Citation
Reproductive strategy of a generalist brood parasite, the shiny cowbird, in the Cauca Valley, Colombia

Material Information

Title:
Reproductive strategy of a generalist brood parasite, the shiny cowbird, in the Cauca Valley, Colombia
Creator:
Kattan, Gustavo H., 1953-
Publication Date:
Language:
English
Physical Description:
vii, 89 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Animal nesting ( jstor )
Bird nesting ( jstor )
Birds ( jstor )
Breeding seasons ( jstor )
Eggs ( jstor )
Female animals ( jstor )
Incubation ( jstor )
Parasite hosts ( jstor )
Parasites ( jstor )
Parasitism ( jstor )
Dissertations, Academic -- Zoology -- UF
Zoology thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 81-88).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Gustavo H. Kattan.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030082269 ( ALEPH )
29996188 ( OCLC )

Downloads

This item has the following downloads:


Full Text








REPRODUCTIVE STRATEGY OF A GENERALIST BROOD PARASITE,
THE SHINY COWBIRD, IN THE CAUCA VALLEY, COLOMBIA









By

GUSTAVO H. KATTAN














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1993













ACKNOWLEDGMENTS

The culmination of this dissertation has involved the contribution of many people and institutions. First I would like to thank the members of my doctoral committee, Drs. Louis Guillette (chairman), Douglas Levey, Richard Kiltie, Scott Robinson, John Fitzpatrick, and formerly Martha Crump, for their support and advice during all phases of the research. My friend and former advisor, Dr. Humberto Alvarez, put me in the tracks of tropical birds, and generously shared with me his experience with House Wrens and Shiny Cowbirds. I enjoyed the many hours we spent discussing cowbirds over countless cups of coffee. I owe a great debt to my field assistant, Natalia G6mez. Her dedication and skills were crucial for the success of the field work. She always did more than I asked and contributed many good ideas. Discussions with Drs. Stephen Rothstein, Spencer Sealy and Alfred Dufty have contributed substantially to my understanding of brood parasitism in birds.
Financial support for this research was provided by the Department of Zoology, University of Florida, the Frank Chapman Memorial Fund of the American Museum of Natural History, the Fundaci6n para la Promoci6n de la Investigaci6n y la Tecnologfa, Banco de la Reptiblica, Bogotd, Colombia, the Fondo de Investigaciones Cientfficas "Francisco Jos6 de Caldas" (COLCIENCIAS), BogotA, Colombia, and the Instituto Vallecaucano de Investigaciones Cientificas (INCIVA), Cali, Colombia. I would like to thank the staff of INCIVA, especially Mrs. Belly Narviez, for logistical support, and


ii








don Alfonso Madrifin for permission to work on his property. Jos6 Kattan and Walter Buttkus helped me build more than 140 nest boxes, with wood donated by don Hernando Murcia. Manuel Giraldo provided company and laboratory and field assistance and helped with logistical support at the Universidad del Valle.
Last but not least, I would like to express my gratitude to my wife,
Carolina Murcia, for being there and for putting up with my constant cowbird talk during the last five years!
































nii












TABLE OF CONTENTS



ACKNOWLEDGMENTS ................................... ii

ABSTRACT ............................................... vi

CHAPTERS

1 INTRODUCTION ..................................... 1

2 STUDY AREA AND GENERAL METHODS ............... 6

3 REPRODUCTIVE SEASONALITY AND FECUNDITY
OF SHINY COWBIRDS IN THE CAUCA VALLEY ......... 8

Introduction ..................................... 8
Methods........................................ 10
Results ............................................ 13
Discussion ........................................ 18

4 HOST NEST SELECTION ............................ . 22

Introduction..................................... 22
Methods........................................ 23
Results .......................................... 24
Discussion ....................................... 31

5 MECHANISMS OF SHORT INCUBATION PERIOD
IN THE SHINY COWBIRD ............................ 35

Introduction ..................................... 35
Methods....................................... 40
Results ......................................... 42
Discussion ..................................... 47

6 IMPACT OF COWBIRD PARASITISM: WHY DO
HOUSE WRENS ACCEPT PARASITIC EGGS? ........... 55

Introduction .................................... 55
M ethods........................................ 57
Results ......................................... 59
Discussion...................................... 67


iv








7 GENERAL DISCUSSION ............................. 73

LITERATURE CITED ................................... 81

BIOGRAPHICAL SKETCH .............................. 89











































V











Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

REPRODUCTIVE STRATEGY OF A GENERALIST BROOD PARASITE,
THE SHINY COWBIRD, IN THE CAUCA VALLEY, COLOMBIA

By

Gustavo H. Kattan

May 1993
Chairman: Dr. Louis J. Guillette
Major Department: Zoology
The reproductive strategy of a brood parasite involves "decisions" on life history traits such as how many eggs to produce and how much to invest in each egg. Parasites may produce a large number of eggs and parasitize as many nests as they can, with little consideration of the circumstances in which each egg is laid (the "shotgun" strategy). Host responses, however, may modulate the parasite's strategy. If hosts reject eggs that are non-mimetic and/or not synchronized with their own laying period, parasites may be forced to spend energy in selecting suitable hosts and closely monitor nests to select the most appropriate time for egg laying. These conditions would result in parasites being able to lay fewer eggs. I studied these aspects of the interaction between Shiny Cowbirds (Molothrus bonariensis) and House Wrens (Troglodyies aedon), their main host in the Cauca Valley, Colombia.
Cowbirds in the Cauca Valley are extremely fecund birds. A conservative estimate gave a fecundity of 120 eggs per year, equivalent to about 9.2 times the female's body mass. This extraordinary fecundity results from cowbirds



vi








laying almost continuously, without regressing the ovary, during the entire 9month breeding season.
As predicted by life history theory, the high fecundity exhibited by
cowbirds involves a trade-off with the energy invested in each egg. I divided investment in two components: A direct investment represented in the amount of energy deposited in the egg and a behavioral investment represented in the process of nest selection. Cowbirds lay eggs with an energy content 24% lower than would be expected from egg mass. Presumably, the energy saved allows cowbirds to lay more eggs than would be possible if each had a higher energy content. A reduced energy content also results in a short incubation period, which gives cowbirds an advantage over their nestmates.
Cowbirds also invest little energy in the careful placing of each egg in circumstances that would maximize its chances of being accepted and incubated successfully by the host. Cowbirds monitor host nests and to some extent choose nests that are at an appropriate stage for parasitism. However, they do not synchronize egg laying with the wren's laying period, and do not avoid multiply parasitized nests. This strategy is possible because wrens do not reject cowbird eggs. Experiments with artificial eggs indicated that wrens do not recognize cowbird eggs as foreign, neither do they reject unsynchronized eggs. Therefore, cowbirds in the Cauca Valley follow the "shotgun" strategy. Instead of carefully placing a few eggs, they direct their energy into egg production and permanently search for nests, parasitizing as many as they can. I discuss the implications of this strategy for the evolution of host specificity in brood parasites.






vii












CHAPTER 1
INTRODUCTION

Avian brood parasites lay their eggs in the nests of other bird species; the host species incubates the eggs and raises the young. Obligate interspecific brood parasitism occurs in about 1% of the world's bird species (about 85 species), including 47 species of cuckoos (Cuculidae) in the Old World (Cuculinae) and three species in America (Neomorphinae), the honeyguides (Indicatoridae) of Africa and Asia, the widowbirds (Viduinae) and the Cuckoo-weaver (Anomalospiza imberbis) of Africa, and five species of cowbirds (Icterinae, Emberizidae) in America.
The unorthodox breeding habit of brood parasites has intrigued
naturalists since ancient times. The parasitic habit of the Common Cuckoo (Cuculus canorus), for example, is known since Aristotle's time (Wyllie 1981) and has been featured extensively in natural history works since then (e. g., Goldsmith 1774). In the New World, the breeding habits of Brown-headed Cowbirds (Molothrus atgr) in North America and Shiny and Screaming Cowbirds (M. bonariensis and M. rufoaxilaris) in South America were described in several XIX century natural history works (Hudson 1920; Peattie 1939).
From a more modern perspective, work on brood parasites was pioneered by H. Friedmann, who wrote extensively on cowbirds, cuckoos, weaverbirds and honeyguides (Friedmann 1929, 1948, 1955, 1960). Since then, the literature on brood parasitism has exploded (see


1





2


reviews in Hamilton and Orians 1965; Payne 1977; Rothstein et al. 1986; Rothstein 1990). Since Hamilton and Orians (1965) placed brood parasitism within a natural selection paradigm, knowledge of the most significant evolutionary questions in this field has progressed enormously. In particular, coevolutionary aspects of the interaction between parasite and host have received much attention (Rothstein 1990). Because parasitism usually has a negative impact on the host, hosts evolve defenses. Parasites, in turn, evolve counterdefenses, resulting in an evolutionary "arms race" (Davies and Brooke 1989b; Rothstein 1990). The interaction between the two species results in the adoption of strategies of host selection by parasites. Some parasites, such as Brown-headed and Shiny Cowbirds, are extreme generalists, parasitizing most passerines with which they occur (Friedmann and Kiff 1985; Rothstein 1990). Others, such as cuckoos, specialize on only one or a few host species (although for species with wide geographic distributions the total host list may be long; Wyllie 1981; Rothstein 1990).
Because brood parasites do not provide parental care, they can be viewed as natural experiments on the effects of parental care on life history traits (Payne 1977). Life history theory predicts a trade-off between the number of offspring produced and the energy invested in each offspring (Roff 1992). Organisms that produce large numbers of offspring invest little in each offspring, and vice-versa. Therefore, brood parasites are expected to exhibit higher fecundity than their nesting relatives. Payne (1977) reviewed reproductive strategies of brood parasites and found that they lay more eggs than non-parasitic species. Payne (1974, 1977) suggested that the small egg size of parasitic cuckoos may be due to allocation of energy to a large number





3


of eggs. In the 16 years since Payne's (1977) review, however, no studies have followed up on these ideas. In particular, the trade-off between fecundity and investment has not been investigated further. Recent estimates of Brownheaded Cowbird fecundity indicate that they lay large numbers of eggs (Scott and Ankney 1983; Rothstein et al. 1986), but egg size is not smaller than the expected from the adult female body size (Briskie and Sealy 1990), as is the case in cuckoos. Is the increased fecundity of cowbirds simply the result of a reallocation of the energy not spent in parental care, or is the allocation of energy to each egg in any way compromised in cowbirds? In this study I test the hypothesis that Shiny Cowbirds direct their reproductive effort into egg production at the cost of reducing the energy invested in each egg.
I studied Shiny Cowbirds in the Cauca Valley, in southwestern Colombia. This cowbird is widely distributed in South America (Friedmann and Kiff 1985) and has recently invaded the Antilles and the southeastern United States (Cruz et al. 1985; Post et al. 1993). It is an extreme generalist, reportedly parasitizing more than 200 species of birds throughout its range (Friedmann and Kiff 1985). Work on Shiny Cowbirds has concentrated mostly on identifying local host assemblages (Ramo and Busto 1981; Salvador 1983; Wiley 1985; Mason 1986a, 1986b; Cavalcanti and Pimentel 1988). Detailed studies of host-parasite interactions are limited to a few species (e. g., Gochfeld 1979b; King 1973; Post and Wiley 1977; Fraga 1983, 1985).
I begin by examining the seasonality of reproduction and estimating annual fecundity in this tropical population of cowbirds (Chapter 3). The hypothesis is that brood parasitism involves a trade-off: emancipation from parental care allows birds to devote extra energy to increase their fecundity (Payne 1977; Roff 1992). Work on Brown-headed Cowbirds in North America has shown that females have a very high fecundity, producing eggs almost





4


continuously during the breeding season (Scott and Ankney 1983; Rothstein et al. 1986). Because the Shiny Cowbird's main host in the Cauca Valley, the House Wren (Troglodyies aedon), has an extended 9-month breeding season (Alvarez et al. 1984), it raises the question, what is the reproductive seasonality and fecundity of cowbirds, both at the population and individual levels? Once I determine how many eggs cowbirds lay, I examine the question of how much cowbirds invest behaviorally and energetically in each egg. Behavioral investment in each egg involves selection of host nests in order to maximize the chances of the parasitic egg being accepted and incubated successfully by the host. In Chapter 4 1 explore aspects of nest selection and test the hypothesis that cowbirds follow host activities in order to time egg laying with the host's laying period.
Next (Chapter 5) I examine the energetic investment in each egg and its implications as an adaptation for a short incubation period. Once the parasitic egg is laid, it is abandoned by the cowbird and nestlings have to compete with foster nestmates for parental resources. Thus, early hatching gives parasitic young a head start. It has long been recognized that brood parasites have short incubation periods (Hamilton and Orians 1965; Payne 1977), but precise measurements of incubation periods are lacking because of the inherent variability associated with intermitent incubation in the field (Briskie and Sealy 1990). In Chapter 5 I present measurements of incubation period of Shiny Cowbird eggs under constant conditions in the laboratory, showing that the incubation period is shorter than expected from egg size. I suggest that a short incubation period is the result of a reduced egg energy content and discuss its implications for the cowbird's reproductive strategy.
Next I examine impact of parasitism and explore the question of why wrens accept cowbird eggs (Chapter 6). My rationale is that the impact of





5


parasitism may determine host responses, which in turn may modulate the cowbird strategy. If hosts are selective, i. e., they reject non-mimetic and nonsynchronized eggs, they may "force" the parasite to specialize. Unselective hosts, in contrast, allow parasites to be generalists. I end (Chapter 7) by discussing how all these pieces can be put together to show that the Shiny Cowbird strategy is that of an extreme generalist, and I discuss the evolution of host specificity in brood parasites.












CHAPTER 2
STUDY AREA AND GENERAL METHODS


The study was conducted between January - June 1988 and March 1989 April 1991 at a 116 ha dairy farm located 15 km south of the city of Cali, in the Cauca Valley (1000 m elevation), in southwestern Colombia. The study area consisted of well-shaded pastures and farm buildings and was surrounded by fields of rice and sorghum, two crops on which cowbirds feed (Finke et al. 1979). There were a few large trees of ceiba (CeCa pentandra) and samdn (Pithecellobium saman), but most of the shade trees in pastures and along fence rows were chiminango (Pithecellobium dulce), gusimo (Guazuma ulmifolia) and tachuelo (Fagara rhoifolia). Rainfall in the area is bimodal, with peaks of precipitation in April and October, and two dry periods, a mild one in December - January, and a harsher one in July August.
At the beginning of the study I mapped 14 wren territories, which were located mostly along fence rows and around shade trees and farm buildings. Wrens nested in a variety of natural and semi-natural cavities such as holes in bamboo (Guadua angustifolia) fence posts and other man-made structures, bromeliads and on the ground in dense clumps of tall grass. To facilitate data collection I placed wood nest boxes (10 x 10 x 15 cm) in all wren territories. During the first eight months of the study I placed a total of 140 boxes, with at least 6 boxes in each mapped territory. Wrens readily used these boxes and this study is based on 34 natural nests and 105 nests in boxes. Most breeding



6





7


wrens were trapped in nest boxes or captured with mist nets and banded. I also banded 37 Shiny Cowbirds that were captured with mist nests.
Wrens were the only host of cowbirds in the study area. All cowbird
fledglings produced during the study were raised by wrens. No other species was parasitized, although nests of some potential hosts, such as Pied WaterTyrants (Fluvicola Vica; Ramo and Busto 1981), were available. Other hosts used by cowbirds elsewhere in the Cauca Valley include Black-billed Thrushes (Turdus ignobilis), Blue-gray Tanagers (Thraupis episconus) and Crimson-backed Tanagers (Ramphocelus dimidiatus). In the mountains surrounding the valley (1200-2000 m), the main host is the Rufous-collared Sparrow (Zonotrichia capensis). Wrens and cowbirds have coexisted in the valley for at least 100 years and probably for several centuries. The earliest ornithological record of both species is from 1911 (Chapman 1917), but the Cauca Valley has been settled since pre-Columbian time. In the 18th century there was already extensive agriculture and cattle ranching by European settlers. These activities probably provided adequate habitat for both wrens and cowbirds.
The general methodology of the study consisted in keeping track of the breeding activities of all wrens. I made observations during all stages of the nesting cycle, from nest building to fledging. Nests were also observed during nest building and egg laying to record cowbird activities. Methods specific to each section of the study are reported in the corresponding chapter.











CHAPTER 3
REPRODUCTIVE SEASONALITY AND FECUNDITY OF SHINY
COWBIRDS IN THE CAUCA VALLEY


Introduction


Life history theory predicts a trade-off between number and quality of offspring (Roff 1992). Organisms that produce large numbers of offspring invest little in each offspring, and vice-versa. Parental care is an investment in the quality of offspring, increasing their chances of survival. Because brood parasites do not invest in parental care, they can be viewed as natural experiments on the trade-off between parental care and fecundity (Payne 1977; Roff 1992). Brood parasites, however, do not abandon their eggs at random. Their offspring require parental care, and the parasitic bird must invest some energy in laying the egg at a place and time that maximize chances of survival (Chapter 4). Emancipation from parental care, nevertheless, makes extra energy available and thus brood parasites are expected to exhibit higher fecundity than other species. Indeed, estimates of Brown-headed Cowbird annual fecundity in the north temperate zone have shown that they are extraordinarily fecund, laying in a pattern that resembles that of domestic chickens (Scott and Ankney 1983). In this chapter I present estimates of annual fecundity (number of eggs produced per year) at a tropical locality in which cowbirds have an extended breeding season.
The first estimates of the number of eggs laid by cowbirds were based on searching for parasitized nests in the field (reviewed in Payne 1977; Rothstein et al. 1986). These estimates relied on the assumption that females


8





9


lay individually distinctive eggs. While there is evidence that individual females always lay similar-looking eggs (Dufty 1983; Fleischer 1985), this is an unreliable method to estimate fecundity for several reasons (Rothstein et al. 1986). First, some host nests may not be found, or cowbird eggs may be removed by hosts or predators before being found. Second, laying ranges of individual females may not be known precisely. Third, laying ranges of several females may overlap and egg appearance may not be distinctive enough to differentiate females.
More recent estimates of Brown-headed Cowbird fecundity have relied on the examination of ovaries. Payne (1965, 1973) examined postovulatory follicles and found that cowbirds lay eggs in sets or clutches, separated by days of non-laying. There was great interindividual variation, with sets varying between 1-6 eggs. Nearly all females laid throughout the breeding season, with the total number of eggs produced by a female depending on the length of the season. Payne (1976) obtained estimates varying from 11 to 25 eggs in different localities in the U. S. Extensive studies by Scott and Ankney (Scott 1978; Scott and Ankney 1980, 1983) obtained a higher estimate of 40 eggs per season in southern Ontario. They confirmed Payne's finding that cowbirds lay in sets of 1-7 eggs, with no regression of the ovary and oviduct until the end of the season. They also showed that individual cowbirds lay on about 70-80% of the days of the breeding season.
Most estimates of cowbird fecundity have been done on Brown-headed Cowbirds in the north temperate zone, where the breeding season is restricted to a few months of the year. Most detailed studies of Shiny Cowbirds have been conducted in the south temperate zone, where there is also marked seasonality (e. g., Fraga 1985; Mason 1985, 1986a). The only estimate of Shiny Cowbird fecundity based on examination of ovaries (Davis





10


1942) is based on a small sample and gave a small estimate. While breeding of cowbirds and other birds in the tropics may be seasonal and related to alternation of wet-dry seasons (e. g., Wiley 1988), some tropical birds may have extended breeding seasons. House Wrens, for example, breed almost year-round in the Cauca Valley (Alvarez et al. 1984). Because wrens are frequently parasitized by cowbirds (Alvarez et al. 1984; pers. obs.), it raises the question, what is the reproductive seasonality and fecundity of Shiny Cowbirds at the population and individual levels? Here I present data on the reproductive seasonality of a population of Shiny Cowbirds in the Cauca Valley and, based on examination of ovaries, estimate seasonality and fecundity of individual females.


Methods


Reproductive seasonality of the House Wren and Shiny Cowbird
populations was determined during January - June 1988 and March 1989 December 1990. Nests were checked at intervals of 1-3 days and eggs were individually marked with permanent ink. For each month I compiled the proportion of parasitized nests and the total number of cowbird eggs laid.
Cowbird fecundity was determined from inspection of ovaries. Monthly samples of females were collected at a rice mill about 5 km SW of the main study area, during the months of November 1990-April 1991. No females were collected at the main study area. Cowbirds exhibited a diurnal cycle of social behavior in which females were solitary in the morning, when they were searching for host nests, and gregarious in the afternoon, when they joined feeding flocks (see also Rothstein et al. 1986). Very few females were observed at the mill in the morning. Flocks started arriving between 11:30








and 12:30 and were composed of both juveniles and adults. I captured cowbirds with mist nets when they entered the mill to feed on rice. This method ensured no bias in capturing females in different reproductive conditions (Scott and Ankney 1979).
Only females in adult plumage were collected. Females were dissected and the fresh ovaries examined under a dissecting scope. In reproductively active females there was usually a series of yellow vitellogenic follicles (i. e., follicles in which yolk is being deposited) of graded sizes and numerous small white follicles. Postovulatory follicles (i. e., the scars left in the ovary after ovulation) were also found in graded series, as the follicle is gradually reabsorbed. Diameters of all vitellogenic follicles and maximum width of postovulatory follicles were measured with calipers. Only yellow follicles were counted as vitellogenic. Follicle sizes were then plotted as a function of size rank for each female, number 1 being the largest irrespective of absolute size (Fig. 3-1). Assuming that Shiny Cowbirds ovulate daily, as found for Brownheaded Cowbirds (Scott and Ankney 1983; Jackson and Roby 1992), the laying history of each cowbird a few days before and after collection can be estimated from follicle sizes. Vitellogenic follicle sizes produced a graded series, with D follicles (size range 9.1-12.0 mm) being ready to ovulate (Fig. 31). C follicles (size range 5.5-9.0 mm) would be ovulated the next day, and so on (assuming no atresia). Most follicles smaller than 2 mm were white and it was not possible to determine if they were initiating vitellogenesis. Thus, this method allowed determination of future ovulations for only four days after collection. A similar rationale was used for postovulatory follicles. Size E postovulatory follicles were found only when there was an egg in the oviduct. Postovulatory follicles regressed rapidly and could be identified with confidence only up to three days after ovulation (Fig. 3-1).





12











OVULATION

12.0

10.0 D

8.0
C
6.0

S4.0
4 +

2.0 Al

0.0

V4 V3 V2 V1 P1 P2 P3 VITELLOGENIC POSTOVULATORY FOLLICLES FOLLICLES

Figure 3-1. Sequence of growth of vitellogenic follicles and regression of postovulatory follicles in cowbird ovaries. V1 is the largest vitellogenic follicle, V2 the next largest, and so on. The same applies to postovulatory follicles. A follicle in the D size range is ready to be ovulated, with follicles in the C, B and A size ranges to be ovulated on consecutive days. A size E postovulatory follicle corresponds to a recently ovulated egg, and F and G indicate follicles regressed for one and two days, respectively.





13


By analyzing the sequence of follicle sizes, I estimated set size and
interval between sets of eggs for each female. The non-laying interval was estimated by gaps in the sequence of follicles. For example, a female with an F postovulatory follicle, a C vitellogenic follicle and no egg in the oviduct, has a gap of two days. Set size was estimated by counting the longest possible sequence of consecutive daily ovulations. Because of the limitations in identifying follicles macroscopically, this method provides only a minimum estimate of set size.


Results


Wrens bred in all months of the year except during the dry period of July August. There were peaks of nesting activity that coincided with the rainy seasons of April-May and October-November, with the first peak higher than the second (Fig. 3-2). Nesting activity was low during the JulySeptember dry season, with a few nests being initiated in early July and late September. Cowbird breeding activity paralleled wren nesting. Nests were parasitized all months in which there were nests available (Fig. 3-2). These data indicate that the cowbird population breeds continuously during the period October-June, with very little or no reproductive activity during the July-September dry season.
I examined 95 ovaries to determine reproductive status and estimate
fecundity. Reproductively active females always had vitellogenic follicles and large oviducts (length=120-150 mm). Non-reproductive females had regressed ovaries and oviducts (length<50 mm). Assuming that cowbirds lay an egg every day, the pattern of graded follicle sizes (Fig. 3-1) indicated that Shiny Cowbirds laid eggs in sets. Set size varied between 1-6 eggs with a





14







-- Wren nests 11
-- Parasitized 10

9

8

7

0
5


4
3

2

1

0- FMAMJ//MAMJ ONDJ MAM SOND
1988 1989 1990


Figure 3-2. Reproductive seasonality of the House Wren and Shiny Cowbird populations in the Cauca Valley. Open symbols indicate number of wren nests with eggs and closed symbols indicate number of parasitized nests.





15


mean of 3.2 (SD=1.4, n=35). This is a minimum estimate because my method does not allow detection of sequences longer than 7 days, although the frequency distribution of sets suggests that this is a precise estimate, as there were few ovaries with sets of 5-6 eggs (Fig. 3-3). Gaps in the size sequence indicated that non-laying intervals varied between 1-5 days, with a mean of
1.64 (SD=0.9, n=28). Modal gap size was one (Fig. 3-3).
From November 1990 to April 1991, between 55% and 90% of the females collected each month were reproductively active (Fig. 3-4). All nonreproductive females were molting, suggesting that molting and breeding are mutually exclusive activities. During August females were neither molting nor breeding, suggesting that cowbirds interrupt all energetically demanding activities during the dry season (June - August). Reproductive females that were in the non-laying interval when collected (i. e., had just finished ovulating a series and follicles were developing for the next series) had fully developed oviducts. This pattern suggests that individual females are inactive during the 3-month dry season, but lay eggs continuously during the 9-month breeding season, interrupting only for molting. Duration of molting in Shiny Cowbirds is unknown. However, for a variety of tropical passerine birds with non-overlapping breeding and molting seasons, duration of the molting season for the population varies between 2-5 months and for individual birds may be 2-3 months (Miller 1961; Worthington 1982; Poulin et al. 1992; Levey and Stiles, in press). Thus, assuming that a cowbird molts once during the 9-month season, and using a conservative estimate of three months for duration of molting, a female cowbird in the Cauca Valley lays eggs during six months of the year. If each female has an average daily laying rate of 0.66 eggs (series of 3.2 eggs separated by 1.64 days), her annual fecundity is approximately 120 eggs.





16






20
18
16
14
0 12 0 10
C4 8
6 4
2
01"M
1 2 3 4 5 GAP SIZE



12

10

8

06

4

2

0
1 2 3 4 5 6 SET SIZE

Figure 3-3. Frequency distribution of set size (number of consecutive ovulations) and gap size (non-laying intervals) in cowbird ovaries.





17










100 10
m 90 11
80 18
70 8
60 18 15
50
40
30
20 10
5
0
AUG // NOV DEC JAN FEB MAR APR




Figure 3-4. Percent of reproductively active female cowbirds in each monthly sample. Numbers above bars are sample sizes. All non-reproductive females were molting in all months except August, when cowbirds were neither breeding nor molting.





18


Discussion


The results indicate that Shiny Cowbirds in the Cauca Valley are
extraordinarily fecund, laying approximately 120 eggs per year. This high fecundity is the result of two factors. First, cowbirds lay almost continuously during the breeding season. Second, the breeding season at this locality is exceedingly long.
Cowbird breeding seasons always coincide with their hosts'. In the north temperate zone, breeding seasons of all birds are determined by strong seasonality, with northern localities having shorter seasons than more southern localities. In Michigan, for example, the breeding season of the Brown-headed Cowbird and its hosts lasts 5-6 weeks, whereas in Oklahoma it lasts 8-10 weeks (Payne 1976). In Ontario, the breeding season of Brownheaded Cowbirds is restricted to about 8 weeks in May-June (Scott and Ankney 1980). Breeding seasons are similarly restricted to a few months of the year in the southern temperate zone. In northern Argentina most passerines, including Shiny Cowbirds, breed during the months of OctoberFebruary, corresponding to the austral spring and summer (Fraga 1985; Mason 1985, 1986a).
The environmental signal that mediates gonadal development in the
temperate zone is photoperiod (Payne 1967, 1977). Environmental signals in the tropics are less obvious. Breeding seasonality in tropical birds is usually related to alternation of dry and wet seasons, and in most localities there tend to be more or less well defined breeding seasons. In Puerto Rico, for example, most passerines breed during the rainy season of March-June; although some cowbird hosts breed year-round, the laying period of the Shiny Cowbird is restricted to the months of March-July (Wiley 1988). In the Cauca





19


Valley, there is a well-defined breeding season that coincides with the MarchJune rainy season (N. G6mez and G. Kattan, unpublished data). Some species, however, have breeding seasons that may extend up to 9 months. House Wrens in the Valley breed almost year-round, interrupting reproductive activities only during the driest period of July-September (Alvarez et al. 1984; this study). Here the Shiny Cowbird population parallels its host's extended breeding season.
As a consequence of the extended breeding season Shiny Cowbirds lay an enormous number of eggs annualy. The extremely high fecundity of cowbirds in the Cauca Valley may be the result of a superabundance of food, as they feed on commercial crops such as rice and sorghum (Finke et al. 1979). My minimum estimate of a daily laying rate of 0.66 eggs is slightly lower than the 0.72-0.80 value estimated for Brown-headed Cowbirds in North America (Scott and Ankney 1979; Rothstein et al. 1986). Most adult females probably lay throughout the season in the temperate zone. Captive Brown-headed Cowbirds had laying rates between 0.18 and 0.91 eggs/day, with some females laying almost continuously during a 68-day period (Jackson and Roby 1992). While it is difficult to know the laying schedule of an individual female without actually following her throughout the breeding season, my data suggest that individual females may be reproductively active during at least two thirds of the 9-month breeding season.
As has been found for Brown-headed Cowbirds in previous studies,
examination of Shiny Cowbird ovaries indicated that eggs are laid in clutches or sets, separated by non-laying days. The significance of these sets is not clear. Laying patterns of captive yearling Brown-headed Cowbirds are highly variable, both within and between individuals, and do not support the notion that cowbirds lay in clutches (Jackson and Roby 1992). It has been suggested





20


that non-laying intervals may result from follicular atresia related to physiological constraints and/or availability of host nests (Payne 1965, 1976; Scott and Ankney 1983; Rothstein et al. 1986). The egg laying pattern of cowbirds, however, resembles that of domestic chickens that are not allowed to brood (Scott and Ankney 1983). Domestic chickens lay eggs in sequences of daily ovulations separated by non-laying days (Phillips et al. 1985). This pattern of laying is the result of the existence of an "open period" in the ovulatory cycle. The open period is a restricted period of the day during which pre-ovulatory releases of luteinizing hormone (LH) occur. The LH surge (and the subsequent ovulation) first occurs at the beginning of the open period. Subsequent LH surges occur a little later each day, until they occur at the end of the open period, ending the sequence of ovulations (Phillips et al. 1985). A similar mechanism could be operating in cowbirds. This does not rule out the possibility, however, that there are waves of follicular recruitment, but the actual number of eggs laid is determined by atresia, as shown for the lizard Sceloporus mucronatus (Mendez de la Cruz et al., in press).
The high laying rate of cowbirds raises questions related to nutritional
constraints. Shiny Cowbirds in the Cauca Valley lay eggs with a mean mass of 4.2 g and an energy content of 14.54 kJ (Chapter 6). Thus, a 55 g female cowbird may be processing a total annual egg mass about 9.2 times her own body mass (504 g), with a total energy content of 1745 kJ. Ankney and Scott (1980) assayed Brown-headed Cowbirds for physiological strain related to the high laying rate, and found no indication of major depletions attributable to egg laying. Rothstein et al. (1986), however, argue that the costs of laying may not be as small as Ankney and Scott (1980) suggested. Besides, certain nutrients such as calcium and iron may impose limitations on the capacity to





21


produce eggs. Rothstein et al. (1986) found that laying females had decreased hematocrit values, suggesting that iron may be a limiting nutrient. The fact that I found no overlap between molting and reproduction in Shiny Cowbirds suggests that these two activities are energetically incompatible (Payne 1972). The question of physiological strain in laying female cowbirds is still unresolved and requires further investigation.
Even with the imprecision of methods available to estimate cowbird fecundity, it is clear that cowbirds are extremely fecund birds. This has important implications for the trade-off between parental care and fecundity. It is tempting to argue that the energy spent in excess of what a nesting bird spends in egg production is the cost of parental care. This is unlikely to be the case, as there are costs associated with nesting, such as predation risks and time constraints, that are difficult to express in energy currency and that limit reproductive effort. The data presented here clearly demonstrate, however, that cowbirds are directing their reproductive effort into production (Roff 1992).











CHAPTER 4
HOST NEST SELECTION


Introduction


Egg laying by a cowbird in a host nest is the culmination of a process of
nest selection (Thompson and Gottfried 1981). The first step in the process is finding an active nest. Once nests are located, cowbirds must select the nests in which they will lay eggs. Nest selection may be important for two reasons. First, cowbirds must lay their eggs at a time that maximizes both probability of acceptance by the host and chances of survival for the parasitic egg and young. Some hosts reject cowbird eggs that are laid before the initiation of their own laying period (Friedmann and Kiff 1985; Wiley 1985). For example, Yellow Warblers (Dendroica petechia) usually reject cowbird eggs laid before their laying period by nest desertion or burying the egg in the nest lining (Clark and Robertson 1981; Weatherhead 1989). On the other hand, parasites laying in nests where incubation is already underway have a low chance of success because the parasitic egg may not hatch, or may hatch too late to compete successfully with its nestmates (e. g., Fraga 1985; Weatherhead 1989).
Second, nest choice by cowbirds may be important in avoiding competition with other cowbirds. Brood parasites are probably limited by the availability of host nests (Payne 1977) and multiple parasitism (more than one egg in a host nest) is likely to result in low cowbird success (Weatherhead 1989). Thus, cowbirds are expected to exhibit mechanisms to avoid competition,


22





23


such as 1) defending nests or laying areas from other females, 2) destroying or removing cowbird eggs from nests, or 3) avoiding laying in already parasitized nests (Post and Wiley 1977; Darley 1983; Fraga 1985; Teather and Robertson 1985; Orians et al. 1989).
Here I test the hypothesis that Shiny Cowbirds invest some effort to
ensure placing each egg in the most advantageous circumstances possible (i. e., they select a nest). An alternative hypothesis is that cowbirds use a "shotgun" strategy (Rothstein 1975). Instead of timing egg laying with a few host nests, cowbirds search for nests in a random manner and parasitize as many as they can, whatever their stage of development. I tested the following predictions:
Prediction 1: Cowbirds should monitor nest progress in order to determine when it will be in an appropriate stage for parasitism. Prediction 2: Cowbirds should time their egg laying with the host's laying period.
Prediction 3: Multiple parasitism should be rare. There should be few multiply parasitized nests.
Prediction 4: Distribution of cowbird eggs in wren nests should be nonrandom, with less multiply-parasitized nests than expected from a Poisson distribution (Orians et al. 1989); cowbirds should remove or destroy previously laid cowbird eggs.


Methods


To test the prediction that cowbirds monitor nest progress, I made 1-h observation sessions of wren nests during three stages: a) Early nest construction, when wrens are adding coarse material; b) Late nest





24


construction, when wrens are finishing a cup of fine material and adding lining; and c) Laying period. During these sessions I recorded number of cowbird visits and cowbird behavior. If cowbirds approached the nest rapidly and directly, I assumed they knew the location of the nest beforehand (Wiley and Wiley 1980). I assumed these rapid visits were visits to check nest progress, as opposed of visits of cowbirds searching for a nest for the first time. To determine synchronization of cowbird egg laying with the host's laying period, I checked all active nests daily during the building and laying stages. Cowbird eggs were marked with permanent ink for identification and dates of laying noted. All nests were checked between 06:30 and 09:00 each morning and observation sessions made between 08:00 and 10:00.
The prediction that multiple parasitism should be rare was tested by comparing the distribution of cowbird eggs in wren nests with a Poisson distribution (Orians et al. 1989). Agreement with the Poisson distribution would indicate that cowbird eggs were randomly distributed in wren nests, while departure from the Poisson would indicate a clumped or regular dispersion of eggs.


Results


Cowbirds made inspection visits to wren nests during all stages of the nesting cycle. During inspection visits cowbirds flew directly to the nest, entered the cavity for less than 15 sec, and left quickly. This direct and rapid approach indicates that cowbirds knew the location of the nest beforehand (Wiley and Wiley 1980). Visitation rates tended to be higher during early stages of nest building, but the difference between stages was non-significant (ANOVA, F=2.84, df=2, 29, P=.075; Fig. 4-1). Inspection visits were mostly by





25









16-26
4


3
10-21

11
26-12





0
a b c NEST STAGE



Figure 4-1. Frequencies of female cowbird inspection visits to wren nests during different stages of nest building. a) Building of nest foundation; b) Building of nest cup and lining; c) Laying period. Error bars indicate one standard deviation. Numbers above bars indicate number of nests and total number of hours of observation (the sample unit is the nest).





26


solitary females, rarely by groups of 2-4 females (Table 4-1). Males sometimes escorted females (Table 4-1), but they did not get close to the nest.
Inspection visits indicated that cowbirds were monitoring nest progress. Patterns of inspection visits (Fig. 4-1) and multiple parasitism indicated that several female cowbirds monitored each nest. Assuming that each female always lays similar-looking eggs, as found for Brown-headed Cowbirds (Dufty 1983), egg variability indicated that at least six different females laid in some nests. Besides, sometimes up to six eggs were laid in a nest in a single day. Resightings of banded females indicated that each female monitored several nests simultaneously over the course of several days. Home ranges of cowbirds were larger than my study area of 14 wren territories, and based on egg variability, at least 15 females were laying in the area. The actual number of females in the area, however, was much larger because I banded 37 females during the study and very rarely did I resight banded females.
Cowbirds tended to lay their eggs in coincidence with the wrens' laying period (Fig. 4-2). Forty-seven percent of 185 eggs laid in 40 nests were laid during the 3-day laying period of wrens. A large number of cowbird eggs, however, were laid before wrens started laying (35%) or after the wrens completed their clutch and initiated incubation (18%; 3(= 24.7, df=2, P<.001). The above sample includes only nests in which wrens actually laid, and it underestimates the proportion of cowbird eggs laid prematurely (before the wrens' laying period), because cowbirds laid 82 eggs in 20 nests that were abandoned by wrens before laying. Sometimes these eggs were laid too early during nest construction and were buried as wrens continued adding nesting material.





27













Table 4-1. Group size of female cowbirds during inspection visits to wren nests.



Escorted by Males

Group size No. observations Proportion Mean No.
(%) visits males/group

1 39 (76.5) 0.18 1.3 2 6 (11.7) 0.17 1.0 3 3 (5.9) 0.66 1.5 4 3 (5.9) 0.66 2.0





28














m 50

40 30
O
O

0,


-7 -6 -5 -4 -3 -2-1 0 1 2 3 4 5 6 7 DAY OF NESTING CYCLE Figure 4-2. Number of cowbird eggs laid in different days of the nesting cycle. Day 0 = Day in which the first wren egg is laid. Modal clutch size of wrens is 3, and wrens start incubating after laying the last egg (day 2).





29


Wrens abandoned nests that were parasitized with more than two eggs, but accepted cowbird eggs when parasitized with one or two eggs, regardless of when were they laid (see Chapter 6). Contrary to predictions 3 and 4, multiple parasitism was common and cowbirds did not avoid multiply parasitized nests. Multiple parasitism was more common and more intense in boxes than in natural nests (Fig. 4-3), reflecting the conspicuousness of nest boxes. The average number of eggs per nest was higher in nest boxes (1.97) than in natural nests (5.65; Mann-Whitney U, P<.001). In natural nests, the distribution of cowbird eggs was significantly different from a Poisson distribution ('=38.52, df=4, P<.001; Fig. 4-3). There were more unparasitized nests than expected, which reflects the inability of cowbirds to enter cavities with small openings (see Chapter 6). Among parasitized natural nests, there were more nests than expected with 5-8 eggs/nest, and less than expected with 1-4 eggs/nest. This result suggests that some nests were easier to find and that cowbirds did not avoid already parasitized nests. The distribution of cowbird eggs in nest boxes was also significantly different from a Poisson distribution ( =29.5, df=4, P<.001; Fig. 4-3). The large number of multiplyparasitized nests suggests that boxes were easier to find than natural nests. There were more nests than expected with 11-12 eggs/nest, again suggesting that cowbirds did not avoid already parasitized nests.
I found no evidence that cowbirds removed previously laid eggs in
multiply-parasitized nests, but in a few cases they may have destroyed eggs by perforating a hole in the shell. Embryos in perforated eggs usually die of dehydration (Carey 1986). I found 9 out of 150 eggs in multiply-parasitized nests that had a single perforation, as would be produced by a peck. Four of the nine perforated eggs were found in a single nest, together with two perforated wren eggs. It is possible that these perforations were made by





30





20- NATURAL NESTS





o 10*


5 0
0 1 2 3 4 5 6 7 8 9 10 11 12




20 - NEST BOXES





S 10


5 .


01
0 1 2 3 4 5 6 7 8 9 10 11 12
NUMBER OF EGGS PER NEST


Figure 4-3. Distribution of cowbird eggs in wren nests. The bars indicate the observed number of nests and the line represents the Poisson expected distribution for a mean of 1.97 (natural nests) and 5.65 eggs/nest (boxes). Both distributions deviate significantly from the expected.





31


other cowbirds, but they also may have been made by wrens, as they frequently destroyed conspecific and heterospecific neighboring nests (N. G6mez and G. Kattan, unpublished).


Discussion


Once cowbirds find a nest, they visit it frequently and monitor its course. Presumably, monitoring a nest would allow a parasite to determine the most appropriate moment for egg laying. Wiley and Wiley (1980) reported that Shiny Cowbirds spent the day searching for nests in colonies of Yellowhooded Blackbirds (Agelaius icterocephalus); once nests were found, cowbirds inspected them frequently. Shiny Cowbirds in Puerto Rico made inspection visits to nests of three host species during pre-laying and laying stages (Wiley 1988). In contrast to this study, Wiley (1988) found that visitation rates peaked around the day hosts laid their first egg, and he suggested that hosts provide some clue that cowbirds use to synchronize their oviposition. My results indicate that more cowbirds tend to visit the nest early during nest construction, but visitation frequencies are not significantly different over the stages (Fig. 4-1). As construction advances, probably some cowbirds lay and then stop visiting.
The high intensity of parasitism found in this study suggests that cowbirds compete for available nests, and it is intriguing that in this population female cowbirds apparently do not exhibit mechanisms to outcompete other females. A high incidence of perforated cowbird eggs has been reported in other populations of Shiny Cowbirds (Sick 1958; Hoy and Ottow 1964; King 1973; Post and Wiley 1977; Fraga 1985); before laying her egg, the cowbird pecks any other cowbird eggs in the nest. In this study I





32


found a low incidence of perforated cowbird eggs (9 of 150 eggs). My analysis of cowbird egg distribution in wren nests also indicates that female cowbirds do not avoid laying in already parasitized nests (Fig. 2-3). In the few cases in which wrens incubated more than two cowbird eggs, only one egg hatched and the embryo partially developed in one or two additional eggs (Chapter 6). Probably in these cases the female that laid the last, was the one that was successful, because her egg was on top.
My results partially supported the hypothesis that cowbirds time their egg laying to coincide with the wren's laying period. Some Shiny Cowbird hosts are discriminative and reject odd-looking or non-synchronyzed eggs (e. g., Fraga 1985; Wiley 1985; Mason and Rothstein 1986). Other hosts, in contrast, accept almost any egg at any time, as is the case for House Wrens (see Chapter 6). Rufous-collared Sparrows (Zonotrichia capensis) also accept nonsynchronized, odd-looking eggs, and as House Wrens do, only reject by desertion when parasitized with more than two cowbird eggs (Sick 1958; King 1973; Fraga 1978, 1983; pers. obs.).
The importance of synchronization depends not only on the host's rejection response, but in the probability of success of the parasitic young. With large hosts, cowbird reproductive success may be low. For example, Fraga (1985) found that 37% of cowbird nestlings raised with broods of Chalk-browed Mockingbirds (Mimus saturninus) starved. With small hosts, in contrast, cowbirds usually have a high success because of the size advantage. Cowbirds usually hatch first, and nestlings are aggressive and grow fast (see Fig. 4-4). Frequently the host young suffer high mortality and the cowbird nestling is raised alone (Fraga 1983; Chapter 6).
The Shiny Cowbird has a large geographic range and parasitizes a large number of species (Friedmann and Kiff 1985). It is, therefore, difficult to





33













70
.--- --- ---- --- . .-- - . adult males
60
----------------------------- adult females
50 ----------------------------- first molt
40 juveniles


Cowbirds (n=12)

0 20

J_ + --I-- --I--* 4
10 a Wrens (n=12)


0 2 4 6 8 10 12 14 16 18

AGE (days)

Figure 4-4. Growth rates of Shiny Cowbird and House Wren nestlings. Error bars indicate one standard deviation. Error bars after day 9 increase for cowbirds because of divergence in size between males and females.





34


make generalizations, and they probably exhibit different strategies under different circumstances. Shiny Cowbirds in the Cauca Valley seem to be adopting the "shotgun" strategy (Rothstein 1975). They monitor host nests and to some degree time their egg laying to coincide with a period in which their eggs are more likely to be successful. The unselective behavior of wrens, however, allows for a wide window of time in which their eggs are accepted. Furthermore, wrens are abundant and have an extended breeding season, providing a high availability of nests almost year-round. The Shiny Cowbird's strategy is to be an extreme generalist that bases its success on high fecundity more than on the precise placement of each egg. It is possible that species that follow other strategies, such as cuckoos (Wyllie 1981), have more precise behavioral and/or physiological mechanisms for synchronization of oviposition.











CHAPTER 5
MECHANISMS OF SHORT INCUBATION PERIOD IN THE SHINY COWBIRD


Introduction

"The comparatively short time the embryo takes to hatch gives it [the
Shiny Cowbird] another and a greater advantage; for whereas the eggs
of other small birds require from fourteen to sixteen days to mature,
that of the cowbird hatches in eleven days and a half from the moment
incubation commences." (W. H. Hudson 1920)


Brood parasitism is a reproductive strategy in which rapid embryonic
development is at a premium. Eggs of brood parasites are abandoned in nests of host species and hatchlings have to compete with foster nestmates for parental resources. Therefore, early hatching gives parasitic young a head start (Payne 1977). One strategy used to eliminate competition in the nest is displayed by parasitic cuckoos. Cuckoos usually hatch earlier than their hosts. Upon hatching, the cuckoo ejects host eggs by pushing them over the rim of the nest with their backs (Wyllie 1981). Cowbirds, in contrast, do not kill nestmates, but early hatching helps them to gain an advantage over their nestmates (Payne 1977).
It has long been recognized that brood parasites have short incubation periods (Friedmann 1927; Hamilton and Orians 1965; Payne 1977; Briskie and Sealy 1990). There is uncertainty, however, about the precise duration of incubation periods because of the inherent variability associated with variable attendance by incubating birds. This problem led Nice (1953) to question the validity of reports of ten-day incubation periods in the Brown35





36


headed Cowbird. One problem with measuring incubation periods in the field is the uncertainty in determining when did incubation begin and how constant it was. Briskie and Sealy (1990) defined incubation period of cowbird eggs as the time from initiation of full incubation by the host to hatching. Initiation of full incubation, however, was defined as "usually just after laying of the penultimate egg" (by the host). Although Briskie and Sealy (1990) found that incubation periods of parasitic cowbirds were shorter than expected from egg size, when compared with non-parasitic relatives, the uncertainty in incubation periods led them to do separate analysis for minimum and mean incubation periods of cowbird eggs.
Here I present measurements of incubation periods of Shiny Cowbird eggs under constant conditions in the laboratory and examine possible mechanisms for the evolution of short incubation periods in parasitic cowbirds. Friedmann (1927) suggested that a short incubation period in Brown-headed Cowbirds represented an adaptive acceleration of embryonic development. Incubation period, however, scales allometrically to the 0.217 power of fresh egg mass (Rahn and Ar 1974). Egg mass, in turn, scales to the
0.675 power of adult female body size (Rahn et al. 1975). Thus, incubation periods in parasitic birds may be short because eggs are smaller than expected (Briskie and Sealy 1990). Parasitic cuckoos lay small eggs and have correspondingly short incubation periods, usually hatching up to four days before their hosts (Rahn et al. 1975; Payne 1977). Cowbirds, in contrast, lay normal-sized eggs, i. e., egg size corresponds to that expected from female body size. Still, their incubation periods are shorter than expected and eggs usually hatch 1-2 days before their hosts, even though cowbird eggs are usually larger (Payne 1977; Briskie and Sealy 1990).





37


I examined two possible mechanisms by which cowbirds may shorten incubation periods. The first mechanism is based on the assumption that there is a window of time at the end of embryonic development during which hatching can occur. Avian embryonic growth can be represented by the parabolic equation W=atb where W is embryo size, t is incubation time and a and b are constants (Fig 5-1a; Ricklefs 1987). Embryonic development can be divided into two phases. During the organogenesis phase there is rapid development of new tissues to form organs, but relatively little increase in embryonic mass. Following the differentiation phase comes a phase of rapid growth (increase in embryonic mass) and functional maturation of tissues (Freeman and Vince 1974; Balinsky 1975). Thus, it is conceivable that there is a time to after which maturation is complete enough to allow hatching (Fig 5-1a). Embryos continue to grow after time to and hatch at time te, presumably because there is an advantage in hatching at a larger size. I hypothesized that cowbirds shorten the incubation period by hatching at to (Fig. 5-1a). Because hatchling size correlates with egg mass (Vleck and Vleck 1987), I predicted that hatchling size (Wo) would be smaller than expected from egg mass (We). I also hypothesized that the mechanism to shorten incubation period within this window of time (to-te) involved a decrease in energy content of the egg. During the final days of development, growth presumably continues until yolk reserves are depleted (except for a small amount that is retained as reserve for the hatchling). Depletion of yolk reserves or some unknown yolk component would signal the embryo that it is time to hatch. This hypothesis is based on the existence of a tight correlation between incubation period and egg energy content (tighter than the correlation between incubation period and egg mass) and the observation that all avian embryos expend about the same proportion of energy stored before




















Figure 5-1. Hypotheses to explain short incubation periods in cowbirds. A) Embryo size as a function of incubation time. The expected incubation time for a given egg size is te, and hatchling size is we. The hypothesis is that there is a time to after which hatching is possible. This hypothesis predicts that hatchling size (wo) will be smaller than expected from egg mass. B) The alternative hypothesis is that cowbirds shorten incubation time by increasing developmental rates. This hypothesis predicts that hatchling size will be the expected from egg size (we). C) Embryonic metabolic rate as a function of embryo size. Hypothesis B predicts that metabolic rate will be higher than expected from embryo size.






39



- ORGANOGENESIS 4- GROWTHWe ----------------------------- --------------WoI
VzI 0I vvo ------------------------------------WeI
I I




































INCUBATION TIME to te
0



SI
I g
I I I I




























INCUBATION TIME C.)


















EMROSIEW





40


hatching (Vleck and Vleck 1987). Assuming that the rate of energy expenditure is constant, a reduced energy content would result in a shorter incubation period. I predicted that cowbirds 1) would have an energy content lower than expected from egg mass, 2) incubation period would be shorter than expected from egg mass but not different from the expected from egg energy content, and 3) hatchling size would correspond to the expected from energy content. Expected values for these variables were derived from Vleck and Vleck's (1987) review on the energetics and metabolism of avian embryos.
An alternative hypothesis is that cowbirds could shorten incubation period by increasing developmental rates (Fig. 5-1b). In this case, I predicted that hatchling size would not be different from that expected from egg size. I also tested this hypothesis by measuring embryonic metabolic rates, under the assumption that an increased developmental rate would result in an increased rate of oxygen consumption (Vleck et al. 1980). Metabolic rates of bird embryos are usually compared using the pre-internal pipping metabolic rate (PIP-V02), defined as the metabolic rate just prior to the perforation of the internal air cell of the egg, which occurs shortly before hatching (Vleck and Vleck 1987). I predicted that cowbird embryos would have higher PIPVO2 than expected from the corresponding embryo size. These two hypothesis are not mutually exclusive and both mechanisms may be operating simultaneously.


Methods


During 1990, I collected fresh Shiny Cowbird eggs from House Wren nests in the study area. I collected only cowbird eggs that were laid before the wrens started laying to ensure that the eggs had not been incubated in the





41


field. Eggs were incubated in the laboratory at a constant temperature of 38 OC. Eggs were turned manually 6-8 times a day. After day 10, I checked the eggs every 4 hours during the day and every 6 hours during the night to determine hatching time. Hatching was defined as the moment in which the shell started to split (actual emergence of the hatchling occurred 2-3 hours after the shell split). Incubation time of cowbird eggs in the field was defined as the time between the laying of the last wren egg (which is when constant incubation by the female wren usually begins) and the day of hatching. Only cowbird eggs that were laid before the last wren egg were used. Also, for determination of incubation period in the field I only used nests in which incubation had proceeded without major interruptions.
Laboratory-incubated eggs were used for measurements of embryonic
metabolic rates. Metabolic rates were measured at 38 OC in a Scholander-type respirometer, using the method described in Hoar and Hickman (1983). In brief, eggs were placed in a vial (egg chamber), together with filter paper soaked in a 10% solution of KOH (C02 absorbent). The compensation vial contained only the C02 absorbent. A V-shaped tube connected vertical tubes placed above the egg and compensation chambers. The V-tube contained a small amount of water colored with anilin. The vertical tube above the animal chamber was connected to a calibrated syringe. The egg was placed in the vial and the apparatus was immersed in a water bath. I waited 10 min for temperature in the chambers to equilibrate with water temperature.
After temperatures equilibrated, I closed stopcock valves on the vertical tubes and started a measurement of oxygen consumption. After each measurement, the manometer fluid was pushed back to its original position with the syringe and the volume of oxygen consumed was read from the syringe graduations. I made at least four consecutive measurements each





42


time and averaged them to obtain a value of oxygen consumption. Fresh air was allowed in the chamber between readings. Oxygen volumes were converted to STPD conditions.
Additional cowbird eggs were collected for measurements of energy
content. I hard boiled 10 eggs and removed the shell. Energy content of the egg was then measured by bomb calorimetry at the Industrial Analysis Laboratory of the Universidad del Valle (Cali, Colombia). Hatchling mass was determined from laboratory-incubated eggs. I removed residual yolk from hatchlings and dried their carcasses at 70 oC until a constant mass was obtained.
For data analyses, I compared the observed values with expected values
for each parameter, derived from least-squares regressions published in Vleck and Vleck (1987). Significance of the deviations was tested by comparing values with the 95% prediction interval, calculated as described in Montgomery (1984), except for egg mass as a function of adult female body size, which was compared with the 68% confidence interval provided in Rahn et al. (1975).


Results


Egg size of Shiny Cowbirds (x = 4.3 g, SD = 0.4, range = 3.5-5.2 g, n=67) was on the lower range of values expected from adult female body size. Egg mass expected for an average female cowbird of 54 g is 5.1 g (68% confidence interval = 4.1-6.2; Rahn et al. 1975; Table 5-1). Incubation period was shorter than expected from egg mass. Incubation period of Shiny Cowbird eggs under constant laboratory conditions ranged between 11.2-12.1 days (x = 11.7 days, SD = 0.5, n=11), slightly overlapping with the lower limit of the 95%





43







Table 5-1. Female size, egg size and incubation period of Shiny Cowbirds and House Wrens in the Cauca Valley, Colombia.


Parameter Shiny Cowbird House Wren


Female size (g) 54.6�3.45 (47) 15.34�1.15 (16) Egg Size (g)

Observed 4.3�0.39 (49) 1.87�0.13 (34)

Expected* 5.1 2.15 Incubation Period (days)

Observed

Laboratory 11.7�0.5 (11) --Field 12.0�0.8 (7) 15.3�0.8 (22) Expected

From Egg Size** 14.25 11.19

From Energy Content*** 13.8 ---------------------------------------------------------------------------* From Rahn, Paganelli and Ar's (1975) equation for passerines (E=0.34WO.677).
** From Vleck and Vleck's (1987) equation for altricial birds (log 1=0.97+0.29 log E).
*** From Vleck and Vleck's (1987) equation for all birds (log I=0.83+0.27 log EC).





44


prediction interval (11.8-16.9 days) derived from Vleck and Vleck's (1987) equation (predicted incubation period = 14.3 days; Table 5-1). Incubation period of cowbird eggs correlated significantly with egg mass (Fig. 5-2). Egg mass, however, explained only 44% of the variation in incubation period, suggesting that other factors are also influencing incubation period (Fig. 5-2).
Estimates of cowbird incubation period in the field agreed with laboratory results. Incubation periods in the field ranged from 11 to 13 days, much shorter than the 14-16 day incubation periods of House Wrens (Table 5-1). This resulted in cowbirds hatching earlier than wrens. In 15 wren nests in which one cowbird egg was laid before the initiation of incubation and that survived the incubation stage, the cowbird always hatched 2-3 days before the wrens. The short incubation period of cowbirds allowed them to hatch synchronously with the wrens even when cowbird eggs were laid after the initiation of incubation. In 8 wren nests in which a cowbird egg was laid 1-4 days after the last wren egg was laid, cowbirds hatched the same day or the next day after the wrens.
As predicted, energy content of Shiny Cowbird eggs (14.54 � 0.19 kJ, n=10) was lower than expected from mean egg mass (expected energy content = 19.05 kJ, 95% prediction interval, 16.3-22.2; Table 5-2). The incubation period of cowbird eggs (11.7 days) was within the limits expected from egg energy content (95% prediction interval, 11.4-16.9; Table 5-2). Yolk-free, dry hatchling mass of Shiny Cowbirds (0.3091 � 0.04 g, n=6) was lower than the lower limit of the prediction interval expected from egg mass (expected value=0.3786, 95% prediction interval, 0.3105-0.4613), but close to the value predicted from egg energy content (expected value=0.2835, 95% prediction interval, 0.2371-0.3350; Table 5-2).





45










12.5



" 12.0



' 11.5



S11.0
3.0 3.5 4.0 4.5 5.0 EGG MASS (g) Figure 5-2. Incubation period under constant laboratory conditions as a function of fresh egg mass for M. bonariensis (P=.03, r2=.44, n=11).





46








Table 5-2. Egg energy content, hatchling mass and embryonic metabolic rate of Shiny Cowbirds. Expected values as a function of egg mass and egg energy content, calculated from equations in Vleck and Vleck (1987). Numbers in parenthesis under the observed column are sample sizes and under the expected column are 95% prediction intervals (Montgomery 1984).


--------------------------------------------------------------------------Parameter Observed Expected from

Egg Mass Energy Content
----------------------------------------------------------------------------Energy Content (kJ) 14.54�0.19 19.05
(10) (16.3-22.2)
Energy Used (kJ)* 4.45 6.95 5.28

Yolk-free, Dry
Hatchling Mass (g) 0.3091�0.04 0.3786 0.2835
(6) (0.3105-0.4613) (0.2371-0.3350
PIP-VO2 (ml/day) 50.19 47.86 (37.9-64.4)
----- ------------------------------------- ------------------------------* Calculated as the area under the metabolic rate function for t=0-11.7 (Fig. 5-3).





47


The metabolic rate of Shiny Cowbird embryos followed the typical pattern of altricial birds, with metabolic rate increasing exponentially with incubation time (Fig. 5-3). PIP-V02 was calculated as the metabolic rate expected from the metabolic rate function at 90% of the mean incubation period (Fig. 5-3). The PIP-V02 (50.19 ml/day) was within the limits of the prediction interval expected from a regression of metabolic rate against energy content for altricial birds (predicted value=47.86, 95% prediction interval, 37.9-64.4; Table 5-2).


Discussion


Incubation of Shiny Cowbird eggs under constant conditions in the
laboratory resulted in a mean incubation period of 11.7 days. This incubation period is shorter than expected from egg size and remarkably short when compared with that of other passerines, even those smaller than cowbirds. For example, House Wrens are much smaller but have an incubation period of 14-16 days (Table 5-1; Alvarez et al. 1984). This incubation period is longer than the value of 11.19 days expected from the wren's egg size (Table 5-1). It would be expected that incubation periods of cowbird eggs under natural conditions in the field may be longer because of intermitent attendance by the incubating bird. Estimates of cowbird incubation periods in the field, however, agree with the laboratory results (Table 5-1; Briskie and Sealy 1990). Incubation periods of Shiny Cowbird eggs incubated in wren nests under natural conditions varied between 11 and 13 days (Table 5-1). This resulted in cowbirds hatching earlier than wrens when cowbird eggs were laid before the initiation of incubation. Even when cowbird eggs were laid 1-4





48













4
0

2- (



1

0 2 4 6 8 10 12 INCUBATION TIME (days) Hatch






Figure 5-3. Oxygen consumption (STPD) of Shiny Cowbird embryos as a function of incubation time. Seventy-eight measurements based on 12 eggs.





49


days after the initiation of incubation, the short incubation period allowed cowbirds to hatch at around the same time that the wrens hatched.
As predicted, I found that Shiny Cowbird eggs had an energy content lower than expected from egg mass (23.8% lower; Table 5-1). Incubation period was 28.8% shorter than expected from egg mass, but was closer to that predicted from energy content (Table 5-1). Similarly, hatchling size was smaller than predicted from egg mass but not significantly different from that predicted from the actual energy content of cowbird eggs. These results support the hypothesis that cowbirds shorten incubation period by decreasing energy content of the egg (Fig 5-1a). The alternative hypothesis that cowbird embryos have accelerated rates of development (Fig 5-1b, 5-1c), in contrast, was not supported. Hatchling size was smaller than expected from egg mass. Measurements of metabolic rates did not support the hypothesis that cowbirds had high growth rates. Pre-internal pipping metabolic rates of Shiny Cowbird embryos were not different from the expected from a regression of PIP-VO2 on egg energy content in altricial birds (Table 5-2), suggesting that developmental rates were not higher than expected based on embryonic size.
My results suggest that the mechanism by which cowbirds shorten incubation period is a combination of laying a slightly smaller egg than predicted by female body size, that has a reduced energy content. One potential problem with this mechanism is that reducing hatchling size could be a disadvantage, because cowbirds have to compete with nestmates for food delivered by the foster parents. Cowbirds, however, usually parasitize species smaller than themselves (Friedmann and Kiff 1985). In this case, the slightly smaller hatchling size would be irrelevant. Wren hatchlings, for example, weigh less than 2 g, as opposed to 3.2-4.5 g for a cowbird hatchling, and this





50


difference increases rapidly with nestling growth (see Fig. 4-4). If the host species is larger than the cowbird, the latter may be at a disadvantage (e. g., Fraga 1985; Carter 1986). The advantage of hatching early, however, probably outweighs the disadvantage of a slightly smaller hatchling.
Energy stored in the egg is a major factor influencing avian embryonic development (Vleck and Vleck 1987). Each egg is provided with a fixed amount of energy, which is used for growth and somatic maintenance of the embryo (except for a small amount that is retained as residual yolk in the hatchling). Energy content of the egg explains most of the interspecific variance in variables associated with embryonic development, such as incubation period, hatchling size and energetic cost of development (i. e., total energy spent during development). When these variables are regressed against egg mass, two separate lines are required to describe altricial and precocial species. This difference disappears when energy content is used as independent variable (Vleck and Vleck 1987).
The correlation between incubation period and energy content, however, does not imply causality and the two variables may be correlated with a third, unknown variable. In a discussion of the evolution of avian altriciality, Vleck and Vleck (1987) argue that a mutation occurred that caused a shortening of the incubation period. The resulting hatchling would be smaller, but would have a large amount of residual yolk. At this stage, there would be selection on the female to invest less energy in each egg. Thus, according to this interpretation, the incubation period got shorter first by an independent mechanism and then the energy density of the egg was decreased (Vleck and Vleck 1987; C. M. Vleck, pers. com.). There is experimental evidence, however, that reducing the energy content of the egg results in a shorter incubation period in other vertebrates. Sinervo (1990)





51


manipulated energy content of eggs of the lizard Sceloporus occidentalis by extracting variable amounts of yolk (10-50% of total egg mass). Some eggs were sham-manipulated (poked with a syringe) but no yolk was extracted. He found that eggs with more yolk removed had shorter incubation periods than unmanipulated or sham-manipulated eggs. Eggs from which yolk was removed developed normally and hatched into viable offspring that were smaller than those from control eggs (Sinervo 1990). Sinervo (1990) also found interpopulational differences in incubation time and hatchling size, among other variables, and his experimental manipulation of yolk content (i. e., energy content) demonstrated that these differences were probably due to differences in yolk content (reflected in egg size) among populations.
Sinervo's (1990) results lend support to the hypothesis that cowbirds could shorten incubation period by reducing the energy content of the egg. The mechanism triggering early hatching may be simply that the embryo is running out of yolk (Sinervo 1990). If there is a threshold time to after which hatching is possible (Fig. 5-1a), the amount of yolk reserve left could determine the timing of hatching. When yolk is almost depleted, the embryo would receive some signal that it is time to hatch. If more reserves are available, the embryo could continue increasing in mass. This hypothesis assumes that incubation period in altricial birds is not already minimized, i. e., that the interval to-te exists. Opposite selection for short incubation period versus large hatchling size is likely to result in the existence of this interval in which there is a balance between the two forces.
Hamilton and Orians (1965) suggested that short incubation periods are a pre-adaptation to brood parasitism (an exaptation in the terminology of Gould and Vrba 1982). Short incubation periods may have evolved in different lineages in response to different selection pressures. For example,





52


Hamilton and Orians (1965) suggested that accelerated incubation periods and nestling growth are an adaptation of cowbirds for feeding among nomadic herds of large mammals. The short incubation period would have evolved in response to the need to be constantly on the move. This then would have contributed to the success of cowbirds as brood parasites.
If short incubation periods are a preadaptation to brood parasitism, then close relatives of cowbirds would be expected to exhibit the same trait. Hamilton and Orians (1965) suggested that the incubation period of parasitic cowbirds was not shorter than that of non-parasitic blackbird relatives. Briskie and Sealy (1990), however, compared incubation periods of 22 species of icterines as a function of egg size, and found that parasitic cowbirds had shorter incubation periods than non-parasitic icterines. They suggested that short incubation periods evolved in cowbirds as an adaptation for brood parasitism. This issue will remain unresolved until a comparative study of female size, egg size, incubation period and embryo energetics of icterines is conducted.
Here I have indicated that short incubation periods in cowbirds may be the result of a reduced energy content. This suggests an alternative scenario for the evolution of short incubation periods in a life history theory framework. Cowbirds are extremely fecund birds, laying almost continuously during the breeding season. Life history theory predicts a trade-off between number and quality of offspring. Increasing the number of offspring implies a decrease in the investment in each offspring (Roff 1992). Cowbirds have adopted a strategy of laying a large number of eggs with a reduced investment in each egg (Chapters 3, 4). Thus, the reduced energy content of cowbird eggs could be the direct result of this trade-off.




53


What is, then, the selective pressure for a reduced energy content in cowbird eggs? Elucidation of this question is difficult because the three factors, namely incubation period, energy content and fecundity, are functionally linked (Fig. 5-4). If there is a selective pressure for shortening incubation period, this would result in a reduced energy content, which then could result in an increase in fecundity. The opposite pathway is also possible. If the reduced energy content is a result of a selective pressure to increase fecundity, then the short incubation period is a side benefit and not a specific adaptation for brood parasitism. Alternatively, both incubation period and fecundity could be acting as selection pressures to decrease energy content of cowbird eggs.
Perhaps the best way to address this problem is by conducting a
comparative study of icterines, as suggested above. Comparison of these characteristics among related species (including non-parasitic species which do not exhibit the high fecundity of cowbirds) would allow to determine which characteristics are primitive or derived, and which of these characteristics, if any, are specific adaptations of cowbirds for brood parasitism.





54









LOW EGG ENERGY CONTENT







HIGH FECUNDITY SHORT INCUBATION PERIOD








Figure 5-4. Functional relationship between fecundity, egg energy content and incubation period in cowbirds. A selection pressure for high fecundity results in a decrease in the energy deposited in each egg, which then results in a short incubation period. In this case the short incubation period is a side benefit and not a specific adaptation for brood parasitism. In the opposite pathway, the selective pressure for a low energy content is the incubation period.











CHAPTER 6
IMPACT OF BROOD PARASITISM: WHY DO WRENS ACCEPT COWBIRD EGGS?


Introduction


Because brood parasitism usually depresses reproductive success of the host, hosts are expected to evolve defenses against parasitism (Payne 1977; Rothstein 1990). One such defense is rejection of parasitic eggs. Hosts of parasitic cowbirds can be classified into two discrete groups, "accepters" and "rejecters," according to their responses to natural and experimental parasitism (Rothstein 1975a; Mason 1986a). Parasitic eggs may be rejected by ejecting the egg, deserting the nest, or constructing a new nest floor over the parasitized clutch.
Why do some hosts accept parasitic eggs? Rothstein (1975a, 1982, 1990) suggested that lack of rejection in some cowbird hosts is best explained by absence of genetic variants in the population with the ability to recognize and reject a foreign egg (evolutionary lag hypothesis). An alternative hypothesis suggests that for some cowbird hosts, rejection incurs costs that exceed any benefits, thereby making acceptance the most adaptive option (cost of rejection hypothesis). For example, egg ejection may be costly for hosts with small bills because they are unable to grasp a cowbird egg between the mandibles. The only way to eject the egg would be to puncture it and hold it by the perforation. Cowbird eggshells, however, are unusually thick and therefore very strong and some of the host's own eggs could be damaged in the puncturing process (Rohwer and Spaw 1988; Rohwer et al. 1989; R0skaft


55





56


et al. 1990). This hypothesis may explain why most small-billed hosts accept eggs of the Brown-headed Cowbird (Rothstein 1975a, 1982; Rohwer and Spaw 1988).
Rejection of parasitic eggs by nest desertion also incurs costs that may exceed the costs of accepting parasitism. Nest desertion involves a time cost, which may be significant if duration of the breeding season is restricted. Besides, good nest sites may be in short supply and the replacement nest may also be parasitized (Rskaft et al. 1990; Petit 1991). Thus, to determine the best strategy for a host to follow, the costs of parasitism must be contrasted with the costs of rejection.
The cost of rejection hypothesis assumes that hosts have the ability to recognize foreign eggs, but they tolerate parasitism because the costs of rejecting are higher. Here I test this hypothesis by examining the impact of Shiny Cowbird parasitism, and host responses in the House Wren. It has been hypothesized that cavity-nesting species experience low rates of parasitism, and as a result these species have not evolved egg discrimination behavior (Friedmann 1963; Friedmann et al. 1977). An alternative hypothesis is that wrens can discriminate foreign eggs, but are unable to grasp-eject cowbird eggs because they are large and heavy. I tested the following hypotheses and predictions:
1) Wrens can discriminate foreign eggs but can not lift cowbird eggs; thus, they should eject artificial odd-looking eggs that are smaller and lighter than cowbird eggs, but similar in size and weight to wren eggs.
2) Wrens can grasp-eject cowbird eggs but do not recognize them as foreign; thus, wrens should accept cowbird-sized artificial eggs but should eject odd-looking objects as large and heavy as cowbird eggs.




57


I also examined the potential costs of nest desertion and contrasted them with the costs of accepting parasitim, to test the hypothesis that wrens tolerate parasitism because the costs of deserting are higher than the costs of accepting.


Methods
For this study, I followed the nesting histories of all pairs of wrens in the area and obtained data on incidence (proportion of nests parasitized) and intensity (number of parasitic eggs per nest) of cowbird parasitism. I evaluated the effects of parasitism on nesting success of wrens by comparing clutch size, brood size and number of wren fledglings in parasitized and unparasitized nests.
To test the discriminatory abilities of wrens, and whether they are capable of grasp-ejecting cowbird eggs, I placed real and artificial cowbird eggs in nests during the pre-laying (1-4 days before wrens laid their first egg) and egg-laying stages (Table 6-1). I tested wrens during the pre-laying stage because this is the period when most cowbird eggs are laid (Chapter 4). For artificial eggs I used commercially obtained plastic egg models. These models are hollow and very light. To make them heavier I filled them with water (Table 6-1). Models were scored as accepted when they were incubated, and rejected when they disappeared from a nest that was not abandoned or lost to predation. Hypothesis 1 was tested by placing small artificial eggs in wren nests. These eggs were only slightly larger than a wren egg but were very contrasting because of their orange-colored markings (Table 6-1). Hypothesis
2 was tested by placing artificial cowbird-sized eggs in wren nests (large artificial eggs in Table 6-1). I placed both light and heavy large eggs to test whether wrens could grasp-eject a large but light egg. To test whether both








Table 6-1. Characteristics of wren and cowbird eggs and egg models placed in House Wren nests in the Cauca Valley.



Model Dimensions (mm) Mass (g) Color of Background / Markings


Wren Egs 18.5 � 0.5 x 1.9 � 0.2 Creamy white / reddish-brown 13.8 � 0.5

Cowbird Eggs 24.1 � 1.4 x 4.2 + 0.4 Creamy-white / brown 18.2 � 0.8

Artificial Eges

Large 23.5 x 18.4 White / brown

Heavy 4.2 Light 1.0

Small 20.0 x 15.7 White / orange

Heavy 2.6 Light 0.6

Glass Beads 16.0 x 16.0 5.2 Varied




59


size and weight constrained the wren's ability to lift a cowbird egg, I placed glass beads in wren nests. The diameter of glass beads was similar to the breadth of a cowbird egg, but they weighed 23% more (Table 6-1). I expected wrens to eject these odd-looking objects. Therefore, acceptance would indicate that wrens can not lift them.


Results


Impact of Parasitism
Parasitism was extremely common in this population of House Wrens, and nest boxes were more frequently parasitized than natural nests. Rate of parasitism in natural cavities was 58.8% (20 of 34 nests), versus 94% (99 of 105 nests) in nest boxes (G=22.46, df=1, P<.001; Table 6-2). Nest boxes were also more intensely parasitized than natural nests. Eleven of 20 (55%) parasitized nests in natural cavities received more than two cowbird eggs. In contrast, 84 of 105 nests (80%) in boxes received more than two cowbird eggs (G=6.65, df=1, P<.01; Table 6-2). The average number of cowbird eggs per nest was also significantly higher in boxes than in natural cavities (MannWhitney U, P<.001; see Fig. 4-3).
Parasitism had a strong negative effect on hatching success of wrens. Only in 55% of parasitized nests did any wrens hatch, compared with 92% of unparasitized nests (Table 6-3). Most wren mortality ocurred during incubation. Cowbirds did not remove host eggs, and wren clutch size at the beginning of incubation was not different between parasitized (2.97�0.5) and unparasitized nests (3.12�0.5; Table 6-4). Clutch size at hatching, however, was significantly reduced in parasitized nests (0.99�1.0 vs. 2.83�0.9; Table 64). Clutch reduction was due to wren eggs disappearing during incubation,





60











Table 6-2. Incidence and intensity of Shiny Cowbird parasitism on House Wrens.

-----------------------------------------------------------------------------Number of Nests (%)

Number Eggs/Nest Natural Boxes
-----------------------------------------------------------------------------0 14 (41.2) 6 (5.7) 1-2 9 (26.5) 15(14.3) 3-12 11 (32.3) 84 (80.0)

x � SD* 1.97 � 2.2 5.65 � 3.2

* Mann Whitne ----------------------------------------- U P<.001
*Mann Whitney U, P<001





61










Table 6-3. Hatching success of wrens (excluding effects of predation).



Number of nests Hatch. Success* G P Parasitized 20 11 5.9 <.025 Unparasitized 13 12

* Number of nests in which at least one wren hatched.





62













Table 6-4. Reproductive success of wrens in parasitized and unparasitized nests (excluding predation). Numbers indicate mean, standard deviation and in parenthesis, sample size.

----------------------------------------------------------------------------Parasitized Unparasitized P*
----------------------------------------------------------------------------Initial Clutch Size 2.97 � 0.5 (20) 3.12 � 0.5 (13) .3 Clutch Size at Hatching 0.99 � 1.0 (15) 2.83 � 0.9 (11) <.01 Fledglings/Nest 0.54 � 0.9 (11) 2.78 � 0.7 (9) <.001
* Mann-Whitney U.
* Mann-Whitney U.





63


probably removed by wrens because they broke when jostled against the heavy, thick-shelled cowbird eggs. During incubation I frequently found dented wren eggs in parasitized nests, and these eggs always disappeared in the next 2 or 3 days.
Fledging success of wrens was also significantly reduced by cowbird parasitism. Wrens laid 3.12�0.5 eggs in unparasitized nests and fledged
2.78�0.7 young, while in parasitized nests they laid 2.97�0.5 eggs and fledged
0.54�0.9 young (Table 6-4). Most wren nestlings in parasitized nests died of starvation within 3-4 days of hatching. Only 4 of 11 parasitized nests (36.4%) that survived the entire nesting cycle produced wren fledglings, as opposed to
8 of 9 (88.9%) unparasitized nests (G=6.22, df=1, P<.02).


Host responses to parasitism
Wrens abandoned 92 of 95 nests that received more than two cowbird eggs but showed no rejection at 24 nests that received one or two parasitic eggs (G=100.8, df=1, P<.001). There were three cases in which wrens accepted and incubated large cowbird clutches (two with 6 and one with 7 cowbird eggs), all in nest boxes. Two of these nests were abandoned halfway through incubation. Only one cowbird hatched in the third nest. Two other cowbird eggs in this nest had advanced embryos but did not hatch.
Thirteen nests received one cowbird egg and were sucessfully incubated by wrens. In nests parasitized with two cowbird eggs, usually one failed to hatch. Only in two of 11 nests with two cowbird eggs did both cowbirds hatch. In one of these, the second cowbird hatched two days after the first one and died three days later. In the other nest, both cowbirds hatched the same day and 10 days later both were growing normally, but then disappeared.





64


Probably both would have fledged, because cowbirds usually fledge at 12 days of age.
Wrens accepted and incubated 14 real cowbird eggs that I placed in their nests, both before they initiated laying and during their laying period (Table 6-5). Wrens also accepted both large and small heavy artificial eggs, but ejected light eggs, except during the laying period, when they were equally likely to accept or reject large light eggs (Table 6-5). This result shows that they can grasp with the bill an object as large as a cowbird egg, and that they can discriminate eggs by weight. Wrens also ejected glass beads (Table 6-5), which were heavier than real cowbird eggs, thereby showing that failure to reject real cowbird eggs and small and large heavy artificial eggs was not due to an inability to grasp and lift such eggs. These experiments indicate that wrens have the ability to grasp-eject parasitic eggs, but do not discriminate foreign eggs based on color or size.


Potential Costs of Nest Desertion
Several factors could limit the benefits of nest desertion. First, wrens could be limited by a short breeding season. At my study site, however, wrens had an extended breeding season, nesting almost year-round (Chapter 3). A similar pattern was reported by Alvarez et al. (1984) for a site about 10 km north of my study site. Birds that deserted nests because of multiple parasitism always started building a new nest within 2-3 weeks (except when desertion occurred close to the onset of the mid-year dry season). During 1989 and 1990, 10 banded females made 3.5�1.35 nesting attempts per year, and some females made up to six attempts per year. Therefore, length of the breeding season is not likely to be a factor that limits renesting possibilities.









Table 6-5. Responses of House Wrens to artificial eggs placed in their nests, during the pre-laying and laying stages of the nesting cycle.



Pre-laying P* Laying P* Accept Reject Accept Reject


Cowbird egge 6 0 <.02 8 0 <.01 Artificial ees

Large

Heavy 7 0 <.01 5 0 <.03 Light 1 9 <.01 3 2 .5 Small

Heavy 8 0 <.01 Light 0 6 <.02 Glass heads 0 4 <.05 Chi-square test. C





66


A second potential factor limiting renesting possiblities is availability of nest sites. Because my study site was saturated with nest boxes and all wrens had at least 6 boxes within their territories, I can not tell whether desertion of parasitized nests was contingent upon availability of nest sites, as Petit (1991) did for Prothonotary Warblers (Protonotaria citrea). Most of the boxes remained in the field for 2.5 years and during this time only one new territory was formed, when in early 1990 a pair of wrens nested in a box placed between two old territories. This suggests that nest site availability was not limiting the establishment of territories. Territories, however, probably varied in the number and quality of nesting cavities. Most wrens used both natural cavities and nest boxes during the study, but in some territories wrens never used nest boxes, while in others they always used nest boxes. Several instances of wrens destroying eggs and nestlings of Spectacled Parrotlets (Fors conspicillatus) and taking over the nest box, suggest competition for cavities (N. G6mez and G. Kattan, unpublished data). In relation to brood parasitism, a good quality cavity is probably one with a small opening. In natural nests, diameter of the entrance of parasitized nests (57.0�9.7 mm, n=12) was significantly larger than the entrance of unparasitized nests (23.0�5.8 mm, n=15; t=8.96, P<.001). All my nest boxes had openings larger than 35 mm, as cowbirds could not enter boxes with smaller openings. Thus, at least in some territories, wrens may be limited by availability of good quality cavities.
A third factor that would limit the benefits of desertion is probability of parasitism of a replacement nest. Because nest boxes were more frequently and intensely parasitized than natural nests, I separated the sample of birds renesting after deserting a parasitized nest into birds renesting in boxes and in natural sites. Of 17 birds renesting in boxes, all were parasitized, while





67


one of six birds that renested in natural cavities were parasitized (G=13.6, df=1, P<.01). Parasitism rate of replacement nests in natural cavities was significantly lower than the 59% overall parasitism rate in natural cavities (G=3.88, df=1, P<.05). After deserting a multiply parasitized nest, birds became wary and difficult to observe, and took a long time to renest (interval between desertion and initiation of a new clutch, x=31.1 � 14.8 days, n=23; sample does not include nests abandoned just prior to the beginning of the dry season). In contrast, birds that abandoned for other reasons (human interference, nest flooding, predation) renested in 13.1 days (SD=7.2, n=18; Mann Whitney U, P<.01).


Discussion


Shiny Cowbird parasitism substantially depressed the nesting success of House Wrens. Most of the cost of parasitism was paid during incubation, probably because wren eggs broke when jostled against the thick shelled cowbird eggs. Disappearance of wren eggs during incubation resulted in a hatching success of only 55% in parasitized nests, as opposed to 92% in unparasitized nests (Table 6-2). Cowbird parasitism also reduced the fledging succes of wrens. Wrens that hatched soon died of starvation, outcompeted by the more aggressively begging cowbird nestling (Gochfeld 1979; pers. obs.). Wrens fledged in only 36% of parasitized nests, while 90% of unparasitized nests produced fledglings.
The frequency of cowbird parasitism reported in this study is very high for a cavity nester. Petit (1991), for example, reported an incidence of parasitism of 21% for the Prothonotary Warbler in Tennessee, USA, in both nest boxes and natural cavities. Shiny Cowbirds are very abundant in the Cauca Valley,





68


a region with extensive plantations of rice and sorghum, two crops on which they feed. In contrast to natural nests, the nest boxes I used in this study were conspicuous and revealed the high potential of cowbird parasitism. Probably a large proportion of wren nests in the Cauca Valley are found by cowbirds, and the only defense wrens have is to nest in cavities with small openings. Woodward and Woodward (1979) also reported high levels of parasitism on a population of Eastern Bluebirds (Sialia sialia) in Virginia, and they also attributed it to the conspicuousness of nest boxes, large size of openings and high density of cowbirds.
Although in some cases cowbird parasitism may have little impact (Smith 1981; Wheatherhead 1989), in most cases it is detrimental to the reproductive success of the host. Detailed studies of both the Brown-headed Cowbird and the Shiny Cowbird show that parasitized birds produce less young than unparasitized birds (e. g., Klaas 1975; Fraga 1985; Wiley 1985). The magnitude of the impact is variable and losses may occur at the incubation stage or at the nestling stage. Brown-headed Cowbirds are reported to habitually remove a host egg when they lay (Sealy 1992), a habit that has also been reported for some populations of Shiny Cowbirds (e. g., Fraga 1985). At my study site I found no evidence of cowbirds removing wren eggs, and very rarely did they puncture wren or other cowbird eggs. Instead, losses during incubation probably occurred because host eggs broke when jostled against cowbird eggs, as has also been reported in other studies (Sick 1958; Blankespoor et al. 1982). Host eggs may also fail to hatch because of improper incubation (e. g., Klaas 1975; Petit 1991). Parasitism also reduces fledging success of the host because nestlings are unable to compete with cowbirds. This occurs primarily with small hosts (e. g., Fraga 1978, 1983; Marvil and Cruz 1989). Large hosts are better able to compete, and thus impact at this





69


stage is less severe for them (Fraga 1985; Weatherhead 1989; Roskaft et al. 1990).
Given that cowbird parasitism has a negative impact, wrens are expected to exhibit defenses. Wrens rejected parasitism by abandoning multiply parasitized nests but accepted when parasitized with only one or two cowbird eggs. This response could be explained if the cost of parasitism was an increasing function of the number of cowbird eggs in the nest. It could be argued that perhaps when parasitized with one or two cowbird eggs, the cost of rejecting is higher than the cost of accepting. This was not the case, however, because losses were total for wrens parasitized with two or more cowbird eggs. Even when parasitized with only one cowbird egg, the cost of accepting was probably higher than the cost of rejecting.
Wrens parasitized with one cowbird egg have the option of rejecting
parasitism, either by ejecting the egg or abandoning the nest. Both of these two alternative modes of rejection have potential costs, which must be balanced against the costs of accepting a parasitic egg. Acceptance of the parasitic egg would be aqdaptive only if losses due to parasitism are not total and if the reproductive success of individuals that accept is higher than the success of individuals that reject. This may occur only for certain hosts for which nest success is not severely depressed by parasitism. For hosts that suffer very high losses, such as House Wrens (see also Rothstein 1975a; Klaas 1978; Fraga 1978, 1983), almost any option should be better than accepting parasitism. Besides, wrens should be able to grasp-eject cowbird eggs. Empty plastic eggs were ejected, probably because they were perceived as empty shells, and this shows that wrens have the ability to discriminate eggs by weight and grasp with the bill an object as large as a cowbird egg. Ejection of glass beads indicated that wrens can eject objects as large and heavy as




70


cowbird eggs. Despite this ability, wrens never ejected real cowbird eggs or artificial eggs of the same size and mass. They also accepted small eggs despite their contrasting colors. These experiments indicate that wrens do not recognize cowbird eggs as foreign. Ortega and Cruz (1988) also showed experimentally that the Red-winged Blackbird (Agelains hoenicius), an accepter species, had the ability to reject objects as large as a cowbird egg. They argued that redwings accept because the frequency and consequences of cowbird parasitism were too slight to be a significant selection pressure. They further suggested that if there are costs associated with ejection that they were unable to detect, selection might even favor acceptance over rejection (see also Roskaft et al. 1990). This is not likely to be true for wrens because the cost of parasitism was very high.
Nest abandonment is another option for wrens. As cavity nesters, wrens may be limited by the availability of nest sites. Acceptance of cowbird eggs by Prothonotary Warblers was found to be dependent on the opportunity to renest (Petit 1991). In this case, nest site limitation and a short breeding season made it more adaptive for females to accept the relatively low cost of parasitism (Petit 1991). At my study site, however, it is unlikely that these factors limit the benefits of nest desertion as opposed to the cost of accepting parasitism. Although wrens in some territories may be limited by the availability of good nesting sites, the long breeding season should make it more adaptive for a female to abandon and make a renesting attempt later. This contention is supported by the fact that wrens readily abandoned multiply-parasitized nests, and replacement natural nests were less likely to be parasitized. Furthermore, the cost of accepting cowbird eggs was very high, i. e., 16 of 20 nests that were parasitized before the initiation of incubation failed to fledge any host young because of parasitism. When a host





71


experiences complete reproductive failure, as occurs with all hosts of the Common Cuckoo (Davies and Brooke 1989a) any form of rejection is likely to be favored by selection. In such circumstances, the only factor that can make acceptance more adaptive than rejection is the occurrence of costs when a host nest is not parasitized (Rothstein 1990). While this may explain acceptance by cuckoo hosts, which typically experience low rates of parasitism (Davies and Brooke 1989a, 1989b), it is unlikely to do so in the cowbird-wren system I studied as most nests (58.8%) in natural cavities were parasitized.
Given this evidence, it is intriguing why House Wrens do not reject
cowbird eggs. The selective advantage of rejection depends on the frequency and cost of parasitism (Rothstein 1975b; Kelly 1987; Davies and Brooke 1989b). If either one or both factors are low, individuals exhibiting the rejection response may have little or no advantage. However, if both factors are high, as is the case for House Wrens, the rejection response would be expected to rapidly become common in the population (Rothstein 1975b). One factor that may explain lack of rejection in House Wrens is that egg discrimination may be dificult inside a dark cavity. However, Mason and Rothstein (1986) found that Rufous Horneros (Furnarius rufus) discriminated cowbird eggs based on size. Also, in my study wrens rejected when parasitized with three or more cowbird eggs. The stimulus for desertion when three or more cowbird eggs are laid may be the total clutch volume (or a high frequency of cowbird intrusions). Because there is no overlap in egg sizes of cowbirds and wrens, and my experiments indicate that wrens can discriminate eggs by weight, rejection on the basis of size or weight alone should incur little or no cost. Thus, evolutionary lag, i. e., the absence of genetic variants with the ability to recognize foreign eggs (Rothstein 1982,





72


1990), seems to be the best explanation for acceptance of parasitic eggs in House Wrens.











CHAPTER 7
GENERAL DISCUSSION


Shiny Cowbirds in the Cauca Valley are extremely fecund. I estimate that each female lays approximately 120 eggs per year, equivalent to about 9.4 times her own body mass. Studies on both the Brown-headed Cowbird and the Shiny Cowbird indicate that they lay in a pattern that resembles that of domestic chickens that are not allowed to brood (Scott and Ankney 1983; Jackson and Roby 1992; this study). They lay almost continuously, without regressing the ovary and oviduct, during the entire breeding season. In the temperate zone, most female cowbirds lay throughout the short breeding season. In the Cauca Valley, the Shiny Cowbird population has an extended 9-month breeding season. Duration of the season for individual females is unknown, but my data suggest that females are reproductively active throughout most of the season, with only one interruption for molting.
As predicted by life history theory (Roff 1992), the high fecundity exhibited by cowbirds involves a trade-off. Laying a high number of eggs implies investing less in each egg. I divided investment into two components: A direct energetic investment, represented in the amount of energy deposited in each egg, and a behavioral investment represented in the process of nest selection. Shiny Cowbirds in the Cauca Valley lay eggs with an energy content lower than would be expected from egg mass (14.54 kJ versus the expected 19.05 kJ). This represents a saving of 24% of the energy that otherwise would be required by the cowbird. Although it is not clear whether the high number of eggs laid represent any physiological strain for the cowbird (Rothstein et al.



73





74


1986), it is conceivable that the energy saved allows the cowbird to lay more eggs than would be possible if each had a higher energy content.
Shiny Cowbirds also invest little energy in the careful placing of each egg in circumstances that would maximize the egg's chances of being accepted and incubated successfully by the host. This would involve monitoring nests in order to lay the egg in coincidence with the host's laying period, as well as having some mechanism to avoid competition from other female cowbirds, such as defending laying territories. These two conditions would probably result in the cowbird being able to lay fewer eggs. Cowbirds monitor host nests, and to a certain extent choose nests that are an appropriate stage for parasitism. Egg laying by cowbirds is clumped around its host's laying period, with 47% of the eggs laid during the three day laying period of wrens (Fig. 22). However, a large proportion of cowbird eggs is laid before wrens lay, sometimes so early during nest construction that these eggs are buried as wrens continue adding nesting material. In addition, cowbirds do not avoid multiply parasitized nests.
These results indicate that Shiny Cowbirds in the Cauca Valley are
following the "shotgun" strategy. Instead of carefully placing a few eggs, they direct their energy into egg production and permanently search for nests, parasitizing as many as they can. This strategy is made possible by the unselective behavior of wrens. Wrens accept cowbird eggs laid at any time before, during or after the laying period. This allows cowbirds a wide window of time in which to lay eggs. The short incubation period and the size advantage of the cowbird nestling over its nestmates contribute to the success of this strategy.
This study concentrated on the interaction between cowbirds and one of their hosts in the Cauca Valley. Cowbirds, however, are generalists,




75


parasitizing most passerines with which they occur (e. g., Cruz et al. 1985, Wiley1885; Mason 1986a). Features of the parasitic strategy of cowbirds are related to their generalist habits. Laying large numbers of eggs requires large numbers of nests. Although there is conflicting evidence regarding female spacing patterns, it appears that female cowbirds have large home ranges that overlap extensively (Rothstein et al. 1986), as shown by patterns of multiple parasitism in this and other studies (e. g., Fleischer 1985).
The generalist strategy of cowbirds contrasts sharply with the host specificity exhibited by the Common Cuckoo. At a given locality cuckoos parasitize only a few species, and individual females are host-specific. Populations of cuckoos are segregated into races or genetic strains ("gentes"), with each "gens" laying eggs that are mimetic to a specific host. Egg mimicry is a consequence of most cuckoo hosts being rejectors (Rothstein 1990). Hosts not only reject non-mimetic eggs, but also reject non-synchronized eggs (Davies and Brooke 1988). As a consequence, female cuckoos have to invest time and energy in selecting a nest and parasitizing it at a certain time. This suggests that it would be advantageous for females to have exclusive use of a laying range. As with cowbirds, there is conflicting evidence regarding the spacing system of female cuckoos, but it appears that females have restricted and separate laying ranges, although there is overlap (Wyllie 1981). Because cuckoos have to invest more than cowbirds in carefully placing each egg, they have comparatively lower fecundity. Estimates of the number of eggs laid by cuckoos in Britain range between 10 and 25 (Rothstein 1990). This is higher than what nesting birds would produce, but not as impressive when compared with cowbird production.
Current theory on brood parasite-host coevolution suggests that parasites become specialized as hosts evolve defenses against parasitism (Rothstein





76


1990). It is assumed that cowbirds are recently evolved parasites that will eventually become host-specific (Rothstein 1990). Will cowbirds, given enough time, evolve cuckoo-like adaptations?
Davies and Brooke (1989b) proposed a model to explain events in an arms race between parasite and host (see also Rothstein 1990). When a host population begins to be parasitized, egg rejection is favored by selection. The rate of spread of the rejection response depends on its adaptive value, determined by impact and frequency of parasitism. As the rejection response spreads in the population, selection will favor egg mimetism. Several outcomes are possible after this stage. For example, egg discrimination may become more refined, or parasitism with mimetic eggs may become too frequent, driving the host to extinction. This model assumes that genetic variation for egg rejection is present in the host population (Davies and Brooke 1989b; Rothstein 1990).
According to this model, different parasites are at different stages in the sequence. Cuckoos are at an advanced stage. Most Common Cuckoo hosts show intermediate to high levels of rejection. As a consequence, female cuckoos lay mimetic eggs and specialize on a single host. Cowbirds are assumed to have evolved parasitism more recently than cuckoos, mainly because there are about 50 species of parasitic cuckoos and only five cowbirds (Rothstein 1990). Thus, cowbirds are at an early stage in the sequence of events leading to specialization. This hypothesis assumes that there has been no time for the spread of rejection in host populations. As rejection spreads, cowbirds should evolve egg mimicry and specialize on one or a few hosts (Harvey and Partridge 1988; Rothstein 1990).
Davies and Brooke's model assumes there is a single pathway for the
evolution of brood parasitism. There is a "progression" from a generalist to a





77


specialist strategy (Harvey and Partridge 1988; Rothstein 1990). This model implies that all avian brood parasites, despite their different origins, should converge on the same evolutionary path. Obligate interspecific brood parasitism has arisen independently in three orders of altricial birds: Cuculiformes, Piciformes and Passeriformes. In the Cuculidae, brood parasitism occurs in two different subfamilies, the Old World Cuculinae and the New World Neomorphinae (in the latter only three of 11 species are parasitic). Among passerines, it evolved independently in two subfamilies in different families: Icterinae (Emberizidae) and Viduinae (Ploceidae). Strategies of parasitism in these lineages differ in a variety of aspects, such as degree of host specificity (Payne 1977; Rothstein 1990).
Given the polyphiletic origin of brood parasitism, therefore, it is
conceivable that there are optional evolutionary routes, not necessarily following the generalist-specialist progression. Brood parasites may remain generalists, or start as specialists and remain specialists or become generalists. A recent phylogeny of the cowbirds, based on mitochondrial DNA (Lanyon 1992), suggests that cowbirds are monophiletic and that the specialist Screaming Cowbird (Molothrus rufoajxilaris. which only parasitizes the non-parasitic Bay-winged Cowbird, M. badius Fraga 1986) is primitive, while the generalist Shiny and Brown-headed Cowbirds are the most derived. This phylogeny, therefore, suggests a pathway opposite to the generalistspecialist progression.
Another recent phylogeny differs from Lanyon's and suggests even
different pathways. Freeman (in press) constructed a phylogeny of 47 species in the subfamily Icterinae, based on mitochondrial DNA. In this phylogeny, brood parasitism evolved independently at least three times. The specialist Screaming Cowbird is distantly related to the other cowbirds, suggesting that





78


the similarity in eggs and nestling plumage with its host, the Bay-winged Cowbird, is related to mimicry (Fraga 1986; Lanyon 1992). The Giant Cowbird (Scanidura oryzivora) also evolved in a non-parasitic lineage, independently of the three generalist species (Brown-headed, Shiny and Bronzed Cowbirds), which share a common ancestor.
Both Lanyon's and Freeman's results require cautious interpretation, because their phylogenies are simply the most parsimonious of a series of possible phylogenies (Freeman, pers. com.). They suggest, however, that multiple pathways are possible in brood parasitism and that being a specialist is not necessarily a derived condition.
There may even be different pathways for the evolution of egg mimicry. Davies and Brooke's (1989b) model assumes that the selection pressure for egg mimicry is host rejection. Brooker and Brooker (1989, 1990) suggested an alternative pathway for the evolution of egg mimicry and host specialization. Although female cuckoos tend to have non-overlapping laying areas, some overlap occurs, particularly in areas where more than one cuckoo species occur. There is evidence that many cuckoo species habitually remove an egg when laying their own. If a second cuckoo parasitizes a nest already containing a cuckoo egg, there is a chance she will remove the first cuckoo egg if it has contrasting colors (with the background or with host eggs). Thus, interspecific competition could explain the evolution of egg traits (mimicry, crypsis and small size) and host specialization in cuckoos (Brooker and Brooker 1990). The model can be extended to explain egg polymorphism in the Common Cuckoo by intraspecific competition (Brooker and Brooker 1990). While this model presents an alternative hypothesis for the evolution of egg mimicry, it still remains a fact that some cuckoo hosts accept mimetic eggs and reject non-mimetic eggs, and the role of their behavior in the evolution of





79


this trait can not be ignored. The two hypothesis are not mutually exclusive and may operate simultaneously.
The fact that a variety of factors may be selecting for brood parasitism to evolve egg mimicry and host specialization prompts the question, how can cowbirds remain generalists? If egg rejection by hosts is the main selective force conducing to host specialization, then the reproductive strategy of cowbirds depends on host responses. With a preponderance of acceptor species, the current strategy of parasitizing all nests found is optimum, because it may be impossible for cowbirds to recognize and avoid rejectors without also avoiding some suitable hosts (Rothstein 1976). The decrease in reproductive succes resulting from parasitizing a few rejectors is probably less than what would result if cowbirds fail to parasitize some suitable hosts (Rothstein 1976). The question then becomes, will cowbird hosts become rejectors? Ample evidence suggests that lack of rejection in some hosts is best explained by absence of genetic variants that exhibit a rejection response (Rothstein 1975a, 1982; this study). Thus, the system depends on stochastic processes (appearance of mutations or recombinations that code for rejection) that are impossible to predict (Rothstein 1982). As long as these variants do not appear, hosts will remain acceptors, allowing cowbirds to remain generalists.
Even if the genetic basis for rejection is present in a population, the costof-rejection hypothesis provides a scenario for hosts not expressing this behavior. One critical difference between cowbirds and cuckoos is the cost of parasitism for the host. For cuckoo hosts, the cost of parasitism is always very high because their losses are total, as cuckoo hatchlings eject host eggs from the nest (Wyllie 1981). Because cowbirds do not kill nestmates, costs for cowbird hosts are variable. For some hosts, such as House Wrens, costs may





80


be high, but for other hosts costs may be intermediate to low (Rothstein 1990). If the costs of accepting parasitism are lower then the costs of rejecting, hosts will continue to accept cowbird eggs, allowing cowbirds to continue parasitizing that population with non-mimmetic and nonsynchronized eggs.
In conclusion, it is likely that being a generalist is a successful strategy that can be maintained indefinitely, and not necessarily a stage in the evolution of host-specificity. Cowbirds direct their reproductive effort into egg production, rather than expending much energy in each egg. This strategy is possible because most cowbird hosts are acceptors (Rothstein 1990). Reproductive success of cowbirds, expressed in number of offspring produced per year, is a function of the probability of success of each egg multiplied by annual fecundity. If hosts become discriminative and reject non-synchronized eggs, cowbirds would be forced to expend more energy in ensuring a better placement of each egg. This could result in an increase in the probability of success of each egg, but cowbirds would have to lay fewer eggs. Differences between the strategies of cowbirds and cuckoos probably reflect different balances of this trade-off.












LITERATURE CITED

Alvarez, H., M. D. Heredia, and M. C. HernAndez. 1984. Reproducci6n del
Cucarachero Comin (Trog~odvtes aedon) en el Valle del Cauca.
Caldasia 14: 86-123.

Ankney, C. D. and D. M. Scott. 1980. Changes in nutrient reserves and diet of
Brown-headed Cowbirds. Auk 97: 684-696.

Balinsky, B. I. 1975. An introduction to embryology. Philadelphia: W. B.
Sanders Co.

Blankespoor, G. W., J. Oolman, and C. Uthe. 1982. Eggshell strength and
cowbird parasitism of Red-winged Blackbirds. Auk 99: 363-365.

Briskie, J. V. and S. G. Sealy. 1990. Evolution of short incubation periods in
the parasitic cowbirds, Molothrus spp. Auk 107: 789-794.

Brooker, L. C. and M. G. Brooker. 1990. Why are cuckoos host specific? Oikos
57: 301-309.

Brooker, M. G. and L. C. Brooker. 1989. The comparative breeding behavior
of two sympatric cuckoos, Horsfield's Bronze-cuckoo Chrvscoccx
basalis, and the Shining Bronze-cuckoo C. lucidus in western
Australia: A new model for the evolution of egg morphology and host
specificity in avian brood parasites. Ibis 131: 528-547.

Carey, C. 1986. Possible manipulation of eggshell conductance of host eggs by
Brown-headed Cowbirds. Condor 88: 388-390.

Carter, M. D. 1986. The parasitic behavior of the Bronzed Cowbird in south
Texas. Condor 88: 11-25.

Cavalcanti, R. B. and T. M. Pimentel. 1988. Shiny Cowbird parasitism in
central Brazil. Condor 90: 40-43.

Clark, K. L. and R. J. Robertson. 1981. Cowbird parasitism and the evolution
of anti- parasite strategies in the Yellow Warbler. Wilson Bull. 92: 244258.

Cruz, A., T. Manolis, and J. W. Wiley. 1985. The Shiny Cowbird: A brood
parasite expanding its range in the Caribbean region. pages 607-620 in P. A. Buckley et al. (eds.), Neotropical Ornithology. Washington, D. C.:
American Ornithologists' Union.


81





82


Darley, J. A. 1983. Territorial behavior of female Brown-headed Cowbirds.
Canadian Journal of Zoology 61: 65-69.

Davies, N. B. and M. L. Brooke. 1988. Cuckoos vs. reed warblers: Adaptations
and counteradaptations. Animal Behaviour 36: 262-284.

Davies, N. B. and M. L. Brooke. 1989a. An experimental study of coevolution
between the cuckoo Cuculus canorus and its hosts. I. Host egg
discrimination. Journal of Animal Ecology 58: 207-224.

Davies, N. B. and M. L. Brooke. 1989b. An experimental study of coevolution
between the cuckoo, Cuculus canorus, and its hosts. II. Host egg
markings, chick discrimination and general discussion. Journal of
Animal Ecology 58: 225-236.

Davis, D. E. 1942. The number of eggs laid by cowbirds. Condor 44: 10-12.

Dufty, A. M. 1983. Variation in the egg markings of the Brown-headed
Cowbird. Condor 85: 109-111.

Finke, E. O., D. Valencia, and W. D. McKay. 1979. Evaluaci6n de dafios por
pdjaros en cultivos de arroz en Colombia. Instituto Colombiano
Agropecuario, Palmira.

Fleischer, R. C. 1985. A new technique to identify and assess the dispersion of
eggs of individual brood parasites. Behavioral Ecology and Sociobiology
17: 91-99.

Fraga, R. M. 1978. The Rufous-collared Sparrow as a host of the Shiny
Cowbird. Wilson Bulletin 90: 271-284.

Fraga, R. M. 1983. Parasitismo de crfa del Renegrido Molothrus bonariensis
sobre el Chingolo Zonotrichia caensis: Nuevas observaciones y
conclusiones. Hornero 12 (Ndimero extraordinario): 245-255.
Fraga, R. M. 1985. Host-parasite interactions between Chalk-browed
Mockingbirds and Shiny Cowbirds. pages 829-844 in P. A. Buckley et
al. (eds.), Neotropical Ornithology. Washington D.C.: American
Ornithologists' Union.

Fraga, R. M. 1986. The Bay-winged Cowbird (Molothrus badius) and its brood
parasites: Interactions, coevolution and comparative efficiency. Ph. D.
Diss., University of California, Santa Barbara.

Freeman, B. M. and M. A. Vince. 1974. Development of the Avian Embryo.
London: Chapman and Hall.





83

Freeman, S. in press. The phylogeny of the blackbirds estimated from
restriction sites in mitochondrial DNA. Systematic Biology

Friedmann, H. 1927. A case of apparently adaptive acceleration of embryonic
growth rate in birds. Biological Bulletin 53: 343-345.

Friedmann, H. 1929. The Cowbirds, A Study in the Biology of Social
Parasitism. Springfield, Ill.: C. C. Thomas.

Friedmann, H. 1948. The parasitic cuckoos of Africa. Washington Academy of
Sciences Monograph 1: 1-204.

Friedmann, H. 1955. The honeyguides. US National Museum Bulletin 1-292.

Friedmann, H. 1960. The parasitic weaverbirds. US National Museum
Bulletin 1-196.

Friedmann, H. 1963. Host relations of the parasitic cowbirds. US National
Museum Bulletin

Friedmann, H. and L. F. Kiff. 1985. The parasitic cowbirds and their hosts.
Proceedings of the Western Foundation of Vertebrate Zoology 2: 227302.

Friedmann, H., L. F. Kiff, and S. I. Rothstein. 1977. A further contribution to
knowledge of host relations of the cowbirds. Smithsonian Contributions
to Zoology

Gochfeld, M. 1979a. Begging by nestling Shiny Cowbirds: Adaptive or
maladaptive? Living Bird 17: 40-50.

Gochfeld, M. 1979b. Brood parasite and host coevolution: Interaction between
Shiny Cowbirds and two species of meadowlarks. American Naturalist
113: 855-870.

Goldsmith, 0. 1774. A history of the earth and animated nature.

Gould, S. J. and E. S. Vrba. 1982. Exaptation: A missing term in the science
of form. Paleobiology 8: 4-15.

Hamilton, W. J. and G. H. Orians. 1965. Evolution of brood parasitism in
altricial birds. Condor 67: 361-382.

Harvey, P. H., and L. Partridge. 1988. Of cuckoo clocks and cowbirds. Nature
335: 586-587.

Hoar, W. S. and C. P. Hickman. 1983. A Laboratory Companion for General
and Comparative Physiology. Englewood Cliffs, N.J.: Prentice Hall.




84

Hoy, G. and J. Ottow. 1964. Biological and oological studies of the molothrine
cowbirds (Icteridae) of Argentina. Auk 82: 186-203.

Jackson, N. H. and D. D. Roby. 1992. Fecundity and egg-laying patterns of
captive yearling Brown-headed Cowbirds. Condor 94: 585-589.

Kelly, C. 1987. A model to explore the rate of spread of mimicry and rejection
in hypothetical populations of cuckoos and their hosts. Journal of
Theoretical Biology 125: 283-299.

King, J. R. 1973. Reproductive relationships of the Rufous-collared Sparrow
and the Shiny Cowbird. Auk 90: 19-34.

Klaas, E. E. 1975. Cowbird parasitism and nesting success in the eastern
phoebe. Occassional Papers of the Museum of Natural History of the
University of Kansas 41:1-18.

Lanyon, S. M. 1992. Interspecific brood parasitism in blackbirds (Icterinae): A
phylogenetic perspective. Science 255: 77-79.

Levey, D. J. and F. G. Stiles. in press. Birds of La Selva: Ecology, Behavior
and Taxonomic Affinities.

Marvil, R. E. and A. Cruz. 1989. Impact of Brown-headed Cowbird parasitism
on the reproductive success of the Solitary Vireo. Auk 106: 476-480.

Mason, P. 1985. The nesting biology of some passerines of Buenos Aires,
Argentina. in P. A. Buckley et al. (eds.), Neotropical Ornithology.
Washington, D. C.: American Ornithologists' Union.

Mason, P. 1986a. Brood parasitism in a host generalist, the Shiny Cowbird: I.
The quality of different species as a host. Auk 103: 52-60.

Mason, P. 1986b. Brood parasitism in a host generalist, the Shiny Cowbird:
II. Host selection. Auk 103: 61-69.

Mason, P. and S. I. Rothstein. 1986. Coevolution and avian brood parasitism:
Cowbird eggs show evolutionary response to host discrimination.
Evolution 40: 1207-1214.

McLean, I. 1987. Response to a dangerous enemy: Should a brood parasite be
mobbed? Ethology 75: 235-245.

Mendez de la Cruz, F., L. J. Guillette and M. Villagrfn Santa Cruz. In press.
Differential atresia of ovarian follicles and its effect on the clutch size
of two populations of the viviparous lizard Sceloporus mucronatus.
Functional Ecology.





85

Miller, A. H. 1961. Molt cycles in equatorial Andean sparrows. Condor 63:
143-161.

Montgomery, D. C. 1984. Design and Analysis of Experiments. New York:
John Wiley and Sons.

Nice, M. M. 1953. The question of ten-day incubation periods. Wilson Bulletin
65: 81-93.

Norman, R. F. and R. J. Robertson. 1975. Nest-searching behavior in the
Brown-headed Cowbird. Auk 92: 610-611.

Orians, G. H., E. Roskaft, and L. D. Beletzky. 1989. Do Brown-headed
Cowbirds lay their eggs at random in the nests of Red-winged
Blackbirds? Wilson Bulletin 101: 599-605.

Ortega, C. P. and A. Cruz. 1988. Mechanisms of egg acceptance by marshdwelling blackbirds. Condor 90: 349-358.

Payne, R. B. 1965. Clutch size and number of eggs laid by Brown-headed
Cowbirds. Condor 67: 44-61.

Payne, R. B. 1967. Gonadal responses of Brown-headed Cowbirds to long daylengths. Condor 69: 289-297.

Payne, R. B. 1972. Mechanisms and control of molt. pages 104-155 in D. S.
Farner and J. R. King (eds.), Avian Biology, Volume 2. New York:
Academic Press.

Payne, R. B. 1973. The breeding season of a parasitic bird, the Brown-headed
Cowbird, in central California. Condor 75: 80-99.

Payne, R. B. 1974. The evolution of clutch size and reproductive rates in
parasitic cuckoos. Evolution 28: 169-181.

Payne, R. B. 1976. The clutch size and numbers of eggs of Brown-headed
Cowbirds: Effects of latitude and breeding season. Condor 78: 337-342.

Petit, L. J. 1991. Adaptive tolerance of cowbird parasitism by Prothonotary
Warblers: A consequence of nest site limitation? Animal Behaviour 41:
425-532.

Phillips, J. G., P. J. Butler and P. J. Sharp. 1985. Physiological strategies in
avian biology. Glasgow: Blackie and Son, Ltd.

Post, W., A. Cruz and D. B. McNair. 1993. The North American invasion
pattern of the Shiny Cowbird. Journal of Field Ornithology 64: 32-41.





86

Post, W. and J. W. Wiley. 1977. Reproductive interactions of the Shiny
Cowbird and the Yellow-shouldered Blackbird. Condor 79: 176-184.

Poulin, B., G. Lefebvre and R. McNeil. 1992. Tropical avian phenology in
relation to abundance and exploitation of food resources. Ecology 73:
2295-2309.

Rahn, H. and A. Ar. 1974. The avian egg: Incubation time and water loss.
Condor 76: 147-152.

Rahn, H., C. V. Paganelli, and A. Ar. 1975. Relation of avian egg weight to
body weight. Auk 92: 750-765.

Ramo, C. and B. Busto. 1981. La reproducci6n de un ave parAsita: El tordomirlo (Molothrus bonariensis) en los Ilanos de Apure (Venezuela).
Dofiana Acta Vertebrata 8: 215-224.

Ricklefs, R. E. 1987. Comparative analysis of avian embryionic growth.
Journal of Experimental Zoology Supplement 1: 309-323.

Robertson, R. J. and R. F. Norman. 1977. The function and evolution of
aggressive behavior towards the Brown-headed Cowbird. Canadian
Journal of Zoology 55: 508-518.

Roff, D. A. 1992. The Evolution of Life Histories. Theory and Analysis. New
York: Chapman and Hall.

Rohwer, S. and C. D. Spaw. 1988. Evolutionary lag versus bill-size
constraints: A comparative study of the acceptance of cowbird eggs by
old hosts. Evolutionary Ecology 2: 27-36.

Rohwer, S., C. D. Spaw, and E. Roskaft. 1989. Costs to Northern Orioles of
puncture-ejecting parasitic cowbird eggs from their nests. Auk 106:
734-738.

Roskaft, E., G. H. Orians, and L. D. Beletsky. 1990. Why do Red-winged
Blackbirds accept eggs of Brown-headed Cowbirds? Evolutionary
Ecology 4: 35-42.

Rothstein, S. I. 1975a. Evolutionary rates and host defenses against avian
brood parasitism. American Naturalist 109: 161-176.

Rothstein, S. I. 1975b. An experimental and teleonomic investigation of avian
brood parasitism. Condor 77: 250-271.

Rothstein, S. I. 1976. Cowbird parasitism of the Cedar Waxwing and its
evolutionary implications. Auk 93: 498-509.





87

Rothstein, S. I. 1982. Successes and failures in avian egg recognition with
comments on the utility of optimality reasoning. American Zoologist
22: 547-560.

Rothstein, S. I. 1990. A model system for coevolution: Avian brood parasitism.
Annual Review Ecology Systematics 21: 481-508.

Rothstein, S. I., D. A. Yokel, and R. C. Fleischer. 1986. Social dominance,
mating and spacing systems, female fecundity, and vocal dialects in
captive and free-ranging Brown-headed Cowbirds. pages 127-185 in R.
F. Johnston (ed.), Current Ornithology. New York: Plenum Press.

Salvador, S. A. 1983. Parasitismo de cria del Renegrido (Molothrus
bonariensif) en Villa Maria, C6rdoba, Argentina. Historia Natural 3:
149-158.

Scott, D. M. 1978. Using sizes of unovulated follicles to estimate the laying
rate of the Brown-headed Cowbird. Canadian Journal of Zoology 56:
2230-2234.

Scott, D. M. and C. D. Ankney. 1979. Evaluation of a method for estimating
the laying rate of Brown-headed Cowbirds. Auk 96: 483-488.

Scott, D. M. and C. D. Ankney. 1980. Fecundity of the Brown-headed Cowbird
in southern Ontario. Auk 97: 677-683.

Scott, D. M. and C. D. Ankney. 1983. The laying cycle of Brown-headed
Cowbirds: Passerine chickens? Auk 100: 583-592.

Sealy, S. G. 1992. Removal of Yellow Warbler eggs in association with
cowbird parasitism. Condor 94: 40-54.

Seppa, J. 1969. The cuckoo's ability to find a nest where it can lay an egg.
Ornis Fennica 46: 78-79.

Sick, H. 1958. Notas biol6gicas s8bre o Gaud6rio Molothrus bonariensis
(Gmelin) (Aves, Icteridae). Revista Brasileira Biologia 18: 417-431.

Sinervo, B. 1990. The evolution of maternal investment in lizards: An
experimental and comparative analysis of egg size and its effects on
offspring performance. Evolution 44: 279-294.

Smith, J. N. M. 1981. Cowbird parasitism, host fitness and age of the host
female in an island Song Sparrow population. Condor 83: 152-161.

Teather, K. L. and R. J. Robertson. 1986. Pair bonds and factors influecing
the diversity of mating systems in Brown-headed Cowbirds. Condor 88:
63-69.




88

Thompson, C. F. and B. M. Gottfried. 1976. How do cowbirds find and select
nests to parasitize? Wilson Bulletin 88: 673-674.

Thompson, C. F. and B. M. Gottfried. 1981. Nest discovery and selection by
Brown-headed Cowbirds. Condor 83: 268-269.

Vleck, C. M. and D. Vleck. 1987. Metabolism and energetics of avian embryos.
Journal of Experimental Zoology Supplement 1: 111-125.

Vleck, D., C. M. Vleck, and D. F. Hoyt. 1980. Patterns of metabolism and
growth in avian embryos. American Zoologist 20: 405-416.

Weatherhead, P. J. 1989. Sex-ratios, host-specific reproductive success and
impact of Brown-headed Cowbirds. Auk 106: 358-366.

Wiley, J. W. 1985. Shiny Cowbird parasitism in two avian communities in
Puerto Rico. Condor 87: 167-176.

Wiley, J. W. 1988. Host selection by the Shiny Cowbird. Condor 90: 289-303.

Wiley, R. H. and M. S. Wiley. 1980. Spacing and timing in the nesting ecology
of a tropical blackbird: Comparison of populations in different
environments. Ecological Monographs 50: 153-178.

Woodward, P. W. and J. C. Woodward. 1979. Brown-headed Cowbird
parasitism on Eastern Bluebirds. Wilson Bulletin 91: 321-322.

Worthington, A. 1982. Population sizes and breeding rythms of two species of
manakins in relation to food supply. pages 213-225 in E. G. Leigh, A. S.
Rand and D. M. Windsor (eds.), The Ecology of a Tropical Forest:
Seasonal Rythms and Long Term Changes. Washington, D.C.:
Smithsonian Institution Press.

Wyllie, I. 1981. The Cuckoo. New York: Universe Books.












BIOGRAPHICAL SKETCH

Gustavo Kattan was born on 26 July 1953 in Cali, Colombia. He attended the Universidad del Valle, where he graduated as a zoologist in 1983, working under the direction of Dr. Humberto Alvarez. In 1984 he married his wife Carolina Murcia, a fellow biologist, and started graduate school at the University of Florida. He obtained his Master's degree in 1987 under the direction of Dr. Harvey Lillywhite and continued in the Ph. D. program with Dr. Lou Guillette as his advisor. His research focuses on the reproductive biology of neotropical vertebrates, from physiological and behavioral pespectives. He is also interested in problems of rarity and vulnerability of species to extinction.























89













I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Los J. Gu ett chairman
Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Richard A. Kilie
Associate Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Douglas J.' Levef
Assistant Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Jfhn W. Fitzsitrick
Professor forest Resources
/ and Conservation









I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Scott K. Robinson
Associate Professor of Ecology University of Illinois


This dissertation was submitted to the Graduate Faculty of the
Department of Zoology in the College of Liberal Arts and Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

May 1993


Dean, Graduate School




Full Text

PAGE 1

REPRODUCTIVE STRATEGY OF A GENERALIST BROOD PARASITE, THE SHINY COWBIRD, IN THE CAUCA VALLEY, COLOMBIA By GUSTAVO H. KATTAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1993

PAGE 2

ACKNOWLEDGMENTS The culmination of this dissertation has involved the contribution of many people and institutions. First I would like to thank the members of my doctoral committee, Drs. Louis Guillette (chairman), Douglas Levey, Richard Kiltie, Scott Robinson, John Fitzpatrick, and formerly Martha Cnmip, for their support and advice during all phases of the research. My friend and former advisor, Dr. Humberto Alvarez, put me in the tracks of tropical birds, and generously shared with me his experience with House Wrens and Shiny Cowbirds. I enjoyed the many hours we spent discussing cowbirds over countless cups of coffee. I owe a great debt to my field assistant, Natalia G6mez. Her dedication and skills were crucial for the success of the field work. She always did more than I asked and contributed many good ideas. Discussions with Drs. Stephen Rothstein, Spencer Sealy and Alfred Dufty have contributed substantially to my understanding of brood parasitism in birds. Financial support for this research was provided by the Department of Zoology, University of Florida, the Frank Chapman Memorial Fund of the American Museum of Natural History, the Ftmdacion para la Promocidn de la Investigaci6n y la Tecnologia, Banco de la Republica, Bogota, Colombia, the Fondo de Investigaciones Cientificas "Francisco Jos6 de Caldas" (COLCIENCIAS), Bogota, Colombia, and the Instituto Vallecaucano de Investigaciones Cientificas (INCIVA), CaH, Colombia. I would like to thank the staff of INCIVA, especially Mrs. Belly Narvaez, for logistical support, and • « u

PAGE 3

don Alfonso Madrindn for permission to work on his property. Jos6 Kattan and Walter Buttkus helped me build more than 140 nest boxes, with wood donated by don Hernando Murcia. Manuel Giraldo provided company and laboratory and field assistance and helped with logistical support at the Universidad del Valle. Last but not least, I wotdd like to express my gratitude to my wife, Carolina Murcia, for being there and for putting up with my constant cowbird talk during the last five years! iii

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS ii ABSTRACT vi CHAPTERS 1 INTRODUCTION 1 2 STUDY AREA AND GENERAL METHODS 6 3 REPRODUCTIVE SEASONALITY AND FECUNDITY OF SHINY COWBIRDS IN THE CAUCA VALLEY 8 Introduction 8 Methods 10 Results 13 Discussion 18 4 HOST NEST SELECTION 22 Introduction 22 Methods 23 Results 24 Discussion 31 5 MECHANISMS OF SHORT INCUBATION PERIOD IN THE SHINY COWBIRD 35 Introduction 35 Methods 40 Results 42 Discussion 47 6 IMPACT OF COWBIRD PARASITISM: WHY DO HOUSE WRENS ACCEPT PARASITIC EGGS? 55 Introduction 55 Methods 57 Results 59 Discussion 67 iv

PAGE 5

7 GENERAL DISCUSSION 73 LITERATURE CITED 81 BIOGRAPHICAL SKETCH 89 V

PAGE 6

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy REPRODUCTIVE STRATEGY OF A GENERALIST BROOD PARASITE, THE SHINY COWBIRD, IN THE CAUCA VALLEY, COLOMBIA By Gustavo H. Kattan May 1993 Chairman: Dr. Louis J. Guillette Major Department: Zoology The reproductive strategy of a brood parasite involves "decisions" on life history traits such as how many eggs to produce and how much to invest in each egg. Parasites may produce a large number of eggs and parasitize as many nests as they can, with little consideration of the circumstances in which each egg is laid (the "shotgun" strategy). Host responses, however, may modulate the parasite's strategy. If hosts reject eggs that are non-mimetic and/or not synchronized with their own laying period, parasites may be forced to spend energy in selecting suitable hosts and closely monitor nests to select the most appropriate time for egg laying. These conditions would result in parasites being able to lay fewer eggs. I studied these aspects of the interaction between Shiny Cowbirds (Molothrua bonariensis ) and House Wrens (Troglodytgg aedon ), their main host in the Cauca Valley, Colombia. Cowbirds in the Cauca Valley are extremely fecund birds. A conservative estimate gave a fecimdity of 120 eggs per year, equivalent to about 9.2 times the female's body mass. This extraordinary fecundity results from cowbirds vi

PAGE 7

laying almost continuously, without regressing the ovary, during the entire 9month breeding season. As predicted by life history theory, the high fecundity exhibited by cowbirds involves a trade-off with the energy invested in each egg. I divided investment in two components: A direct investment represented in the amount of energy deposited in the egg and a behavioral investment represented in the process of nest selection. Cowbirds lay eggs with an energy content 24% lower than would be expected from egg mass. Presumably, the energy saved allows cowbirds to lay more eggs than would be possible if each had a higher energy content. A reduced energy content also results in a short incubation period, which gives cowbirds an advantage over their nestmates. Cowbirds also invest little energy in the careful placing of each egg in circumstances that woiold maximize its chances of being accepted and incubated successfully by the host. Cowbirds monitor host nests and to some extent choose nests that are at an appropriate stage for parasitism. However, they do not synchronize egg laying with the wren's laying period, and do not avoid multiply parasitized nests. This strategy is possible because wrens do not reject cowbird eggs. Experiments with artificial eggs indicated that wrens do not recognize cowbird eggs as foreign, neither do they reject unsynchronized eggs. Therefore, cowbirds in the Cauca Valley follow the "shotgun" strategy. Instead of carefully placing a few eggs, they direct their energy into egg production and permanently search for nests, parasitizing as many as they can. I discuss the implications of this strategy for the evolution of host specificity in brood parasites. vii

PAGE 8

CHAPTER 1 INTRODUCTION Avian brood parasites lay their eggs in the nests of other bird species; the host species incubates the eggs and raises the young. Obligate interspecific brood parasitism occurs in about 1% of the world's bird species (about 85 species), including 47 species of cuckoos (Cuculidae) in the Old World (Cuculinae) and three species in America (Neomorphinae), the honeyguides (Indicatoridae) of Africa and Asia, the widowbirds (Viduinae) and the Cuckooweaver (Anomalospiza imberbis ) of Africa, and five species of cowbirds (Icterinae, Emberizidae) in America. The unorthodox breeding habit of brood parasites has intrigued naturalists since ancient times. The parasitic habit of the Common Cuckoo (Cuculus canorus) . for example, is known since Aristotle's time (Wyllie 1981) and has been featured extensively in natural history works since then (e. g., Goldsmith 1774). In the New World, the breeding habits of Brown-headed Cowbirds ( Molothrus ater) in North America and Shiny and Screaming Cowbirds (Mbonariensis and Mrufoaxilaris ) in South America were described in several XIX century natural history works (Hudson 1920; Peattie 1939). From a more modem perspective, work on brood parasites was pioneered by H. Friedmann, who wrote extensively on cowbirds, cuckoos, weaverbirds and honeyguides (Friedmann 1929, 1948, 1955, 1960). Since then, the literature on brood parasitism has exploded (see 1

PAGE 9

2 reviews in Hamilton and Orians 1965; Payne 1977; Rothstein et al. 1986; Rothstein 1990). Since Hamilton and Orians (1965) placed brood parasitism within a natural selection paradigm, knowledge of the most significant evolutionary questions in this field has progressed enormously. In particular, coevolutionary aspects of the interaction between parasite and host have received much attention (Rothstein 1990). Because parasitism usually has a negative impact on the host, hosts evolve defenses. Parasites, in tvim, evolve counterdefenses, resulting in an evolutionary "arms race" (Davies and Brooke 1989b; Rothstein 1990). The interaction between the two species results in the adoption of strategies of host selection by parasites. Some parasites, such as Brown-headed and Shiny Cowbirds, are extreme generalists, parasitizing most passerines with which they occur (Friedmann and KifT 1985; Rothstein 1990). Others, such as cuckoos, specialize on only one or a few host species (although for species with wide geographic distributions the total host list may be long; Wyllie 1981; Rothstein 1990). Because brood parasites do not provide parental care, they can be viewed as natiiral experiments on the effects of parental care on life history traits (Payne 1977). Life history theory predicts a trade-off between the number of offspring produced and the energy invested in each offspring (Roflf 1992). Organisms that produce large numbers of offspring invest little in each offspring, and vice-versa. Therefore, brood parasites are expected to exhibit higher fecimdity than their nesting relatives. Payne (1977) reviewed reproductive strategies of brood parasites and foimd that they lay more eggs than non-parasitic species. Payne (1974, 1977) suggested that the small egg size of parasitic cuckoos may be due to allocation of energy to a large number

PAGE 10

of eggs. In the 16 years since Pajnie's (1977) review, however, no studies have followed up on these ideas. In particular, the trade-off between fecundity and investment has not been investigated further. Recent estimates of Brownheaded Cowbird fecimdity indicate that they lay large nimibers of eggs (Scott and Ankney 1983; Rothstein et al. 1986), but egg size is not smaller than the expected from the adult female body size (Briskie and Sealy 1990), as is the case in cuckoos. Is the increased fecundity of cowbirds simply the result of a reallocation of the energy not spent in parental care, or is the allocation of energy to each egg in any way compromised in cowbirds? In this study I test the h3T)othesis that Shiny Cowbirds direct their reproductive effort into egg production at the cost of reducing the energy invested in each egg. I studied Shiny Cowbirds in the Cauca Valley, in southwestern Colombia. This cowbird is widely distributed in South America (Friedmann and Kiff 1985) and has recently invaded the Antilles and the southeastern United States (Cruz et al. 1985; Post et al. 1993). It is an extreme generaUst, reportedly parasitizing more than 200 species of birds throughout its range (Friedmann and Kiff 1985). Work on Shiny Cowbirds has concentrated mostly on identifying local host assemblages (Ramo and Busto 1981; Salvador 1983; Wiley 1985; Mason 1986a, 1986b; Cavalcanti and Pimentel 1988). Detailed studies of host-parasite interactions are limited to a few species (e, g., Gochfeld 1979b; King 1973; Post and Wiley 1977; Fraga 1983, 1985). I begin by examining the seasonaHty of reproduction and estimating annual fecvmdity in this tropical population of cowbirds (Chapter 3). The hypothesis is that brood parasitism involves a trade-off: emancipation from parental care allows birds to devote extra energy to increase their fecundity (Payne 1977; Roff 1992). Work on Brown-headed Cowbirds in North America has shown that females have a very high fecundity, producing eggs almost

PAGE 11

4 continuously during the breeding season (Scott and Ankney 1983; Rothstein et al. 1986). Because the Shiny Cowbird's main host in the Cauca Valley, the House Wren (Troglodytes aedon) , has an extended 9-month breeding season (Alvarez et al. 1984), it raises the question, what is the reproductive seasonality and fecimdity of cowbirds, both at the population and individual levels? Once I determine how many eggs cowbirds lay, I examine the question of how much cowbirds invest behaviorally and energetically in each egg. Behavioral investment in each egg involves selection of host nests in order to maximize the chances of the parasitic egg being accepted and incubated successfully by the host. In Chapter 4 I explore aspects of nest selection and test the h)T)othesis that cowbirds follow host activities in order to time egg laying with the host's laying period. Next (Chapter 5) I examine the energetic investment in each egg and its implications as an adaptation for a short incubation period. Once the parasitic egg is laid, it is abandoned by the cowbird and nestlings have to compete with foster nestmates for parental resources. Thus, early hatching gives parasitic young a head start. It has long been recognized that brood parasites have short incubation periods (Hamilton and Orians 1965; Payne 1977), but precise measvirements of incubation periods are lacking because of the inherent variability associated with intermitent incubation in the field (Briskie and Sealy 1990). In Chapter 5 I present measurements of incubation period of Shiny Cowbird eggs under constant conditions in the laboratory, showing that the incubation period is shorter than expected from egg size. I suggest that a short incubation period is the result of a reduced egg energy content and discuss its impHcations for the cowbird's reproductive strategy. Next I examine impact of parasitism and explore the question of why wrens accept cowbird eggs (Chapter 6). My rationale is that the impact of

PAGE 12

5 parasitism may determine host responses, which in turn may modulate the cowbird strategy. If hosts are selective, i. e., they reject non-mimetic and nonsynchronized eggs, they may "force" the parasite to specialize. Unselective hosts, in contrast, allow parasites to be generalists. I end (Chapter 7) by discussing how all these pieces can be put together to show that the Shiny Cowbird strategy is that of an extreme generalist, and I discuss the evolution of host specificity in brood parasites.

PAGE 13

CHAPTER 2 STUDY AREA AND GENERAL METHODS The study was conducted between January June 1988 and March 1989 April 1991 at a 116 ha dairy farm located 15 km south of the city of Cali, in the Cauca Valley (1000 m elevation), in southwestern Colombia. The study area consisted of well-shaded pastures and farm buildings and was surrounded by fields of rice and sorghum, two crops on which cowbirds feed (Finke et al. 1979). There were a few large trees of ceiba (Ceiba pentandra) and samdn (Pithecellobium saman) . but most of the shade trees in pastures and along fence rows were chiminango (Pithecellobiimi dulce ), gu^simo (Guazuma ulmifolia ) and tachuelo (Fa gara rhoifolia ). Rainfall in the area is bimodal, with peaks of precipitation in April and October, and two dry periods, a mild one in December January, and a harsher one in July August. At the beginning of the study I mapped 14 wren territories, which were located mostly along fence rows and around shade trees and farm buildings. Wrens nested in a variety of natural and semi-natural cavities such as holes in bamboo (Guadua an gustifoUa) fence posts and other man-made structures, bromeliads and on the groimd in dense clumps of tall grass. To faciUtate data collection I placed wood nest boxes (10 x 10 x 15 cm) in all wren territories. During the first eight months of the study I placed a total of 140 boxes, with at least 6 boxes in each mapped territory. Wrens readily used these boxes and this study is based on 34 natural nests and 105 nests in boxes. Most breeding 6

PAGE 14

7 wrens were trapped in nest boxes or captured with mist nets and banded. I also banded 37 Shiny Cowbirds that were captured with mist nests. Wrens were the only host of cowbirds in the study area. All cowbird fledglings produced during the study were raised by wrens. No other species was parasitized, although nests of some potential hosts, such as Pied WaterTyrants (Fluvicola pica ; Ramo and Busto 1981), were available. Other hosts used by cowbirds elsewhere in the Cauca Valley include Black-billed Thrushes ( Turdus i gnobilis ), Blue-gray Tanagers (Thraupis episcopus ) and Crimson-backed Tanagers ( Ramphocelus dimidiatus ). In the mountains surrounding the valley (1200-2000 m), the main host is the Rufous-collared Sparrow (Zonotrichia capensis ). Wrens and cowbirds have coexisted in the valley for at least 100 years and probably for several centuries. The earliest ornithological record of both species is from 1911 (Chapman 1917), but the Cauca Valley has been settled since pre-Columbian time. In the 18th century there was already extensive agriculture and cattle ranching by European settlers. These activities probably provided adequate habitat for both wrens and cowbirds. The general methodology of the study consisted in keeping track of the breeding activities of all wrens. I made observations during all stages of the nesting cycle, from nest building to fledging. Nests were also observed during nest building and egg laying to record cowbird activities. Methods specific to each section of the study are reported in the corresponding chapter.

PAGE 15

CHAPTER 3 REPRODUCTIVE SEASONALITY AND FECUNDITY OF SHINY COWBIRDS IN THE CAUCA VALLEY Introduction Life history theory predicts a trade-off between number and quality of offspring (Roff 1992). Organisms that produce large numbers of offspring invest little in each offspring, and vice-versa. Parental care is an investment in the quality of offspring, increasing their chances of survival. Because brood parasites do not invest in parental care, they can be viewed as natural experiments on the trade-off between parental care and fecundity (Payne 1977; Roff 1992). Brood parasites, however, do not abandon their eggs at random. Their offspring require parental care, and the parasitic bird must invest some energy in laying the egg at a place and time that maximize chances of survival (Chapter 4). Emancipation from parental care, nevertheless, makes extra energy available and thus brood parasites are expected to exhibit higher fecundity than other species. Indeed, estimates of Brown-headed Cowbird annual fecundity in the north temperate zone have shown that they are extraordinarily fecund, laying in a pattern that resembles that of domestic chickens (Scott and Ankney 1983). In this chapter I present estimates of annual fecundity (number of eggs produced per year) at a tropical locality in which cowbirds have an extended breeding season. The first estimates of the number of eggs laid by cowbirds were based on searching for parasitized nests in the field (reviewed in Payne 1977; Rothstein et al. 1986). These estimates relied on the assumption that females 8

PAGE 16

9 lay individually distinctive eggs. While there is evidence that individual females always lay similar-looking eggs (Dufty 1983; Fleischer 1985), this is an imrehable method to estimate fecundity for several reasons (Rothstein et al. 1986). First, some host nests may not be found, or cowbird eggs may be removed by hosts or predators before being found. Second, laying ranges of individual females may not be known precisely. Third, laying ranges of several females may overlap and egg appearance may not be distinctive enough to differentiate females. More recent estimates of Brown-headed Cowbird fecundity have relied on the examination of ovaries. Payne (1965, 1973) examined postovulatory follicles and found that cowbirds lay eggs in sets or clutches, separated by days of non-la5dng. There was great interindividual variation, with sets varying between 1-6 eggs. Nearly all females laid throughout the breeding season, with the total ntraiber of eggs produced by a female depending on the length of the season. Payne (1976) obtained estimates varying from 11 to 25 eggs in different localities in the U. S. Extensive studies by Scott and Ankney (Scott 1978; Scott and Ankney 1980, 1983) obtained a higher estimate of 40 eggs per season in southern Ontario. They confirmed Payne's finding that cowbirds lay in sets of 1-7 eggs, with no regression of the ovary and oviduct until the end of the season. They also showed that individual cowbirds lay on about 70-80% of the days of the breeding season. Most estimates of cowbird fecundity have been done on Brown-headed Cowbirds in the north temperate zone, where the breeding season is restricted to a few months of the year. Most detailed studies of Shiny Cowbirds have been conducted in the south temperate zone, where there is also marked seasonality (e. g., Fraga 1985; Mason 1985, 1986a). The only estimate of Shiny Cowbird fecimdity based on examination of ovaries (Davis

PAGE 17

10 1942) is based on a small sample and gave a small estimate. While breeding of cowbirds and other birds in the tropics may be seasonal and related to alternation of wet-dry seasons (e. g., Wiley 1988), some tropical birds may have extended breeding seasons. House Wrens, for example, breed almost year-roimd in the Cauca Valley (Alvarez et al. 1984). Because wrens are frequently parasitized by cowbirds (Alvarez et al. 1984; pers. obs.), it raises the question, what is the reproductive seasonality and fecimdity of Shiny Cowbirds at the population and individual levels? Here I present data on the reproductive seasonality of a population of Shiny Cowbirds in the Cauca Valley and, based on examination of ovaries, estimate seasonality and feciindity of individual females. Methods Reproductive seasonality of the House Wren and Shiny Cowbird populations was determined during January June 1988 and March 1989 December 1990. Nests were checked at intervals of 1-3 days and eggs were individually marked with permanent ink. For each month I compiled the proportion of parasitized nests and the total number of cowbird eggs laid. Cowbird fecundity was determined from inspection of ovaries. Monthly samples of females were collected at a rice mill about 5 km SW of the main study area, during the months of November 1990April 1991. No females were collected at the main study area. Cowbirds exhibited a diurnal cycle of social behavior in which females were solitary in the morning, when they were searching for host nests, and gregarious in the afternoon, when they joined feeding flocks (see also Rothstein et al. 1986). Very few females were observed at the mill in the morning. Flocks started arriving between 11:30

PAGE 18

11 and 12:30 and were composed of both juveniles and adults. I captured cowbirds with mist nets when they entered the mill to feed on rice. This method ensured no bias in capturing females in different reproductive conditions (Scott and Ankney 1979). Only females in adult plxmiage were collected. Females were dissected and the fresh ovaries examined under a dissecting scope. In reproductively active females there was usually a series of yellow vitellogenic follicles (i. e., follicles in which yolk is being deposited) of graded sizes and numerous small white follicles. Postovulatory follicles (i. e., the scars left in the ovary after ovTilation) were also found in graded series, as the folUcle is gradually reabsorbed. Diameters of all vitellogenic follicles and maximimi width of postovulatory follicles were measured with calipers. Only yellow follicles were counted as vitellogenic. Follicle sizes were then plotted as a fimction of size rank for each female, number 1 being the largest irrespective of absolute size (Fig. 3-1). Assuming that Shiny Cowbirds ovulate daily, as foimd for Brownheaded Cowbirds (Scott and Ankney 1983; Jackson and Roby 1992), the laying history of each cowbird a few days before and after collection can be estimated from follicle sizes. Vitellogenic follicle sizes produced a graded series, with D follicles (size range 9.1-12.0 mm) being ready to ovulate (Fig. 31). C folUcles (size range 5.5-9.0 mm) would be ovulated the next day, and so on (assuming no atresia). Most follicles smaller than 2 mm were white and it was not possible to determine if they were initiating vitellogenesis. Thus, this method allowed determination of future ovulations for only four days after collection. A similar rationale was used for postovulatory follicles. Size E postovulatory follicles were found only when there was an egg in the oviduct. Postovtdatory folUcles regressed rapidly and could be identified with confidence only up to three days after ovulation (Fig. 3-1).

PAGE 19

12 OVULATION V4 V3 V2 VI PI P2 P3 VITELLOGENIC POSTOVULATORY FOLLICLES FOLLICLES Figure 3-1. Sequence of growth of vitellogenic follicles and regression of postovulatory follicles in cowbird ovaries. VI is the largest vitellogenic follicle, V2 the next largest, and so on. The same applies to postovulatory follicles. A follicle in the D size range is ready to be ovulated, with folhcles in the C, B and A size ranges to be ovulated on consecutive days. A size E postovulatory follicle corresponds to a recently ovulated egg, and F and G indicate folHcles regressed for one and two days, respectively.

PAGE 20

13 By analyzing the sequence of follicle sizes, I estimated set size and interval between sets of eggs for each female. The non-laying interval was estimated by gaps in the sequence of follicles. For example, a female with an F postovulatory follicle, a C vitellogenic follicle and no egg in the oviduct, has a gap of two days. Set size was estimated by coimting the longest possible sequence of consecutive daily ovulations. Because of the limitations in identifying follicles macroscopically, this method provides only a minimiam estimate of set size. Wrens bred in all months of the year except during the dry period of July August. There were peaks of nesting activity that coincided with the rainy seasons of April-May and October-November, with the first peak higher than the second (Fig. 3-2). Nesting activity was low during the JulySeptember dry season, with a few nests being initiated in early July and late September. Cowbird breeding activity paralleled wren nesting. Nests were psirasitized all months in which there were nests available (Fig. 3-2). These data indicate that the cowbird population breeds continuously during the period October-June, with very little or no reproductive activity during the July-September dry season. I examined 95 ovaries to determine reproductive status and estimate fecundity. Reproductively active females always had vitellogenic follicles and large oviducts (length=120-150 mm). Non-reproductive females had regressed ovaries and oviducts (length<50 mm). Assuming that cowbirds lay an egg every day, the pattern of graded follicle sizes (Fig. 3-1) indicated that Shiny Cowbirds laid eggs in sets. Set size varied between 1-6 eggs with a

PAGE 21

14 Figure 3-2. Reproductive seasonality of the House Wren and Shiny Cowbird populations in the Cauca Valley. Open sjonbols indicate number of wren nests with eggs and closed symbols indicate number of parasitized nests.

PAGE 22

15 mean of 3.2 (SD=1.4, n=35). This is a minimum estimate because my method does not allow detection of sequences longer than 7 days, although the frequency distribution of sets suggests that this is a precise estimate, as there were few ovaries with sets of 5-6 eggs (Fig. 3-3). Gaps in the size sequence indicated that non-laying intervals varied between 1-5 days, with a mean of 1.64 (SD=0.9, n=28). Modal gap size was one (Fig. 3-3). From November 1990 to April 1991, between 55% and 90% of the females collected each month were reproductively active (Fig. 3-4). All nonreproductive females were molting, suggesting that molting and breeding are mutually exclusive activities. Diiring August females were neither molting nor breeding, suggesting that cowbirds interrupt all energetically demanding activities during the dry season (June August). Reproductive females that were in the non-laying interval when collected (i. e., had just finished ovulating a series and follicles were developing for the next series) had fully developed oviducts. This pattern suggests that individual females are inactive during the 3-month dry season, but lay eggs continuously during the 9-month breeding season, interrupting only for molting. Duration of molting in Shiny Cowbirds is imknown. However, for a variety of tropical passerine birds with nonoverlapping breeding and molting seasons, duration of the molting season for the population varies between 2-5 months and for individual birds may be 2-3 months (Miller 1961; Worthington 1982; Poulin et al. 1992; Levey and Stiles, in press). Thus, assuming that a cowbird molts once during the 9-month season, and using a conservative estimate of three months for duration of molting, a female cowbird in the Cauca Valley lays eggs during six months of the year. If each female has an average daily laying rate of 0.66 eggs (series of 3.2 eggs separated by 1.64 days), her annual fecimdity is approximately 120 eggs.

PAGE 23

20-1 ^ 181 2 3 4 5 6 SET SIZE Figure 3-3. Frequency distribution of set size (number of consecutive ovxolations) and gap size (non-laying intervals) in cowbird ovaries.

PAGE 24

17 Figure 3-4. Percent of reproductively active female cowbirds in each monthly sample. Numbers above bars are sample sizes. All non-reproductive females were molting in all months except August, when cowbirds were neither breeding nor molting.

PAGE 25

Discussion The results indicate that Shiny Cowbirds in the Cauca Valley are extraordinarily fecund, laying approximately 120 eggs per year. This high fecvindity is the result of two factors. First, cowbirds lay almost continuously during the breeding season. Second, the breeding season at this locality is exceedingly long. Cowbird breeding seasons always coincide with their hosts'. In the north temperate zone, breeding seasons of all birds are determined by strong seasonality, with northern localities having shorter seasons than more southern localities. In Michigan, for example, the breeding season of the Brown-headed Cowbird and its hosts lasts 5-6 weeks, whereas in Oklahoma it lasts 8-10 weeks (Payne 1976). In Ontario, the breeding season of Brownheaded Cowbirds is restricted to about 8 weeks in May-June (Scott and Ankney 1980). Breeding seasons are similarly restricted to a few months of the year in the southern temperate zone. In northern Argentina most passerines, including Shiny Cowbirds, breed during the months of OctoberFebruary, corresponding to the austral spring and summer (Fraga 1985; Mason 1985, 1986a). The environmental signal that mediates gonadal development in the temperate zone is photoperiod (Payne 1967, 1977). Environmental signals in the tropics are less obvious. Breeding seasonality in tropical birds is usually related to alternation of dry and wet seasons, and in most localities there tend to be more or less well defined breeding seasons. In Puerto Rico, for example, most passerines breed during the rainy season of March-June; although some cowbird hosts breed year-roimd, the laying period of the Shiny Cowbird is restricted to the months of MarchJuly (Wiley 1988). In the Cauca

PAGE 26

19 Valley, there is a well-defined breeding season that coincides with the MarchJune rainy season (N. Gomez and G. Kattan, unpublished data). Some species, however, have breeding seasons that may extend up to 9 months. House Wrens in the Valley breed almost year-round, interrupting reproductive activities only during the driest period of July-September (Alvarez et al. 1984; this study). Here the Shiny Cowbird population parallels its host's extended breeding season. As a consequence of the extended breeding season Shiny Cowbirds lay an enormous number of eggs annualy. The extremely high fecxmdity of cowbirds in the Cauca Valley may be the result of a superabundance of food, as they feed on commercial crops such as rice and sorghum (Pinke et al. 1979). My minimimi estimate of a daily laying rate of 0.66 eggs is slightly lower than the 0.72-0.80 value estimated for Brown-headed Cowbirds in North America (Scott and Ankney 1979; Rothstein et al. 1986). Most adult females probably lay throughout the season in the temperate zone. Captive Brown-headed Cowbirds had laying rates between 0.18 and 0.91 eggs/day, with some females laying almost continuously during a 68-day period (Jackson and Roby 1992). While it is difficult to know the laying schedule of an individual female without actually following her throughout the breeding season, my data suggest that individual females may be reproductively active during at least two thirds of the 9-month breeding season. As has been foimd for Brown-headed Cowbirds in previous studies, examination of Shiny Cowbird ovaries indicated that eggs are laid in clutches or sets, separated by non-laying days. The significance of these sets is not clear. Laying patterns of captive yearling Brown-headed Cowbirds are highly variable, both within and between individuals, and do not support the notion that cowbirds lay in clutches (Jackson and Roby 1992). It has been suggested

PAGE 27

20 that non-laying intervals may result from follicular atresia related to physiological constraints and/or availability of host nests (Payne 1965, 1976; Scott and Ankney 1983; Rothstein et al. 1986). The egg laying pattern of cowbirds, however, resembles that of domestic chickens that are not allowed to brood (Scott and Ankney 1983). Domestic chickens lay eggs in sequences of daily ovulations separated by non-laying days (Phillips et al. 1985). This pattern of laying is the result of the existence of an "open period" in the ovulatory cycle. The open period is a restricted period of the day during which pre-ovulatory releases of luteinizing hormone (LH) occur. The LH surge (and the subsequent ovulation) first occurs at the beginning of the open period. Subsequent LH surges occur a little later each day, until they occur at the end of the open period, ending the sequence of ovulations (Phillips et al. 1985). A similar mechanism could be operating in cowbirds. This does not rule out the possibility, however, that there are waves of follicular recruitment, but the actual number of eggs laid is determined by atresia, as shown for the lizard Sceloporus mucronatus (Mendez de la Cruz et al., in press). The high laying rate of cowbirds raises questions related to nutritional constraints. Shiny Cowbirds in the Cauca Valley lay eggs with a mean mass of 4.2 g and an energy content of 14.54 kJ (Chapter 6). Thus, a 55 g female cowbird may be processing a total annual egg mass about 9.2 times her own body mass (504 g), with a total energy content of 1745 kJ. Ankney and Scott (1980) assayed Brown-headed Cowbirds for physiological strain related to the high laying rate, and foxmd no indication of major depletions attributable to egg laying. Rothstein et al. (1986), however, argue that the costs of laying may not be as small as Ankney and Scott (1980) suggested. Besides, certain nutrients such as calcium and iron may impose limitations on the capacity to

PAGE 28

21 produce eggs. Rothstein et al. (1986) found that la5dng females had decreased hematocrit values, suggesting that iron may be a limiting nutrient. The fact that I found no overlap between molting and reproduction in Shiny Cowbirds suggests that these two activities are energetically incompatible (Pa3nie 1972). The question of physiological strain in laying female cowbirds is still unresolved and requires further investigation. Even with the imprecision of methods available to estimate cowbird feciindity, it is clear that cowbirds are extremely fecund birds. This has important implications for the trade-off between parental care and fecundity. It is tempting to argue that the energy spent in excess of what a nesting bird spends in egg production is the cost of parental care. This is imlikely to be the case, as there are costs associated with nesting, such as predation risks and time constraints, that are difficult to express in energy currency and that limit reproductive effort. The data presented here clearly demonstrate, however, that cowbirds are directing their reproductive effort into production (Roff 1992).

PAGE 29

CHAPTER 4 HOST NEST SELECTION Introduction Egg laying by a cowbird in a host nest is the culmination of a process of nest selection (Thompson and Gottfried 1981). The first step in the process is finding an active nest. Once nests are located, cowbirds must select the nests in which they will lay eggs. Nest selection may be important for two reasons. First, cowbirds must lay their eggs at a time that maximizes both probability of acceptance by the host and chances of survival for the parasitic egg and yovmg. Some hosts reject cowbird eggs that are laid before the initiation of their own lajdng period (Friedmann and Kiff 1985; Wiley 1985). For example, Yellow Warblers (Dendroica petechia) usually reject cowbird eggs laid before their laying period by nest desertion or burying the egg in the nest lining (Clark and Robertson 1981; Weatherhead 1989). On the other hand, parasites laying in nests where incubation is already underway have a low chance of success because the parasitic egg may not hatch, or may hatch too late to compete successfully with its nestmates (e. g., Fraga 1985; Weatherhead 1989). Second, nest choice by cowbirds may be important in avoiding competition with other cowbirds. Brood parasites are probably limited by the availability of host nests (Payne 1977) and multiple parasitism (more than one egg in a host nest) is Ukely to result in low cowbird success (Weatherhead 1989). Thus, cowbirds are expected to exhibit mechanisms to avoid competition, 22

PAGE 30

23 such as 1) defending nests or laying areas from other females, 2) destroying or removing cowbird eggs from nests, or 3) avoiding laying in already parasitized nests (Post and Wiley 1977; Darley 1983; Fraga 1985; Teather and Robertson 1985; Orians et al. 1989). Here I test the h5T)othesis that Shiny Cowbirds invest some effort to ensiire placing each egg in the most advantageous circimistances possible (i. e., they select a nest). An alternative hypothesis is that cowbirds use a "shotgun" strategy (Rothstein 1975). Instead of timing egg laying with a few host nests, cowbirds search for nests in a random manner and parasitize as many as they can, whatever their stage of development. I tested the following predictions: Prediction 1 : Cowbirds shordd monitor nest progress in order to determine when it will be in an appropriate stage for parasitism. Prediction 2 : Cowbirds should time their egg laying with the host's laying period. Prediction 3 : Multiple parasitism should be rare. There should be few m\iltiply parasitized nests. Prediction 4 : Distribution of cowbird eggs in wren nests should be nonrandom, with less multiply-parasitized nests than expected from a Poisson distribution (Orians et al. 1989); cowbirds shoiild remove or destroy previously laid cowbird eggs. Methods To test the prediction that cowbirds monitor nest progress, I made 1-h observation sessions of wren nests during three stages: a) Early nest construction, when wrens are adding coarse material; b) Late nest

PAGE 31

24 construction, when wrens are finishing a cup of fine material and adding lining; and c) Laying period. During these sessions I recorded number of cowbird visits and cowbird behavior. If cowbirds approached the nest rapidly and directly, I assumed they knew the location of the nest beforehand (Wiley and Wiley 1980). I assumed these rapid visits were visits to check nest progress, as opposed of visits of cowbirds searching for a nest for the first time. To determine synchronization of cowbird egg la5dng with the host's laying period, I checked all active nests daily during the building and laying stages. Cowbird eggs were marked with permanent ink for identification and dates of laying noted. All nests were checked between 06:30 and 09:00 each morning and observation sessions made between 08:00 and 10:00. The prediction that multiple parasitism should be rare was tested by comparing the distribution of cowbird eggs in wren nests with a Poisson distribution (Orians et al. 1989). Agreement with the Poisson distribution would indicate that cowbird eggs were randomly distributed in wren nests, while departure from the Poisson would indicate a clumped or regular dispersion of eggs. Results Cowbirds made inspection visits to wren nests during all stages of the nesting cycle. During inspection visits cowbirds flew directly to the nest, entered the cavity for less than 15 sec, and left quickly. This direct and rapid approach indicates that cowbirds knew the location of the nest beforehand (Wiley and Wiley 1980). Visitation rates tended to be higher during early stages of nest building, but the difference between stages was non-significant (ANOVA, F=2.84, df=2, 29, P=.075; Fig. 4-1). Inspection visits were mostly by

PAGE 32

25 16-26 4 n — , — a b NEST STAGE Figure 4-1. Frequencies of female cowbird inspection visits to wren nests during different stages of nest building, a) Building of nest foundation; b) Building of nest cup and lining; c) Laying period. Error bars indicate one standard deviation. Numbers above bars indicate number of nests and total nixmber of hours of observation (the sample unit is the nest).

PAGE 33

26 solitary females, rarely by groups of 2-4 females (Table 4-1). Males sometimes escorted females (Table 4-1), but they did not get close to the nest. Inspection visits indicated that cowbirds were monitoring nest progress. Patterns of inspection visits (Fig. 4-1) and multiple parasitism indicated that several female cowbirds monitored each nest. Assuming that each female always lays similar-looking eggs, as foxmd for Brown-headed Cowbirds (Dufty 1983), egg variability indicated that at least six different females laid in some nests. Besides, sometimes up to six eggs were laid in a nest in a single day. Resightings of banded females indicated that each female monitored several nests simultaneously over the course of several days. Home ranges of cowbirds were larger than my study area of 14 wren territories, and based on egg variability, at least 15 females were laying in the area. The actual nimiber of females in the area, however, was much larger because I banded 37 females during the study and very rarely did I resight banded females. Cowbirds tended to lay their eggs in coincidence with the wrens' laying period (Pig. 4-2). Forty-seven percent of 185 eggs laid in 40 nests were laid during the 3-day laying period of wrens. A large number of cowbird eggs, however, were laid before wrens started laying (35%) or after the wrens completed their clutch and initiated incubation (18%; 'X^= 24.7, df=2, P<.001). The above sample includes only nests in which wrens actually laid, and it underestimates the proportion of cowbird eggs laid prematurely (before the wrens' laying period), because cowbirds laid 82 eggs in 20 nests that were abandoned by wrens before laying. Sometimes these eggs were laid too early during nest construction and were buried as wrens continued adding nesting material.

PAGE 34

27 Table 4-1. Group size of female cowbirds during inspection visits to wren nests. Escorted by Males Group size No. observations Proportion Mean No. (%) visits males/group 1 39(76.5) 0.18 1.3 2 6(11.7) 0.17 1.0 3 3(5.9) 0.66 1.5 4 3(5.9) 0.66 2.0

PAGE 35

CO o O o o fa o 50n 40 3020100 I I Ml III lllilll i .7^-5^-3 -2 -1 0 1 2 3 4 5 6 7 DAY OF NESTING CYCLE Figure 4-2. Number of cowbird eggs laid in different days of the nesting cycle. Day 0 = Day in which the first wren egg is laid. Modal clutch size of wrens is 3, and wrens start incubating after laying the last egg (day 2).

PAGE 36

29 Wrens abandoned nests that were parasitized with more than two eggs, but accepted cowbird eggs when parasitized with one or two eggs, regardless of when were they laid (see Chapter 6). Contrary to predictions 3 and 4, multiple parasitism was common and cowbirds did not avoid multiply parasitized nests. Multiple parasitism was more common and more intense in boxes than in natural nests (Fig. 4-3), reflecting the conspicuousness of nest boxes. The average number of eggs per nest was higher in nest boxes (1.97) than in natural nests (5.65; Mann-Whitney U, P<.001). In natural nests, the distribution of cowbird eggs was significantly different from a Poisson distribution ('X=38.52, df=4, P<.001; Fig. 4-3). There were more unparasitized nests than expected, which reflects the inability of cowbirds to enter cavities with small openings (see Chapter 6). Among parasitized natural nests, there were more nests than expected with 5-8 eggs/nest, and less than expected with 1-4 eggs/nest. This result suggests that some nests were easier to find and that cowbirds did not avoid already parasitized nests. The distribution of cowbird eggs in nest boxes was also significantly different from a Poisson distribution 0(^=29.5, df=4, P<.001; Fig. 4-3). The large number of multiplyparasitized nests suggests that boxes were easier to find than natural nests. There were more nests than expected with 11-12 eggs/nest, again suggesting that cowbirds did not avoid already parasitized nests. I found no evidence that cowbirds removed previously laid eggs in multiply-parasitized nests, but in a few cases they may have destroyed eggs by perforating a hole in the shell. Embryos in perforated eggs usually die of dehydration (Carey 1986). I found 9 out of 150 eggs in multiply-parasitized nests that had a single perforation, as would be produced by a peck. Four of the nine perforated eggs were foimd in a single nest, together with two perforated wren eggs. It is possible that these perforations were made by

PAGE 37

30 NATURAL NESTS 0 123456789 10 11 12 201 NEST BOXES 0 123456789 10 11 12 NUMBER OF EGGS PER NEST Figure 4-3. Distribution of cowbird eggs in wren nests. The bars indicate the observed number of nests and the line represents the Poisson expected distribution for a mean of 1.97 (natural nests) and 5.65 eggs/nest (boxes). Both distributions deviate significantly from the expected.

PAGE 38

31 other cowbirds, but they also may have been made by wrens, as they frequently destroyed conspecific and heterospecific neighboring nests (N, G6mez and G. Kattan, unpublished). Discussion Once cowbirds find a nest, they visit it frequently and monitor its course. Presumably, monitoring a nest would allow a parasite to determine the most appropriate moment for egg laying. Wiley and Wiley (1980) reported that Shiny Cowbirds spent the day searching for nests in colonies of Yellowhooded Blackbirds (A gelaius icterocephalus) ; once nests were found, cowbirds inspected them frequently. Shiny Cowbirds in Puerto Rico made inspection visits to nests of three host species during pre-lajdng and laying stages (Wiley 1988). In contrast to this study, Wiley (1988) found that visitation rates peaked arovmd the day hosts laid their first egg, and he suggested that hosts provide some clue that cowbirds use to synchronize their oviposition. My results indicate that more cowbirds tend to visit the nest early during nest construction, but visitation frequencies are not significantly different over the stages (Fig. 4-1). As construction advances, probably some cowbirds lay and then stop visiting. The high intensity of parasitism foimd in this study suggests that cowbirds compete for available nests, and it is intriguing that in this population female cowbirds apparently do not exhibit mechanisms to outcompete other females. A high incidence of perforated cowbird eggs has been reported in other populations of Shiny Cowbirds (Sick 1958; Hoy and Ottow 1964; King 1973; Post and Wiley 1977; Fraga 1985); before laying her egg, the cowbird pecks any other cowbird eggs in the nest. In this study I

PAGE 39

32 foiind a low incidence of perforated cowbird eggs (9 of 150 eggs). My analysis of cowbird egg distribution in wren nests also indicates that female cowbirds do not avoid laying in already parasitized nests (Fig, 2-3). In the few cases in which wrens incubated more than two cowbird eggs, only one egg hatched and the embryo partially developed in one or two additional eggs (Chapter 6). Probably in these cases the female that laid the last, was the one that was successful, because her egg was on top. My results partially supported the h5T)othesis that cowbirds time their egg laying to coincide with the wren's laying period. Some Shiny Cowbird hosts are discriminative and reject odd-looking or non-sjnichronyzed eggs (e. g., Fraga 1985; Wiley 1985; Mason and Rothstein 1986). Other hosts, in contrast, accept almost any egg at any time, as is the case for House Wrens (see Chapter 6). Rufous-collared Sparrows (Zonotrichia capensis ) also accept nonsynchronized, odd-looking eggs, and as House Wrens do, only reject by desertion when parasitized with more than two cowbird eggs (Sick 1958; King 1973; Fraga 1978, 1983; pers. obs.). The importance of synchronization depends not only on the host's rejection response, but in the probability of success of the parasitic yoimg. With large hosts, cowbird reproductive success may be low. For example, Fraga (1985) found that 37% of cowbird nestlings raised with broods of Chalk-browed Mockingbirds (Mimyg satiiminus ) starved. With small hosts, in contrast, cowbirds usually have a high success because of the size advantage. Cowbirds usually hatch first, and nestlings are aggressive and grow fast (see Fig. 4-4). Frequently the host young suffer high mortality and the cowbird nestling is raised alone (Fraga 1983; Chapter 6). The Shiny Cowbird has a large geographic range and parasitizes a large number of species (Friedmann and Kiff 1985). It is, therefore, difficult to

PAGE 40

33 701 eo 50 40 30 20 10 i 0 liH adult males adult females first molt juveniles i Cowbirds (n=12) ^ ;-f-,rf-i-f^-I-.-f — Wrens (n=12) " 1 ' — I ' 1 ' 1 ' 1 1 1 1 1 r— I 1 1 0 2 4 6 8 10 12 14 16 18 AGE (days) Figure 4-4. Growth rates of Shiny Cowbird and House Wren nestUngs. Error bars indicate one standard deviation. Error bars after day 9 increase for cowbirds because of divergence in size between males and females.

PAGE 41

34 make generalizations, and they probably exhibit different strategies under different circumstances. Shiny Cowbirds in the Cauca Valley seem to be adopting the "shotgun" strategy (Rothstein 1975). They monitor host nests and to some degree time their egg laying to coincide with a period in which their eggs are more likely to be successful. The vmselective behavior of wrens, however, allows for a wide window of time in which their eggs are accepted. Furthermore, wrens are abimdant and have an extended breeding season, providing a high availability of nests almost year-round. The Shiny Cowbird's strategy is to be an extreme generalist that bases its success on high fecimdity more than on the precise placement of each egg. It is possible that species that follow other strategies, such as cuckoos (Wyllie 1981), have more precise behavioral and/or physiological mechanisms for synchronization of oviposition.

PAGE 42

CHAPTER 5 MECHANISMS OF SHORT INCUBATION PERIOD IN THE SHINY COWBIRD Introduction "The comparatively short time the embryo takes to hatch gives it [the Shiny Cowbird] another and a greater advantage; for whereas the eggs of other small birds require from fourteen to sixteen days to mature, that of the cowbird hatches in eleven days and a half from the moment incubation commences." (W. H. Hudson 1920) Brood parasitism is a reproductive strategy in which rapid embryonic development is at a premium. Eggs of brood parasites are abandoned in nests of host species and hatchlings have to compete with foster nestmates for parental resources. Therefore, early hatching gives parasitic young a head start (Pajnie 1977). One strategy used to eliminate competition in the nest is displayed by parasitic cuckoos. Cuckoos usually hatch earlier than their hosts. Upon hatching, the cuckoo ejects host eggs by pushing them over the rim of the nest with their backs (WylHe 1981). Cowbirds, in contrast, do not kill nestmates, but early hatching helps them to gain an advantage over their nestmates (Payne 1977). It has long been recognized that brood parasites have short incubation periods (Friedmann 1927; Hamilton and Orians 1965; Payne 1977; Briskie and Sealy 1990). There is uncertainty, however, about the precise duration of incubation periods because of the inherent variability associated with variable attendance by incubating birds. This problem led Nice (1953) to question the validity of reports of ten-day incubation periods in the Brown35

PAGE 43

36 headed Cowbird. One problem with measuring incubation periods in the field is the uncertainty in determining when did incubation begin £uid how constant it was. Briskie and Sealy (1990) defined incubation period of cowbird eggs as the time firom initiation of fiill incubation by the host to hatching. Initiation of full incubation, however, was defined as "usually just after laying of the penultimate egg" (by the host). Although Briskie and Sealy (1990) found that incubation periods of parasitic cowbirds were shorter than expected firom egg size, when compared with non-parasitic relatives, the imcertainty in incubation periods led them to do separate analysis for minimimi and mean incubation periods of cowbird eggs. Here I present measurements of incubation periods of Shiny Cowbird eggs under constant conditions in the laboratory and examine possible mechanisms for the evolution of short incubation periods in parasitic cowbirds. Friedmann (1927) suggested that a short incubation period in Brown-headed Cowbirds represented an adaptive acceleration of embryonic development. Incubation period, however, scales allometrically to the 0.217 power of fresh egg mass (Rahn and Ar 1974). Egg mass, in turn, scales to the 0.675 power of adult female body size (Rahn et al. 1975). Thus, incubation periods in parasitic birds may be short because eggs are smaller than expected (Briskie and Sealy 1990). Parasitic cuckoos lay small eggs and have correspondingly short incubation periods, usually hatching up to four days before their hosts (Rahn et al. 1975; Payne 1977). Cowbirds, in contrast, lay normal-sized eggs, i. e., egg size corresponds to that expected from female body size. Still, their incubation periods are shorter than expected and eggs usually hatch 1-2 days before their hosts, even though cowbird eggs are usually larger (Payne 1977; Briskie and Sealy 1990).

PAGE 44

37 I examined two possible mechanisms by which cowbirds may shorten incubation periods. The first mechanism is based on the assumption that there is a window of time at the end of embryonic development during which hatching can occur. Avian embryonic growth can be represented by the parabolic equation W=atb where W is embryo size, t is incubation time and a and b are constants (Fig 5la; Ricklefs 1987). Embryonic development can be divided into two phases. During the organogenesis phase there is rapid development of new tissues to form organs, but relatively little increase in embryonic mass. Following the differentiation phase comes a phase of rapid growth (increase in embryonic mass) and functional maturation of tissues (Freeman and Vince 1974; Balinsky 1975). Thus, it is conceivable that there is a time to after which maturation is complete enough to allow hatching (Fig 5-la). Embryos continue to grow after time to and hatch at time te, presumably because there is an advantage in hatching at a larger size. I hypothesized that cowbirds shorten the incubation period by hatching at to (Fig. 5-la). Because hatchling size correlates with egg mass (Vleck and Vleck 1987), I predicted that hatchling size (Wq) would be smaller than expected from egg mass (We). I also hypothesized that the mechanism to shorten incubation period within this window of time (to-te) involved a decrease in energy content of the egg. During the final days of development, growth presumably continues until yolk reserves are depleted (except for a small amount that is retained as reserve for the hatchling). Depletion of yolk reserves or some unknown yolk component would signal the embryo that it is time to hatch. This hypothesis is based on the existence of a tight correlation between incubation period and egg energy content (tighter than the correlation between incubation period and egg mass) and the observation that all avian embryos expend about the same proportion of energy stored before

PAGE 45

Figure 5-1. Hypotheses to explain short incubation periods in cowbirds. A) Embryo size as a function of incubation time. The expected incubation time for a given egg size is te, and hatchling size is we. The hypothesis is that there is a time to after which hatching is possible. This hypothesis predicts that hatchling size (wq) will be smaller than expected from egg mass. B) The alternative hypothesis is that cowbirds shorten incubation time by increasing developmental rates. This hypothesis predicts that hatchling size will be the expected from egg size (we). C) Embryonic metabolic rate as a fimction of embryo size. Hypothesis B predicts that metabolic rate will be higher than expected from embryo size.

PAGE 46

EMBRYO SIZE We

PAGE 47

40 hatching (Vleck and Vleck 1987). Assuming that the rate of energy expenditure is constant, a reduced energy content would result in a shorter incubation period. I predicted that cowbirds 1) would have an energy content lower than expected from egg mass, 2) incubation period would be shorter than expected from egg mass but not different from the expected from egg energy content, and 3) hatchling size would correspond to the expected from energy content. Expected values for these variables were derived from Vleck and Vleck's (1987) review on the energetics and metabolism of avian embryos. An alternative hypothesis is that cowbirds could shorten incubation period by increasing developmental rates (Fig. 5lb). In this case, I predicted that hatchling size would not be different from that expected from egg size. I also tested this hypothesis by measuring embryonic metabolic rates, under the assimiption that an increased developmental rate would result in an increased rate of oxygen consumption (Vleck et al. 1980). Metabolic rates of bird embryos are usually compared using the pre-intemal pipping metabolic rate (PIP-V02), defined as the metabolic rate just prior to the perforation of the internal air cell of the egg, which occurs shortly before hatching (Vleck and Vleck 1987). I predicted that cowbird embryos would have higher PIP• V02 than expected from the corresponding embryo size. These two hypothesis are not mutually exclusive and both mechanisms may be operating simultaneously. Methods During 1990, 1 collected fresh Shiny Cowbird eggs from House Wren nests in the study area. I collected only cowbird eggs that were laid before the wrens started laying to ensure that the eggs had not been incubated in the

PAGE 48

41 field. Eggs were incubated in the laboratory at a constant temperature of 38 °C. Eggs were turned manually 6-8 times a day. After day 10, 1 checked the eggs every 4 hours during the day and every 6 hours during the night to determine hatching time. Hatching was defined as the moment in which the shell started to split (actual emergence of the hatchling occurred 2-3 hours after the shell split). Incubation time of cowbird eggs in the field was defined as the time between the laying of the last wren egg (which is when constant incubation by the female wren usually begins) and the day of hatching. Only cowbird eggs that were laid before the last wren egg were used. Also, for determination of incubation period in the field I only used nests in which incubation had proceeded without major interruptions. Laboratory-incubated eggs were used for measurements of embryonic metabolic rates. Metabolic rates were measured at 38 °C in a Scholander-type respirometer, using the method described in Hoar and Hickman (1983). In brief, eggs were placed in a vial (egg chamber), together with filter paper soaked in a 10% solution of KOH (CO2 absorbent). The compensation vial contained only the CO2 absorbent. A V-shaped tube connected vertical tubes placed above the egg and compensation chambers. The V-tube contained a small amount of water colored with anilin. The vertical tube above the animal chamber was connected to a calibrated syringe. The egg was placed in the vial and the apparatus was immersed in a water bath. I waited 10 min for temperature in the chambers to equilibrate with water temperature. After temperatures equiUbrated, I closed stopcock valves on the vertical tubes and started a measurement of oxygen consumption. After each measurement, the manometer fluid was pushed back to its original position with the syringe and the volume of oxygen consumed was read from the syringe graduations. I made at least four consecutive measurements each

PAGE 49

42 time and averaged them to obtain a value of oxygen consumption. Fresh air was allowed in the chamber between readings. Oxygen volumes were converted to STPD conditions. Additional cowbird eggs were collected for measurements of energy content. I hard boiled 10 eggs and removed the shell. Energy content of the egg was then measured by bomb calorimetry at the Industrial Analysis Laboratory of the Universidad del Valle (Cali, Colombia). Hatchling mass was determined from laboratory-incubated eggs. I removed residual yolk from hatchlings and dried their carcasses at 70 °C until a constant mass was obtained. For data analyses, I compared the observed values with expected values for each parameter, derived from least-squares regressions published in Vleck and Vleck (1987). Significance of the deviations was tested by comparing values with the 95% prediction interval, calculated as described in Montgomery (1984), except for egg mass as a fvmction of adult female body size, which was compared with the 68% confidence interval provided in Rahn et al. (1975). Results Egg size of Shiny Cowbirds (x = 4.3 g, SD = 0.4, range = 3.5-5.2 g, n=67) was on the lower range of values expected from adult female body size. Egg mass expected for an average female cowbird of 54 g is 5.1 g (68% confidence interval = 4.1-6.2; Rahn et al. 1975; Table 5-1). Incubation period was shorter than expected from egg mass. Incubation period of Shiny Cowbird eggs under constant laboratory conditions ranged between 11.2-12.1 days (x = 11.7 days, SD = 0.5, n=ll), sHghtly overlapping with the lower limit of the 95%

PAGE 50

43 Table 5-1. Female size, egg size and incubation period of Shiny Cowbirds and House Wrens in the Cauca Valley, Colombia. Parameter Shiny Cowbird House Wren Female size (g) 54.6+3.45(47) 15.34±1.15 (16) Egg Size (g) Observed 4.3±0.39 (49) 1.87+0.13 (34) Expected* 5.1 2.15 Incubation Period (days) Observed Laboratory 11.7±0.5 (11) — Field 12.0±0.8 (7) 15.3±0.8 (22) Expected From Egg Size** 14.25 11.19 From Energy Content*** 13.8 — * From Rahn, Paganelli and Ar's (1975) equation for passerines (E=0.34W0-677). ** From Vleck and Vleck's (1987) equation for altricial birds (log 1=0.97+0.29 log E). *** From Vleck and Vleck's (1987) equation for all birds (log 1=0.83+0.27 log EC).

PAGE 51

44 prediction interval (11.8-16.9 days) derived from Vleck and Vleck's (1987) equation (predicted incubation period = 14.3 days; Table 5-1). Incubation period of cowbird eggs correlated significantly with egg mass (Fig. 5-2). Egg mass, however, explained only 44% of the variation in incubation period, suggesting that other factors are also influencing incubation period (Fig. 5-2). Estimates of cowbird incubation period in the field agreed with laboratory results. Incubation periods in the field ranged from 11 to 13 days, much shorter than the 14-16 day incubation periods of House Wrens (Table 5-1). This resulted in cowbirds hatching earlier than wrens. In 15 wren nests in which one cowbird egg was laid before the initiation of incubation and that survived the incubation stage, the cowbird always hatched 2-3 days before the wrens. The short incubation period of cowbirds allowed them to hatch synchronously with the wrens even when cowbird eggs were laid after the initiation of incubation. In 8 wren nests in which a cowbird egg was laid 1-4 days after the last wren egg was laid, cowbirds hatched the same day or the next day after the wrens. As predicted, energy content of Shiny Cowbird eggs (14.54 ± 0.19 kJ, n=10) was lower than expected fi-om mean egg mass (expected energy content = 19.05 kJ, 95% prediction interval, 16.3-22.2; Table 5-2). The incubation period of cowbird eggs (11.7 days) was within the limits expected from egg energy content (95% prediction interval, 11.4-16.9; Table 5-2). Yolk-fi-ee, dry hatchhng mass of Shiny Cowbirds (0.3091 ± 0.04 g, n=6) was lower than the lower limit of the prediction interval expected from egg mass (expected value=0.3786, 95% prediction interval, 0.3105-0.4613), but close to the value predicted from egg energy content (expected value=0.2835, 95% prediction interval, 0.2371-0.3350; Table 5-2).

PAGE 52

46 Figure 5-2. Incubation period iinder constant laboratory conditions as a function of fresh egg mass for Mbonariensia (P=.03, r2=.44, n=ll).

PAGE 53

46 Table 5-2. Egg energy content, hatchling mass £ind embryonic metabolic rate of Shiny Cowbirds. Expected values as a fimction of egg mass and egg energy content, calculated from equations in Vleck and Vleck (1987). Nvmibers in parenthesis imder the observed column are sample sizes and imder the expected column are 95% prediction intervals (Montgomery 1984). Parameter Observed Expected from Egg Mass Energy Content Energy Content (kJ) Energy Used (kJ)* Yolk-free, Dry Hatchling Mass (g) PIP-VO2 (ml/day) 14.54+0.19 (10) 4.45 0.3091+0.04 (6) 50.19 19.05 (16.3-22.2) 6.95 0.3786 (0.3105-0.4613) 5.28 0.2835 (0.2371-0.3350 47.86 (37.9-64.4) * Calculated as the area under the metaboHc rate fimction for t=0-ll 7 (Fig 5-3). ^'

PAGE 54

47 The metabolic rate of Shiny Cowbird embryos followed the typical pattern of altricial birds, with metabolic rate increasing exponentially with incubation time (Fig. 5-3). PIP-V02 was calculated as the metabolic rate expected from the metabolic rate function at 90% of the mean incubation period (Fig. 5-3). The PIP-V02 (50.19 ml/day) was within the Umits of the prediction interval expected from a regression of metabolic rate against energy content for altricial birds (predicted value=47.86, 95% prediction interval, 37.9-64.4; Table 5-2). Discussion Incubation of Shiny Cowbird eggs under constant conditions in the laboratory resulted in a mean incubation period of 11.7 days. This incubation period is shorter than expected from egg size and remarkably short when compared with that of other passerines, even those smaller than cowbirds. For example, House Wrens are much smaller but have an incubation period of 14-16 days (Table 5-1; Alvarez et al. 1984). This incubation period is longer than the value of 11.19 days expected from the wren's egg size (Table 5-1). It would be expected that incubation periods of cowbird eggs under natural conditions in the field may be longer because of intermitent attendance by the incubating bird. Estimates of cowbird incubation periods in the field, however, agree with the laboratory results (Table 5-1; Briskie and Sealy 1990). Incubation periods of Shiny Cowbird eggs incubated in wren nests under natural conditions varied between 11 and 13 days (Table 5-1). This resulted in cowbirds hatching earlier than wrens when cowbird eggs were laid before the initiation of incubation. Even when cowbird eggs were laid 1-4

PAGE 55

48 4l INCUBATION TIME (days) Hatch Figure 5-3. Oxygen consumption (STPD) of Shiny Cowbird embryos as a function of incubation time. Seventy-eight measurements based on 12 eggs.

PAGE 56

49 days after the initiation of incubation, the short incubation period allowed cowbirds to hatch at around the same time that the wrens hatched. As predicted, I foimd that Shiny Cowbird eggs had an energy content lower than expected from egg mass (23.8% lower; Table 5-1). Incubation period was 28.8% shorter than expected from egg mass, but was closer to that predicted from energy content (Table 5-1). Similarly, hatchling size was smaller than predicted from egg mass but not significantly different from that predicted from the actual energy content of cowbird eggs. These results support the hypothesis that cowbirds shorten incubation period by decreasing energy content of the egg (Fig 5-la). The alternative h3T)othesis that cowbird embryos have accelerated rates of development (Fig 5lb, 5-lc), in contrast, was not supported. Hatchling size was smaller than expected from egg mass. Measurements of metabolic rates did not support the hypothesis that cowbirds had high growth rates. Pre-internal pipping metabolic rates of Shiny Cowbird embryos were not different from the expected from a regression of PIP-V02 on egg energy content in altricial birds (Table 5-2), suggesting that developmental rates were not higher than expected based on embryonic size. My results suggest that the mechanism by which cowbirds shorten incubation period is a combination of laying a slightly smaller egg than predicted by female body size, that has a reduced energy content. One potential problem with this mechanism is that reducing hatchling size could be a disadvantage, because cowbirds have to compete with nestmates for food deUvered by the foster parents. Cowbirds, however, usually parasitize species smaller than themselves (Friedmann and Kiff 1985). In this case, the slightly smaller hatchling size would be irrelevant. Wren hatchlings, for example, weigh less than 2 g, as opposed to 3.2-4.5 g for a cowbird hatchling, and this

PAGE 57

50 difference increases rapidly with nestling growth (see Fig. 4-4). If the host species is larger than the cowbird, the latter may be at a disadvantage (e. g., Fraga 1985; Carter 1986). The advantage of hatching early, however, probably outweighs the disadvantage of a slightly smaller hatchling. Energy stored in the egg is a major factor influencing avian embryonic development (Vleck and Vleck 1987). Each egg is provided with a fixed amount of energy, which is used for growth and somatic maintenance of the embryo (except for a small amount that is retained as residual yolk in the hatchling). Energy content of the egg explains most of the interspecific variance in variables associated with embryonic development, such as incubation period, hatchling size and energetic cost of development (i. e., total energy spent during development). When these variables are regressed against egg mass, two separate lines are required to describe altricial and precocial species. This difference disappears when energy content is used as independent variable (Vleck and Vleck 1987). The correlation between incubation period and energy content, however, does not imply causality and the two variables may be correlated with a third, unknown variable. In a discussion of the evolution of avian altriciality, Vleck and Vleck (1987) argue that a mutation occurred that caused a shortening of the incubation period. The resulting hatchling would be smaller, but would have a large amount of residual yolk. At this stage, there would be selection on the female to invest less energy in each egg. Thus, according to this interpretation, the incubation period got shorter first by an independent mechanism and then the energy density of the egg was decreased (Vleck and Vleck 1987; CM. Vleck, pers. com.). There is experimental evidence, however, that reducing the energy content of the egg results in a shorter incubation period in other vertebrates. Sinervo (1990)

PAGE 58

51 manipulated energy content of eggs of the lizard Sceloponis occidentalis by extracting variable amounts of yolk (10-50% of total egg mass). Some eggs were sham-manipulated (poked with a sjainge) but no yolk was extracted. He foimd that eggs with more yolk removed had shorter incubation periods than unmanipulated or sham-manipulated eggs. Eggs from which yolk was removed developed normally and hatched into viable offspring that were smaller than those from control eggs (Sinervo 1990). Sinervo (1990) also foimd interpopulational differences in incubation time and hatchling size, among other variables, and his experimental manipulation of yolk content (i. e., energy content) demonstrated that these differences were probably due to differences in yolk content (reflected in egg size) among populations. Sinervo's (1990) results lend support to the hypothesis that cowbirds could shorten incubation period by reducing the energy content of the egg. The mechanism triggering early hatching may be simply that the embryo is running out of yolk (Sinervo 1990). If there is a threshold time to after which hatching is possible (Fig. 5la), the amount of yolk reserve left could determine the timing of hatching. When yolk is almost depleted, the embryo would receive some signal that it is time to hatch. If more reserves are available, the embryo could continue increasing in mass. This hypothesis assumes that incubation period in altricial birds is not already minimized, i. e., that the interval to-te exists. Opposite selection for short incubation period versus large hatchling size is likely to result in the existence of this interval in which there is a balance between the two forces. Hamilton and Orians (1965) suggested that short incubation periods are a pre-adaptation to brood parasitism (an exaptation in the terminology of GovJd and Vrba 1982). Short incubation periods may have evolved in different lineages in response to different selection pressures. For example,

PAGE 59

52 Hamilton and Orians (1965) suggested that accelerated incubation periods and nestling growth are an adaptation of cowbirds for feeding among nomadic herds of large mammals. The short incubation period would have evolved in response to the need to be constantly on the move. This then would have contributed to the success of cowbirds as brood parasites. If short incubation periods are a preadaptation to brood parasitism, then close relatives of cowbirds would be expected to exhibit the same trait. Hamilton and Orians (1965) suggested that the incubation period of parasitic cowbirds was not shorter than that of non-parasitic blackbird relatives. Briskie and Sealy (1990), however, compared incubation periods of 22 species of icterines as a function of egg size, and foimd that parasitic cowbirds had shorter incubation periods than non-parasitic icterines. They suggested that short incubation periods evolved in cowbirds as an adaptation for brood parasitism. This issue will remain imresolved until a comparative study of female size, egg size, incubation period and embryo energetics of icterines is conducted. Here I have indicated that short incubation periods in cowbirds may be the result of a reduced energy content. This suggests an alternative scenario for the evolution of short incubation periods in a life history theory framework. Cowbirds are extremely fecund birds, laying almost continuously during the breeding season. Life history theory predicts a trade-off between number and quality of offspring. Increasing the number of offspring impUes a decrease in the investment in each offspring (Roff 1992). Cowbirds have adopted a strategy of laying a large number of eggs with a reduced investment in each egg (Chapters 3, 4). Thus, the reduced energy content of cowbird eggs could be the direct result of this trade-off.

PAGE 60

53 What is, then, the selective pressure for a reduced energy content in cowbird eggs? Elucidation of this question is difficult because the three factors, namely incubation period, energy content and fecundity, £ire functionally linked (Fig. 5-4). If there is a selective pressure for shortening incubation period, this would result in a reduced energy content, which then could result in an increase in fecundity. The opposite pathway is also possible. If the reduced energy content is a result of a selective pressure to increase fecundity, then the short incubation period is a side benefit and not a specific adaptation for brood parasitism. Alternatively, both incubation period and fecimdity could be acting as selection pressures to decrease energy content of cowbird eggs. Perhaps the best way to address this problem is by conducting a comparative study of icterines, as suggested above. Comparison of these characteristics among related species (including non-parasitic species which do not exhibit the high fecimdity of cowbirds) would allow to determine which characteristics are primitive or derived, and which of these characteristics, if any, sire specific adaptations of cowbirds for brood parasitism.

PAGE 61

LOW EGG ENERGY CONTENT HIGH FECUNDITY SHORT INCUBATION PERIOD Figure 5-4. Functional relationship between fecundity, egg energy content and incubation period in cowbirds. A selection pressure for high fecundity results in a decrease in the energy deposited in each egg, which then results in a short incubation period. In this case the short incubation period is a side benefit and not a specific adaptation for brood parasitism. In the opposite pathway, the selective pressure for a low energy content is the incubation period.

PAGE 62

CHAPTER 6 IMPACT OF BROOD PARASITISM: WHY DO WRENS ACCEPT COWBIRD EGGS? Introduction Because brood parasitism usually depresses reproductive success of the host, hosts are expected to evolve defenses against parasitism (Payne 1977; Rothstein 1990). One such defense is rejection of peirasitic eggs. Hosts of parasitic cowbirds can be classified into two discrete groups, "accepters" and "rejecters," according to their responses to natural and experimental parasitism (Rothstein 1975a; Mason 1986a). Parasitic eggs may be rejected by ejecting the egg, deserting the nest, or constructing a new nest floor over the parasitized clutch. Why do some hosts accept parasitic eggs? Rothstein (1975a, 1982, 1990) suggested that lack of rejection in some cowbird hosts is best explained by absence of genetic variants in the population with the ability to recognize and reject a foreign egg (evolutionary lag hypothesis). An alternative hypothesis suggests that for some cowbird hosts, rejection incurs costs that exceed any benefits, thereby making acceptance the most adaptive option (cost of rejection hypothesis). For example, egg ejection may be costly for hosts with small bills because they are unable to grasp a cowbird egg between the mandibles. The only way to eject the egg would be to pimcture it and hold it by the perforation. Cowbird eggshells, however, are imusually thick and therefore very strong and some of the host's own eggs could be damaged in the puncturing process (Rohwer and Spaw 1988; Rohwer et al. 1989; R0skaft 55

PAGE 63

66 et al. 1990). This hypothesis may explain why most small-billed hosts accept eggs of the Brown-headed Cowbird (Rothstein 1975a, 1982; Rohwer and Spaw 1988). Rejection of parasitic eggs by nest desertion also incurs costs that may exceed the costs of accepting parasitism. Nest desertion involves a time cost, which may be significant if duration of the breeding season is restricted. Besides, good nest sites may be in short supply and the replacement nest may also be parasitized (R08kaft et al. 1990; Petit 1991). Thus, to determine the best strategy for a host to follow, the costs of parasitism must be contrasted with the costs of rejection. The cost of rejection hypothesis assimies that hosts have the ability to recognize foreign eggs, but they tolerate parasitism because the costs of rejecting are higher. Here I test this hypothesis by examining the impact of Shiny Cowbird parasitism, and host responses in the House Wren. It has been hypothesized that cavity-nesting species experience low rates of parasitism, and as a result these species have not evolved egg discrimination behavior (Friedmann 1963; Friedmann et al. 1977). An alternative hypothesis is that wrens can discriminate foreign eggs, but are unable to grasp-eject cowbird eggs because they are large and heavy. I tested the following hypotheses and predictions: 1) Wrens can discriminate foreign eggs but can not Uft cowbird eggs; thus, they should eject artificial odd-looking eggs that are smaller and fighter than cowbird eggs, but similar in size and weight to wren eggs. 2) Wrens can grasp-eject cowbird eggs but do not recognize them as foreign; thus, wrens should accept cowbird-sized artificial eggs but should eject odd-looking objects as large and heavy as cowbird eggs.

PAGE 64

57 I also examined the potential costs of nest desertion and contrasted them with the costs of accepting parasitim, to test the hypothesis that wrens tolerate parasitism because the costs of deserting are higher than the costs of accepting. Methods For this study, I followed the nesting histories of all pairs of wrens in the area and obtained data on incidence (proportion of nests parasitized) and intensity (nimiber of parasitic eggs per nest) of cowbird parasitism. I evaluated the effects of peirasitism on nesting success of wrens by comparing clutch size, brood size and number of wren fledglings in parasitized and unparasitized nests. To test the discriminatory abilities of wrens, and whether they are capable of grasp-ejecting cowbird eggs, I placed real and artificial cowbird eggs in nests during the pre-laying (1-4 days before wrens laid their first egg) and egg-laying stages (Table 6-1). I tested wrens during the pre-laying stage because this is the period when most cowbird eggs are laid (Chapter 4). For artificial eggs I used commercially obtained plastic egg models. These models are hollow and very Ught. To make them heavier I filled them with water (Table 6-1). Models were scored as accepted when they were incubated, and rejected when they disappeared from a nest that was not abandoned or lost to predation. Hypothesis 1 was tested by placing smaD artificial eggs in wren nests. These eggs were only slightly larger than a wren egg but were very contrasting because of their orange-colored markings (Table 6-1). Hypothesis 2 was tested by placing artificial cowbird-sized eggs in wren nests (large artificial eggs in Table 6-1). I placed both light and heavy large eggs to test whether wrens could grasp-eject a large but light egg. To test whether both

PAGE 65

OQ i M o CO Cm O U O o Xi I OQ 2 5 2 Xi I U) ^ — ' CQ CQ CO (N O -H (31 -H (N CD U5 d CD -H -H la 00 od CO 00 -H -H tH (N 00 00 tH X CO (N M O o CO M o CO CO 0) T3 o d as I O O b. •c < bfi CO -a CO IQC CO o OQ OQ OQ Cfl o

PAGE 66

69 size and weight constrained the wren's ability to lift a cowbird egg, I placed glass beads in wren nests. The diameter of glass beads was similar to the breadth of a cowbird egg, but they weighed 23% more (Table 6-1). I expected wrens to eject these odd-looking objects. Therefore, acceptance would indicate that wrens can not lift them. Results Impact of Parasitism Parasitism was extremely common in this population of House Wrens, and nest boxes were more frequently parasitized than natured nests. Rate of parasitism in natural cavities was 58.8% (20 of 34 nests), versus 94% (99 of 105 nests) in nest boxes (G=22.46, df=l, P<.001; Table 6-2). Nest boxes were also more intensely parasitized than natural nests. Eleven of 20 (55%) parasitized nests in natural cavities received more than two cowbird eggs. In contrast, 84 of 105 nests (80%) in boxes received more than two cowbird eggs (G=6.65, df=l, P<.01; Table 6-2). The average number of cowbird eggs per nest was also significantly higher in boxes than in natural cavities (MannWhitney U, P<.001; see Fig. 4-3). Parasitism had a strong negative effect on hatching success of wrens. Only in 55% of parasitized nests did any wrens hatch, compared with 92% of imparasitized nests (Table 6-3). Most wren mortaUty ocurred during incubation. Cowbirds did not remove host eggs, and wren clutch size at the beginning of incubation was not different between parasitized (2.9710.5) and unparasitized nests (3.1210.5; Table 6-4). Clutch size at hatching, however, was significantly reduced in parasitized nests (0.9911.0 vs. 2.8310.9; Table 64). Clutch reduction was due to wren eggs disappearing during incubation,

PAGE 67

60 Table 6-2. Incidence and intensity of Shiny Cowbird parasitism on House Wrens. Number of Nests (%) Number Eggs/Nest Natural Boxes 0 14 (41.2) 6(5.7) 1-2 9 (26.5) 15 (14.3) 3-12 11 (32.3) 84(80.0) x±SD* 1.97 ±2.2 5.65 ±3.2 * Mann Whitney U, P<.001

PAGE 68

61 Table 6-3. Hatching success of wrens (excluding eflfects of predation). Number of nests Hatch. Success* G P Parasitized 20 11 5.9 <.025 Unparasitized 13 12 * Number of nests in which at least one wren hatched.

PAGE 69

62 Table 6-4, Reproductive success of wrens in peirasitized and unparasitized nests (excluding predation). Numbers indicate mean, standard deviation and in parenthesis, sample size. Parasitized Unparasitized p* Initied Clutch Size 2.97 ±0.5 (20) 3.12 ±0.5 (13) .3 Clutch Size at Hatching 0.99 ±1.0 (15) 2.83 ±0.9 (11) <.01 Fledglings/Nest 0.54 ±0.9 (11) 2.78 ±0.7 (9) <.001 * Mann-Whitney U.

PAGE 70

63 probably removed by wrens because they broke when jostled against the heavy, thick-shelled cowbird eggs. During incubation I frequently found dented wren eggs in parasitized nests, and these eggs always disappeared in the next 2 or 3 days. Fledging success of wrens was also significantly reduced by cowbird parasitism. Wrens laid 3.12±0.5 eggs in unparasitized nests and fledged 2.78±0.7 young, while in parasitized nests they laid 2.97±0.5 eggs and fledged 0.54±0.9 young (Table 6-4). Most wren nestlings in parasitized nests died of starvation within 3-4 days of hatching. Only 4 of 11 parasitized nests (36.4%) that survived the entire nesting cycle produced wren fledglings, as opposed to 8 of 9 (88.9%) imparasitized nests (G=6.22, df=l, P<.02). Host resp onses to parasitism Wrens abandoned 92 of 95 nests that received more than two cowbird eggs but showed no rejection at 24 nests that received one or two parasitic eggs (G=100.8, df=l, P<.001). There were three cases in which wrens accepted and incubated large cowbird clutches (two with 6 and one with 7 cowbird eggs), all in nest boxes. Two of these nests were abandoned halfway through incubation. Only one cowbird hatched in the third nest. Two other cowbird eggs in this nest had advanced embryos but did not hatch. Thirteen nests received one cowbird egg and were sucessfully incubated by wrens. In nests parasitized with two cowbird eggs, usually one failed to hatch. Only in two of 11 nests with two cowbird eggs did both cowbirds hatch. In one of these, the second cowbird hatched two days after the first one and died three days later. In the other nest, both cowbirds hatched the same day and 10 days later both were growing normally, but then disappeared.

PAGE 71

64 Probably both would have fledged, because cowbirds usually fledge at 12 days of age. Wrens accepted and incubated 14 real cowbird eggs that I placed in their nests, both before they initiated la)dng and during their laying period (Table 6-5). Wrens also accepted both large and small heavy artificial eggs, but ejected light eggs, except during the laying period, when they were equally likely to accept or reject large light eggs (Table 6-5). This result shows that they can grasp with the bill an object as large as a cowbird egg, £ind that they can discriminate eggs by weight. Wrens also ejected glass beads (Table 6-5), which were heavier than real cowbird eggs, thereby showing that failure to reject real cowbird eggs and small and large heavy artificial eggs was not due to an inability to grasp and lift such eggs. These experiments indicate that wrens have the ability to grasp-eject parasitic eggs, but do not discriminate foreign eggs based on color or size. Potential Costa of Nes t Desertion Several factors could limit the benefits of nest desertion. First, wrens could be limited by a short breeding season. At my study site, however, wrens had an extended breeding season, nesting almost year-round (Chapter 3). A similar pattern was reported by Alvarez et al. (1984) for a site about 10 km north of my study site. Birds that deserted nests because of multiple parasitism always started building a new nest within 2-3 weeks (except when desertion occurred close to the onset of the mid-year dry season). During 1989 and 1990, 10 banded females made 3.5±1.35 nesting attempts per year, and some females made up to six attempts per year. Therefore, length of the breeding season is not likely to be a factor that limits renesting possibilities.

PAGE 72

66 * o V CO o 3 (N 00 CO * Oh (N O V o V o V o V o V U3 O V 05 CO I i ^ 8 CO 00 bi 01 o U DC bj (1) bo 03 T3 CO 0) CO CO CO I La lO

PAGE 73

66 A second potential factor limiting renesting possiblities is availability of nest sites. Because my study site was saturated with nest boxes and all wrens had at least 6 boxes within their territories, I can not tell whether desertion of parasitized nests was contingent upon availability of nest sites, as Petit (1991) did for Prothonotary Warblers (Protonotaria citrea) . Most of the boxes remained in the field for 2.5 years and dtiring this time only one new territory was formed, when in early 1990 a pair of wrens nested in a box placed between two old territories. This suggests that nest site availabiUty was not limiting the establishment of territories. Territories, however, probably varied in the number and quaUty of nesting cavities. Most wrens used both natural cavities and nest boxes during the study, but in some territories wrens never used nest boxes, while in others they always used nest boxes. Several instances of wrens destroying eggs and nestlings of Spectacled Parrotiets (Forpua conspicillatuH) and taking over the nest box, suggest competition for cavities (N. G6mez and G. Kattan, unpubHshed data). In relation to brood parasitism, a good quality cavity is probably one with a small opening. In natural nests, diameter of the entrance of parasitized nests (57.0±9.7 mm, n=12) was significantly larger than the entrance of imparasitized nests (23.0±5.8 mm, n=15; t=8.96, P<.001). All my nest boxes had openings larger than 35 mm, as cowbirds could not enter boxes with smaller openings. Thus, at least in some territories, wrens may be limited by availability of good quality cavities. A third factor that would Hmit the benefits of desertion is probabihty of parasitism of a replacement nest. Because nest boxes were more frequentiy and intensely parasitized than natural nests, I separated the sample of birds renesting after deserting a parasitized nest into birds renesting in boxes and in natural sites. Of 17 birds renesting in boxes, all were parasitized, while

PAGE 74

67 one of six birds that renested in natural cavities were parasitized (G=13.6, df=l, P<.01). Parasitism rate of replacement nests in natural cavities was significantly lower than the 59% overall parasitism rate in natural cavities (G=3.88, df=l, P<.05). After deserting a multiply parasitized nest, birds became wary and difficult to observe, and took a long time to renest (interval between desertion and initiation of a new clutch, x=31.1 ± 14.8 days, n=23; sample does not include nests abandoned just prior to the beginning of the dry season). In contrast, birds that abandoned for other reasons (himian interference, nest flooding, predation) renested in 13.1 days (SD=7.2, n=18; Mann Whitney U, P<.01). Discussion Shiny Cowbird parasitism substantially depressed the nesting success of House Wrens. Most of the cost of parasitism was paid during incubation, probably because wren eggs broke when jostled against the thick shelled cowbird eggs. Disappearance of wren eggs during incubation resulted in a hatching success of only 55% in parasitized nests, as opposed to 92% in unparasitized nests (Table 6-2). Cowbird parasitism also reduced the fledging succes of wrens. Wrens that hatched soon died of starvation, outcompeted by the more aggressively begging cowbird nestling (Gochfeld 1979; pers. obs.). Wrens fledged in only 36% of parasitized nests, while 90% of unparasitized nests produced fledglings. The frequency of cowbird parasitism reported in this study is very high for a cavity nester. Petit (1991), for example, reported an incidence of parasitism of 21% for the Prothonotary Warbler in Tennessee, USA, in both nest boxes and natural cavities. Shiny Cowbirds are very abundant in the Cauca Valley,

PAGE 75

68 a region with extensive plantations of rice and sorghum, two crops on which they feed. In contrast to natural nests, the nest boxes I used in this study were conspicuous and revealed the high potential of cowbird parasitism. Probably a Isirge proportion of wren nests in the Cauca Vedley are foimd by cowbirds, and the only defense wrens have is to nest in cavities with small openings. Woodward and Woodward (1979) also reported high levels of parasitism on a population of Eastern Bluebirds (Sialia sialis) in Virginia, and they also attributed it to the conspicuoiisness of nest boxes, large size of openings and high density of cowbirds. Although in some cases cowbird parasitism may have little impact (Smith 1981; Wheatherhead 1989), in most cases it is detrimental to the reproductive success of the host. Detailed studies of both the Brown-headed Cowbird and the Shiny Cowbird show that parasitized birds produce less yoimg than unparasitized birds (e. g., Klaas 1975; Fraga 1985; Wiley 1985). The magnitude of the impact is variable and losses may occur at the incubation stage or at the nestling stage. Brown-headed Cowbirds are reported to habitually remove a host egg when they lay (Sealy 1992), a habit that has silso been reported for some populations of Shiny Cowbirds (e. g., Fraga 1985). At my study site I found no evidence of cowbirds removing wren eggs, and very rarely did they puncture wren or other cowbird eggs. Instead, losses during incubation probably occurred because host eggs broke when jostled against cowbird eggs, as has also been reported in other studies (Sick 1958; Blankespoor et al. 1982). Host eggs may also fail to hatch because of improper incubation (e. g., Klaas 1975; Petit 1991). Parasitism also reduces fledging success of the host because nestlings are imable to compete with cowbirds. This occurs primarily with small hosts (e. g., Fraga 1978, 1983; Marvil and Cruz 1989). Large hosts are better able to compete, and thus impact at this

PAGE 76

69 stage is less severe for them (Fraga 1985; Weatherhead 1989; R0skaft et al. 1990). Given that cowbird parasitism has a negative impact, wrens are expected to exhibit defenses. Wrens rejected parasitism by abandoning multiply parasitized nests but accepted when parasitized with only one or two cowbird eggs. This response could be explained if the cost of parasitism was an increasing function of the nimiber of cowbird eggs in the nest. It could be argued that perhaps when psirasitized with one or two cowbird eggs, the cost of rejecting is higher than the cost of accepting. This was not the case, however, because losses were total for wrens parasitized with two or more cowbird eggs. Even when parasitized with only one cowbird egg, the cost of accepting was probably higher than the cost of rejecting. Wrens pEirasitized with one cowbird egg have the option of rejecting parasitism, either by ejecting the egg or abandoning the nest. Both of these two alternative modes of rejection have potential costs, which must be balanced against the costs of accepting a peirasitic egg. Acceptance of the parasitic egg would be aqdaptive only if losses due to parasitism are not total and if the reproductive success of individuals that accept is higher than the success of individuals that reject. This may occur only for certain hosts for which nest success is not severely depressed by parasitism. For hosts that suffer very high losses, such as House Wrens (see also Rothstein 1975a; Klaas 1978; Fraga 1978, 1983), almost any option should be better than accepting parasitism. Besides, wrens should be able to grasp-eject cowbird eggs. Empty plastic eggs were ejected, probably because they were perceived as empty shells, and this shows that wrens have the abiHty to discriminate eggs by weight and grasp with the bill an object as large as a cowbird egg. Ejection of glass beads indicated that wrens can eject objects as large and heavy as

PAGE 77

70 cowbird eggs. Despite this ability, wrens never ejected real cowbird eggs or artificial eggs of the same size and mass. They also accepted small eggs despite their contrasting colors. These experiments indicate that wrens do not recognize cowbird eggs as foreign. Ortega and Cruz (1988) also showed experimentally that the Red-winged Blackbird (Agelaiiis phoeniciua) . an accepter species, had the ability to reject objects as large as a cowbird egg. They argued that redwings accept because the frequency and consequences of cowbird parasitism were too slight to be a significant selection pressure. They further suggested that if there are costs associated with ejection that they were unable to detect, selection might even favor acceptance over rejection (see also R0skaft et al. 1990). This is not likely to be true for wrens because the cost of parasitism was very high. Nest abandonment is another option for wrens. As cavity nesters, wrens may be Umited by the availabihty of nest sites. Acceptance of cowbird eggs by Prothonotary Warblers was foimd to be dependent on the opportimity to renest (Petit 1991). In this case, nest site Umitation and a short breeding season made it more adaptive for females to accept the relatively low cost of parasitism (Petit 1991), At my study site, however, it is unlikely that these factors Hmit the benefits of nest desertion as opposed to the cost of accepting parasitism. Although wrens in some territories may be limited by the availability of good nesting sites, the long breeding season should make it more adaptive for a female to abandon and make a renesting attempt later. This contention is supported by the fact that wrens readily abandoned multiply-parasitized nests, and replacement natural nests were less likely to be parasitized. Furthermore, the cost of accepting cowbird eggs was very high, i. e., 16 of 20 nests that were parasitized before the initiation of incubation failed to fledge any host young because of parasitism. When a host

PAGE 78

71 experiences complete reproductive failure, as occurs with all hosts of the Common Cuckoo (Davies and Brooke 1989a) any form of rejection is likely to be favored by selection. In such circmnstances, the only factor that can make acceptance more adaptive than rejection is the occurrence of costs when a host nest is not parasitized (Rothstein 1990). While this may explain acceptance by cuckoo hosts, which typically experience low rates of parasitism (Davies and Brooke 1989a, 1989b), it is unlikely to do so in the cowbird-wren system I studied as most nests (58.8%) in natural cavities were parasitized. Given this evidence, it is intriguing why House Wrens do not reject cowbird eggs. The selective advantage of rejection depends on the frequency and cost of parasitism (Rothstein 1975b; Kelly 1987; Davies and Brooke 1989b). If either one or both factors are low, individuals exhibiting the rejection response may have Uttle or no advantage. However, if both factors are high, as is the case for House Wrens, the rejection response would be expected to rapidly become common in the population (Rothstein 1975b). One factor that may explain lack of rejection in House Wrens is that egg discrimination may be dificult inside a dark cavity. However, Mason and Rothstein (1986) found that Rufous Homeros (Fumarius rufus) discriminated cowbird eggs based on size. Also, in my study wrens rejected when parasitized with three or more cowbird eggs. The stimulus for desertion when three or more cowbird eggs are laid may be the total clutch volume (or a high frequency of cowbird intrusions). Because there is no overlap in egg sizes of cowbirds and wrens, and my experiments indicate that wrens can discriminate eggs by weight, rejection on the basis of size or weight alone should incur little or no cost. Thus, evolutionary lag, i. e., the absence of genetic variants with the ability to recognize foreign eggs (Rothstein 1982,

PAGE 79

72 1990), seems to be the best explanation for acceptance of parasitic eggs in House Wrens.

PAGE 80

CHAPTER 7 GENERAL DISCUSSION Shiny Cowbirds in the Cauca Valley are extremely fecund. I estimate that each female lays approximately 120 eggs per year, eq\aivalent to about 9.4 times her own body mass. Studies on both the Brown-headed Cowbird and the Shiny Cowbird indicate that they lay in a pattern that resembles that of domestic chickens that are not allowed to brood (Scott and Ankney 1983; Jackson and Roby 1992; this study). They lay almost continuously, without regressing the ovary and oviduct, during the entire breeding season. In the temperate zone, most female cowbirds lay throughout the short breeding season. In the Cauca Valley, the Shiny Cowbird population has an extended 9-month breeding season. Diiration of the season for individual females is unknown, but my data suggest that females are reproductively active throughout most of the season, with only one interruption for molting. As predicted by life history theory (Roff 1992), the high fecimdity exhibited by cowbirds involves a trade-off. Laying a high nmnber of eggs impHes investing less in each egg. I divided investment into two components: A direct energetic investment, represented in the amoimt of energy deposited in each egg, and a behavioral investment represented in the process of nest selection. Shiny Cowbirds in the Cauca Valley lay eggs with an energy content lower than would be expected from egg mass (14.54 kJ versus the expected 19.05 kJ). This represents a saving of 24% of the energy that otherwise would be required by the cowbird. Although it is not clear whether the high number of eggs laid represent any physiological strain for the cowbird (Rothstein et al. 73

PAGE 81

74 1986), it is conceivable that the energy saved allows the cowbird to lay more eggs than would be possible if each had a higher energy content. Shiny Cowbirds also invest little energy in the carefial placing of each egg in circumstances that would maximize the egg's chances of being accepted and incubated successfully by the host. This would involve monitoring nests in order to lay the egg in coincidence with the host's laying period, as well as having some mechanism to avoid competition from other female cowbirds, such as defending laying territories. These two conditions would probably result in the cowbird being able to lay fewer eggs. Cowbirds monitor host nests, £uid to a certain extent choose nests that are £ui appropriate stage for parasitism. Egg la5dng by cowbirds is clumped around its host's laying period, with 47% of the eggs laid during the three day laying period of wrens (Fig. 22). However, a large proportion of cowbird eggs is laid before wrens lay, sometimes so early during nest construction that these eggs are buried as wrens continue adding nesting material. In addition, cowbirds do not avoid multiply parasitized nests. These results indicate that Shiny Cowbirds in the Cauca Vsdley are following the "shotgun" strategy. Instead of carefully placing a few eggs, they direct their energy into egg production and perm£inently search for nests, parasitizing as many as they can. This strategy is made possible by the imselective behavior of wrens. Wrens accept cowbird eggs laid at any time before, during or after the laying period. This allows cowbirds a wide window of time in which to lay eggs. The short incubation period and the size advantage of the cowbird nestling over its nestmates contribute to the success of this strategy. This study concentrated on the interaction between cowbirds and one of their hosts in the Cauca Valley. Cowbirds, however, are generalists.

PAGE 82

75 parasitizing most passerines with which they occur (e. g., Cruz et al. 1985, Wileyl885; Mason 1986a). Features of the parasitic strategy of cowbirds Eire related to their generalist habits. Laying large numbers of eggs requires large numbers of nests. Although there is conflicting evidence regarding female spacing patterns, it appeeirs that female cowbirds have Isirge home ranges that overlap extensively (Rothstein et al. 1986), as shown by patterns of multiple parasitism in this and other studies (e. g., Fleischer 1985). The generalist strategy of cowbirds contrasts sharply with the host specificity exhibited by the Common Cuckoo. At a given locality cuckoos psirasitize only a few species, and individual females are host-specific. Populations of cuckoos are segregated into races or genetic strains ("gentes"), with each "gens" laying eggs that are mimetic to a specific host. Egg mimicry is a consequence of most cuckoo hosts being rejectors (Rothstein 1990). Hosts not only reject non-mimetic eggs, but also reject non-synchronized eggs (Davies and Brooke 1988). As a consequence, fenmle cuckoos have to invest time and energy in selecting a nest and parasitizing it at a certsiin time. This suggests that it would be advantageous for females to have exclusive use of a laying range. As with cowbirds, there is conflicting evidence regarding the spacing system of female cuckoos, but it appears that females have restricted £uid separate laying ranges, although there is overlap (WylUe 1981). Because cuckoos have to invest more than cowbirds in carefiilly placing each egg, they have comparatively lower fecundity. Estimates of the number of eggs laid by cuckoos in Britain range between 10 and 25 (Rothstein 1990). This is higher than what nesting birds would produce, but not as impressive when compared with cowbird production. Current theory on brood parasite-host coevolution suggests that parasites become specialized as hosts evolve defenses against parasitism (Rothstein

PAGE 83

76 1990). It is assumed that cowbirds are recently evolved parasites that will eventimlly become host-specific (Rothstein 1990). Will cowbirds, given enough time, evolve cuckoo-like adaptations? Davies and Brooke (1989b) proposed a model to explain events in an arms race between parasite and host (see also Rothstein 1990). When a host popvdation begins to be parasitized, egg rejection is favored by selection. The rate of spread of the rejection response depends on its adaptive value, determined by impact and frequency of psirasitism. As the rejection response spreads in the population, selection will favor egg mimetism. Several outcomes are possible sifter this stage. For example, egg discrimination may become more refined, or parasitism with mimetic eggs may become too frequent, driving the host to extinction. This model assumes that genetic vEiriation for egg rejection is present in the host population (Davies and Brooke 1989b; Rothstein 1990). According to this model, different parasites are at different stages in the sequence. Cuckoos are at an advanced stage. Most Common Cuckoo hosts show intermediate to high levels of rejection. As a consequence, female cuckoos lay mimetic eggs and specialize on a single host. Cowbirds are assumed to have evolved parasitism more recently than cuckoos, mainly because there are about 50 species of parasitic cuckoos and only five cowbirds (Rothstein 1990). Thus, cowbirds are at an early stage in the sequence of events leading to specialization. This hypothesis assumes that there has been no time for the spread of rejection in host populations. As rejection spreads, cowbirds should evolve egg mimicry and specialize on one or a few hosts (Harvey and Partridge 1988; Rothstein 1990). Davies and Brooke's model assimies there is a single pathway for the evolution of brood parasitism. There is a "progression" fi-om a generalist to a

PAGE 84

77 specialist strategy (Harvey and Partridge 1988; Rothstein 1990). This model implies that all avian brood parasites, despite their different origins, should converge on the same evolutionary path. Obligate interspecific brood peirasitism has arisen independently in three orders of altridal birds: Cuculiformes, Piciformes and Passeriformes. In the Cuculidae, brood parasitism occurs in two different subfamilies, the Old World Cuculinae and the New World Neomorphinae (in the latter only three of 11 species are peirasitic). Among passerines, it evolved independently in two subfamilies in different families: Icterinae (Emberizidae) and Viduinae (Ploceidae). Strategies of parasitism in these lineages differ in a variety of aspects, such as degree of host specificity (Payne 1977; Rothstein 1990). Given the polyphiletic origin of brood parasitism, therefore, it is conceivable that there are optional evolutionary routes, not necessarily following the generalist-speciaHst progression. Brood parasites may remain generalists, or start as specialists and remain specialists or become generalists. A recent phylogeny of the cowbirds, based on mitochondrial DNA (Lanyon 1992), suggests that cowbirds are monophiletic and that the specialist Screaming Cowbird ( Molothrus mfnayillfiris which only parasitizes the non-parasitic Bay-winged Cowbird, Mhadilia; Fraga 1986) is primitive, while the generalist Shiny and Brown-headed Cowbirds £u*e the most derived. This phylogeny, therefore, suggests a pathway opposite to the generalistspecialist progression. Another recent phylogeny differs fi'om Lanyon's and suggests even different pathways. Freeman (in press) constructed a phylogeny of 47 species in the subfamily Icterinae, based on mitochondrijd DNA, In this phylogeny, brood parasitism evolved independently at least three times. The specialist Screaming Cowbird is distantly related to the other cowbirds, suggesting that

PAGE 85

78 the similarity in eggs and nestling plumage with its host, the Baywinged Cowbird, is related to mimicry (Fraga 1986; Lanyon 1992). The Giant Cowbird (Scaphidura orvzivora) also evolved in a non-parasitic lineage, independently of the three generalist species (Brown-headed, Shiny £ind Bronzed Cowbirds), which share a common ancestor. Both Lanyon's and Freeman's results require cautious interpretation, because their phylogenies are simply the most peirsimonious of a series of possible phylogenies (Freeman, pers. com.). They suggest, however, that multiple pathways are possible in brood parasitism and that being a speciaUst is not necessarily a derived condition. There may even be different pathways for the evolution of egg mimicry. Davies and Brooke's (1989b) model assumes that the selection pressure for egg mimicry is host rejection. Brooker and Brooker (1989, 1990) suggested an alternative pathway for the evolution of egg mimicry and host specisilization. Although female cuckoos tend to have non-overlapping laying areas, some overlap occurs, particularly in areas where more than one cuckoo species occur. There is evidence that many cuckoo species habitually remove an egg when laying their own. If a second cuckoo parasitizes a nest already cont£iining a cuckoo egg, there is a chance she will remove the first cuckoo egg if it has contrasting colors (with the backgroimd or with host eggs). Thus, interspecific competition could explain the evolution of egg traits (mimicry, crypsis and small size) and host specialization in cuckoos (Brooker and Brooker 1990). The model can be extended to explain egg polymorphism in the Common Cuckoo by intraspecific competition (Brooker and Brooker 1990). While this model presents an alternative h)T)othesis for the evolution of egg mimicry, it still remains a fact that some cuckoo hosts accept mimetic eggs and reject non-mimetic eggs, and the role of their behavior in the evolution of

PAGE 86

79 this trait can not be ignored. The two hjrpothesis are not mutually exclusive and may operate simultaneously. The fact that a variety of factors may be selecting for brood parasitism to evolve egg mimicry and host specialization prompts the question, how can cowbirds remain generalists? If egg rejection by hosts is the main selective force conducing to host specialization, then the reproductive strategy of cowbirds depends on host responses. With a preponderance of acceptor species, the current strategy of parasitizing all nests foimd is optimtun, because it may be impossible for cowbirds to recognize and avoid rejectors without eJso avoiding some suitable hosts (Rothstein 1976). The decrease in reproductive succes resulting from parasitizing a few rejectors is probably less than what would result if cowbirds fail to parasitize some suitable hosts (Rothstein 1976). The question then becomes, will cowbird hosts become rejectors? Ample evidence suggests that lack of rejection in some hosts is best explained by absence of genetic variants that exhibit a rejection response (Rothstein 1975a, 1982; this study). Thus, the system depends on stochastic processes (appeeirance of mutations or recombinations that code for rejection) that are impossible to predict (Rothstein 1982). As long as these variants do not appear, hosts will remain acceptors, allowing cowbirds to remain generalists. Even if the genetic basis for rejection is present in a population, the costof-rejection hypothesis provides a scenario for hosts not expressing this behavior. One critical difference between cowbirds and cuckoos is the cost of parasitism for the host. For cuckoo hosts, the cost of parasitism is always very high because their losses are total, as cuckoo hatchlings eject host eggs from the nest (Wyllie 1981). Because cowbirds do not kill nestmates, costs for cowbird hosts are variable. For some hosts, such as House Wrens, costs may

PAGE 87

80 be high, but for other hosts costs may be intermediate to low (Rothstein 1990). If the costs of accepting parasitism are lower then the costs of rejecting, hosts will continue to accept cowbird eggs, allowing cowbirds to continue parasitizing that population with non-mimmetic and nonsynchronized eggs. In conclusion, it is likely that being a generalist is a successful strategy that can be maintained indefinitely, and not necessarily a stage in the evolution of host-specificity. Cowbirds direct their reproductive effort into egg production, rather than expending much energy in each egg. This strategy is possible because most cowbird hosts are acceptors (Rothstein 1990). Reproductive success of cowbirds, expressed in number of offspring produced per year, is a function of the probability of success of each egg multiplied by annual fecimdity. If hosts become discriminative and reject non-S5nichronized eggs, cowbirds would be forced to expend more energy in ensuring a better placement of each egg. This could result in an increase in the probability of success of each egg, but cowbirds would have to lay fewer eggs. Differences between the strategies of cowbirds and cuckoos probably reflect different balances of this trade-off.

PAGE 88

LITERATURE CITED Alvarez, H., M. D. Heredia, and M. C. Hem^dez. 1984. Reproduccion del Cucarachero Comun ( Troglodytes aedon) en el VaUe del Cauca. Caldasia 14: 86-123. Ankney, C. D. and D. M. Scott. 1980. Changes in nutrient reserves and diet of Brown-headed Cowbirds. Auk 97: 684-696. Balinsky, B. I. 1975. An introduction to embryology. Philadelphia: W. B. Sanders Co. Blankespoor, G. W., J. Oohnan, and C. Uthe. 1982. Eggshell strength and cowbird parasitism of Redwinged Blackbirds. Auk 99: 363-365. Briskie, J. V. and S. G. Sealy. 1990. Evolution of short incubation periods in the parasitic cowbirds, Molothrus spp. Auk 107: 789-794. Brooker, L. C. and M. G. Brooker. 1990. Why are cuckoos host specific? Oikos 57: 301-309. Brooker, M. G. and L. C. Brooker. 1989. The comparative breeding behavior of two sympatric cuckoos, Horsfield's Bronze-cuckoo Chrvsococcvx basalia. and the Shining Bronze-cuckoo Q. lucidus . in western Australia: A new model for the evolution of egg morphology find host specificity in avian brood parasites. Ibis 131: 528-547. Carey, C. 1986. Possible manipulation of eggshell conductance of host eggs by Brown-headed Cowbirds. Condor 88: 388-390. Carter, M. D. 1986. The parasitic behavior of the Bronzed Cowbird in south Texas. Condor 88: 11-25. Cavalcanti, R. B. and T. M. Pimentel. 1988. Shiny Cowbird parasitism in central Brazil. Condor 90: 40-43. Clark, K. L. and R. J. Robertson. 1981. Cowbird parasitism and the evolution of antiparasite strategies in the Yellow Warbler. Wilson Bull. 92: 244258. Cruz, A., T. Manolis, and J. W. Wiley. 1985. The Shiny Cowbird: A brood parasite expanding its range in the Caribbean region, pages 607-620 in P. A. Buckley et al. (eds.), Neotropical Ornithology. Washington, D. C: American Ornithologists' Union, 81

PAGE 89

82 Darley, J. A. 1983. Territorial behavior of female Brown-headed Cowbirds. Canadian Journal of Zoology 61: 65-69. Davies, N. B. and M. L. Brooke. 1988. Cuckoos vs. reed warblers: Adaptations £uid counteradaptations. Animal Behaviour 36: 262-284. Davies, N. B. and M. L, Brooke. 1989a. An experimental study of coevolution between the cuckoo Cuculus canorus and its hosts. I. Host egg discrimination. Journal of Animal Ecology 58: 207-224. Davies, N. B. and M. L. Brooke. 1989b. An experimental study of coevolution between the cuckoo, Cuciilus canorus , and its hosts. II. Host egg markings, chick discrimination and general discussion. Journal of Animal Ecology 58: 225-236. Davis, D. E. 1942. The number of eggs laid by cowbirds. Condor 44: 10-12. Dufty, A. M. 1983. Variation in the egg markings of the Brown-headed Cowbird. Condor 85: 109-111. Finke, E. O., D. Valencia, and W. D. McKay. 1979. Evaluaci6n de daiios por pdjaros en cultivos de arroz en Colombia. Instituto Colombiano Agropecuario, Palmira. Fleischer, R. C. 1985. A new technique to identify and assess the dispersion of eggs of individual brood parasites. Behavioral Ecology and Sociobiology 17: 91-99. Fraga, R. M. 1978. The Rufous-collared Sparrow as a host of the Shiny Cowbird. Wilson Bulletin 90: 271-284. Fraga, R. M. 1983. Parasitismo de cria del Renegrido Molothrus bonariensis sobre el Chingolo Zonotrichia caoenais : Nuevas observaciones y conclusiones. Homero 12 (Niimero extraordinario): 245-255. Fraga, R. M. 1985. Host-parasite interactions between Chalk-browed Mockingbirds and Shiny Cowbirds. pages 829-844 in P. A. Buckley et al. (eds.). Neotropical Ornithology. Washington D.C.: American Ornithologists' Union. Fraga, R. M. 1986. The Bay-winged Cowbird (Molothrus badius) and its brood parasites: Interactions, coevolution and comparative efficiency. Ph. D. Diss., University of California, Santa Barbara. Freeman, B. M. and M. A. Vince. 1974. Development of the Avian Embryo. London: Chapman and Hall.

PAGE 90

83 Freeman, S. in press. The phylogeny of the blackbirds estimated from restriction sites in mitochondrial DNA. Systematic Biology Friedmann, H. 1927. A case of apparently adaptive acceleration of embryonic growth rate in birds. Biological Biilletin 53: 343-345. Friedmann, H. 1929. The Cowbirds, A Study in the Biology of Social Parasitism. Springfield, 111.: C. C. Thomas. Friedmann, H. 1948. The parasitic cuckoos of Africa. Washington Academy of Sciences Monograph 1: 1-204. Friedmann, H. 1955. The honeyguides. US National Museum Bulletin 1-292. Friedmann, H. 1960. The parasitic weaverbirds. US National Musemn Bulletin 1-196. Friedmann, H. 1963. Host relations of the parasitic cowbirds. US National Museum Bulletin Friedmann, H. and L. F. Kiff. 1985. The parasitic cowbirds and their hosts. Proceedings of the Western Foimdation of Vertebrate Zoology 2: 227302. Friedmann, H., L. F. Kiff, and S. I. Rothstein. 1977. A further contribution to knowledge of host relations of the cowbirds. Smithsonian Contributions to Zoology Gochfeld, M. 1979a. Begging by nestling Shiny Cowbirds: Adaptive or maladaptive? Living Bird 17: 40-50. Gochfeld, M. 1979b. Brood parasite and host coevolution: Interaction between Shiny Cowbirds and two species of meadowlarks. American Naturalist 113: 855-870. Goldsmith, O. 1774. A history of the earth and animated nature. Gould, S. J. and E. S. Vrba. 1982. Exaptation: A missing term in the science of form. Paleobiology 8: 4-15. Hamilton, W. J. and G. H. Orians. 1965. Evolution of brood parasitism in altricial birds. Condor 67: 361-382. Harvey, P. H., and L. Partridge. 1988. Of cuckoo clocks and cowbirds. Nature 335: 586-587. Hoar, W. S. and C. P. Hickman. 1983. A Laboratory Companion for General and Comparative Physiology. Englewood Cliffs, N.J.: Prentice Hall.

PAGE 91

84 Hoy, G. and J. Ottow. 1964. Biological and oological studies of the molothrine cowbirds (Icteridae) of Argentina. Auk 82: 186-203. Jackson, N. H. and D. D. Roby. 1992. Fecundity and egg-laying patterns of captive yearling Brown-headed Cowbirds. Condor 94: 585-589. Kelly, C. 1987. A model to explore the rate of spread of mimicry and rejection in hypothetical populations of cuckoos and their hosts. Journal of Theoretical Biology 125: 283-299. King, J. R. 1973. Reproductive relationships of the Rufous-collared Sparrow and the Shiny Cowbird. Auk 90: 19-34. Klaas, E. E, 1975. Cowbird parasitism and nesting success in the eastern phoebe. Occassional Papers of the Museum of Natural History of the University of Kansas 41:1-18. Lanyon, S. M. 1992. Interspecific brood parasitism in blackbirds (Icterinae): A phylogenetic perspective. Science 255: 77-79. Levey, D. J. and F. G. Stiles, in press. Birds of La Selva: Ecology, Behavior and Taxonomic Affinities. Marvil, R. E. and A. Cruz. 1989. Impact of Brown-headed Cowbird psirasitism on the reproductive success of the Solitary Vireo. Auk 106: 476-480. Mason, P. 1985. The nesting biology of some passerines of Buenos Aires, Argentina, in P. A. Buckley et al. (eds.), Neotropical Ornithology. Washington, D. C: American Ornithologists' Union. Mason, P. 1986a. Brood parasitism in a host generalist, the Shiny Cowbird: I. The quality of difierent species as a host. Auk 103: 52-60. Mason, P. 1986b. Brood parasitism in a host generaUst, the Shiny Cowbird: II. Host selection. Auk 103: 61-69. Mason, P. and S. I. Rothstein. 1986. Coevolution and avian brood parasitism: Cowbird eggs show evolutionary response to host discrimination. Evolution 40: 1207-1214. McLean, I. 1987. Response to a dangerous enemy: Should a brood parasite be mobbed? Ethology 75: 235-245. Mendez de la Cruz, F., L. J. Guillette and M. Villagr^ Santa Cruz. In press. Differential atresia of ovarian foUicles and its effect on the clutch size of two populations of the viviparous lizard Sceloporus mucronatus . Functional Ecology.

PAGE 92

85 Miller, A. H. 1961. Molt cycles in equatorial Andean sparrows. Condor 63: 143-161. Montgomery, D. C. 1984. Design and Analysis of Experiments. New York: John Wiley and Sons. Nice, M. M. 1953. The question of ten-day incubation periods. Wilson Bulletin 65: 81-93. Norman, R. F. and R. J. Robertson. 1975. Nest-searching behavior in the Brown-headed Cowbird. Auk 92: 610-611. Orians, G. H., E. R0skaft, and L. D. Beletzky. 1989. Do Brown-headed Cowbirds lay their eggs at random in the nests of Redwinged Blackbirds? Wilson Bulletin 101: 599-605. Ortega, C. P. and A. Cruz. 1988. Mechanisms of egg acceptance by marshdwelling blackbirds. Condor 90: 349-358. Payne, R. B. 1965. Clutch size and nvmiber of eggs laid by Brown-headed Cowbirds. Condor 67: 44-61. Payne, R. B. 1967. Gonadal responses of Brown-headed Cowbirds to long daylengths. Condor 69: 289-297. Payne, R. B. 1972. Mechanisms and control of molt, pages 104-155 in D. S. Earner and J. R. King (eds.), Avian Biology, Volume 2. New York: Academic Press. Payne, R. B. 1973. The breeding season of a parasitic bird, the Brown-headed Cowbird, in central California. Condor 75: 80-99. Payne, R. B. 1974. The evolution of clutch size and reproductive rates in parasitic cuckoos. Evolution 28: 169-181. Payne, R. B. 1976. The clutch size and numbers of eggs of Brown-headed Cowbirds: Effects of latitude and breeding season. Condor 78: 337-342. Petit, L. J. 1991. Adaptive tolerance of cowbird parasitism by Prothonotary Warblers: A consequence of nest site limitation? Animal Behaviour 41* 425-532. Phillips, J. G., P. J. Butler and P. J. Sharp. 1985. Physiological strategies in avian biology. Glasgow: Blackie and Son, Ltd. Post, W., A. Cruz and D. B. McNair. 1993. The North American invasion pattern of the Shiny Cowbird. Journal of Field Ornithology 64: 32-41.

PAGE 93

86 Post, W. and J. W. Wiley. 1977. Reproductive interactions of the Shiny Cowbird and the Yellow-shouldered Blackbird. Condor 79: 176-184. Poulin, B., G. Lefebvre and R. McNeil. 1992. Tropical avian phenology in relation to abundance and exploitation of food resources. Ecology 73: 2295-2309. Rahn, H. and A. Ar. 1974. The avian egg: Incubation time and water loss. Condor 76: 147-152. Rahn, H., C. V. Paganelh, and A. Ar. 1975. Relation of avian egg weight to body weight. Auk 92: 750-765. Ramo, C. and B. Busto. 1981. La reproduccidn de un ave par^sita: El tordomirlo (Molothrus bonariensia) en los llanos de Apure (Venezuela). Donana Acta Vertebrata 8: 215-224. Ricklefs, R. E. 1987. Comparative an2Jysis of avian embryionic growth. Journal of Experimental Zoology Supplement 1: 309-323. Robertson, R. J. and R. F. Norman. 1977. The fimction and evolution of aggressive behavior towards the Brown-headed Cowbird. Canadian Journal of Zoology 55: 508-518. RofF, D. A. 1992. The Evolution of Life Histories. Theory and Analysis. New York: Chapman and Hall. Rohwer, S. and C. D. Spaw. 1988. Evolutionary lag versus bill-size constraints: A comparative study of the acceptance of cowbird eggs by old hosts. Evolutionary Ecology 2: 27-36. Rohwer, S., C. D. Spaw, and E. R0skaft. 1989. Costs to Northern Orioles of pimcture-ejecting parasitic cowbird eggs from their nests. Auk 106: 734-738. R0skaft, E., G. H. Orians, and L. D. Beletsky. 1990. Why do Red-winged Blackbirds accept eggs of Brown-headed Cowbirds? Evolutionary Ecology 4: 35-42. Rothstein, S. I. 1975a. Evolutionary rates and host defenses against avian brood parasitism. American Naturahst 109: 161-176. Rothstein, S. I. 1975b. An experimental and teleonomic investigation of avian brood parasitism. Condor 77: 250-271. Rothstein, S. I. 1976. Cowbird parasitism of the Cedar Waxwing and its evolutionary impUcations. Auk 93: 498-509.

PAGE 94

87 Rothstein, S. I. 1982. Successes and failures in avian egg recognition with comments on the utiHty of optimality reasoning. American Zoologist 22: 547-560. Rothstein, S. I. 1990. A model system for coevolution: Avi£in brood parasitism. Annual Review Ecology Systematics 21: 481-508. Rothstein, S. I., D. A. Yokel, and R. C. Fleischer. 1986. Social dominance, mating and spacing systems, female fecundity, and vocal dialects in captive and free-ranging Brown-headed Cowbirds. pages 127-185 in R. F. Johnston (ed.). Current Ornithology. New York: Plenum Press. Salvador, S. A. 1983. Parasitismo de cn'a del Renegrido (Molothrus bonariensia) en Villa Maria, C6rdoba, Argentina. Historia Natural 3: 149-158. Scott, D. M. 1978. Using sizes of unovulated folHcles to estimate the laying rate of the Brown-headed Cowbird. Canadian Journal of Zoology 56: 2230-2234. Scott, D. M. and C. D. Ankney. 1979. Evaluation of a method for estimating the laying rate of Brown-headed Cowbirds. Auk 96: 483-488. Scott, D. M. and C. D. Ankney. 1980. Fecimdity of the Brown-headed Cowbird in southern Ontario. Auk 97: 677-683. Scott, D. M. and C. D. Ankney. 1983. The laying cycle of Brown-headed Cowbirds: Passerine chickens? Auk 100: 583-592. Sealy, S. G. 1992. Removal of Yellow Warbler eggs in association with cowbird parasitism. Condor 94: 40-54. Seppa, J. 1969. The cuckoo's abiUty to find a nest where it can lay an egg. Omis Fennica 46: 78-79. Sick, H. 1958. Notas bioldgicas sSbre o Gaud^rio Molothrus bonariensis (Gmelin) (Aves, Icteridae). Revista Brasileira Biologia 18: 417-431. Sinervo, B. 1990. The evolution of maternal investment in lizards: An experimental and comparative analysis of egg size and its effects on offspring performance. Evolution 44: 279-294. Smith, J. N. M. 1981. Cowbird parasitism, host fitness and age of the host female in an island Song Sparrow population. Condor 83: 152-161. Teather, K. L. and R. J. Robertson. 1986. Pair bonds and factors influecing the diversity of mating systems in Brown-headed Cowbirds. Condor 8863-69.

PAGE 95

88 Thompson, C. F, and B. M. Gottfried. 1976. How do cowbirds find and select nests to parasitize? Wilson Bulletin 88: 673-674. Thompson, C. F. and B. M. Gottfried. 1981. Nest discovery and selection by Brown-headed Cowbirds. Condor 83: 268-269. Vleck, C. M. and D. Vleck. 1987. Metabolism and energetics of avian embryos. Journal of Experimental Zoology Supplement 1: 111-125. Vleck, D., C. M. Vleck, and D. F. Hoyt. 1980. Patterns of metabohsm and growth in avian embryos. American Zoologist 20: 405-416. Weatherhead, P. J. 1989. Sex-ratios, host-specific reproductive success and impact of Brown-headed Cowbirds. Auk 106: 358-366. Wiley, J. W. 1985. Shiny Cowbird parasitism in two avian commimities in Puerto Rico. Condor 87: 167-176. Wiley, J. W. 1988. Host selection by the Shiny Cowbird. Condor 90: 289-303. Wiley, R. H. and M. S. Wiley. 1980. Spacing and timing in the nesting ecology of a tropical blackbird: Comparison of populations in different environments. Ecological Monographs 50: 153-178. Woodward, P. W. and J. C. Woodward. 1979. Brown-headed Cowbird parasitism on Eastern Bluebirds. Wilson Bulletin 91: 321-322. Worthington, A. 1982, Population sizes and breeding rythms of two species of manakins in relation to food supply, pages 213-225 in E. G. Leigh, A. S. Rand and D. M. Windsor (eds.), The Ecology of a Tropical Forest: Seasoned Rythms and Long Term Changes. Washington, D.C.: SmithsoniEin Institution Press. WylUe, I. 1981. The Cuckoo. New York: Universe Books.

PAGE 96

BIOGRAPHICAL SKETCH Gustavo Kattan was born on 26 July 1953 in Cali, Colombia. He attended the Universidad del Valle, where he graduated as a zoologist in 1983, working under the direction of Dr. Humberto Alvarez. In 1984 he married his wife Carolina Murcia, a fellow biologist, and started graduate school at the University of Florida. He obtained his Master's degree in 1987 imder the direction of Dr. Harvey Lillywhite and continued in the Ph. D. program with Dr. Lou Guillette as his advisor. His research focuses on the reproductive biology of neotropical vertebrates, from physiological and behavioral pespectives. He is also interested in problems of rarity and vulnerability of species to extinction. 89

PAGE 97

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Louis J. GuiIIetteTthairman Professor of Zoology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Richard A. Kiltie Associate Professor of Zoology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Assistant Professor of Zoology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. W. F;itzJ^trick Professor bf^orest Resources / and Conservation

PAGE 98

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^^^^ ^^;fe^>^ Scott K. Robinson Associate Professor of Ecology University of Illinois This dissertation was submitted to the Graduate Faculty of the Department of Zoology in the College of Liberal Arts and Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy, May 1993 Dean, Graduate School


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESHGQ5O42_HWQ2WK INGEST_TIME 2015-04-02T21:32:03Z PACKAGE AA00030000_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES