of southern Chile and south- the ground or a tree; enclosednests made of veg- western Argentina. The Andean and Patagonian etational matter; or enclosed mud nests, typical regions could be considered as sub-units of a to the genus Furnarius and some Hirundinidae. larger Andino-Patagonian region. ANNUAL MOVEMENTS MATERIALS AND METHODS Although most birds of the study area are year- We studied several factors related to clutch size round residents, some speciesmigrate or move of passerinesby comparing clutch sizes in rela- from one region to another during the year.