Behavioral Ecology Vol. 14 No. 6: 757-770
© 2003 International Society for Behavioral Ecology
When will rejection of parasite nestlings by hosts of nonevicting avian brood parasites be favored? A misimprinting-equilibrium model
a School of Botany and Zoology b School of Mathematics, Statistics, and Information Technology, University of Natal (Pietermaritzburg), P/Bag X01, Scottsville, 3209 South Africa
Address correspondence to M.J. Lawes. E-mail: lawes{at}nu.ac.za.
Received 4 July 2001; revised 16 September 2002; accepted 2 November 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
It has been suggested that discrimination and rejection of the nestlings of avian brood parasites are most likely to evolve when the parasite nestling is raised alongside the host nestlings, for example, many cowbird-host systems. Under these circumstances, the benefits of discrimination are high because the host parents may save most of their brood. However, there is a general absence of nestling rejection behavior among hosts of nonevicting parasites. In a cost-benefit equilibrium model, based on the premise that host species learn to recognize their offspring through imprinting on first breeding, we show that nestling recognition can be adaptive for hosts of cowbirds, but only under strict conditions. Namely, when host nestling survival alongside the parasite is low, rates of parasitism are high and the average clutch size is large. All of these conditions are seldom simultaneously achieved in real systems. Most importantly, the parasite nestling, on average, does not sufficiently depress host nestling survival to outweigh the costs of nestling recognition and rejection errors. Thus, we argue that nestling acceptance behaviors by hosts of nonevicting brood parasites may be explained as an evolutionary equilibrium in which recognition costs act as a stabilizing selection pressure against rejection when most of the host's offspring survive parasitism.
Key words: brood parasitism, cost-benefit equilibrium, cowbirds, nestling rejection, nonevicting parasites, probability tree model.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
The coevolutionary arms race in which egg discrimination by a host is countered by sophisticated egg mimicry in many avian parasites is well known. Investigations of the rate of spread and evolution of host defenses have used the common cuckoo (Cuculus canorus) as a model (Kelly, 1987
; Lotem, 1993; Takasu et al., 1993). The cuckoo is an obligate brood parasite that on laying removes a host's egg and later evicts all the host's nestlings. An evolutionary paradox arises in which the host makes no meaningful contribution to its individual fitness by its action of raising an unrelated cuckoo chick. Although host species have been known to develop an acute ability to discriminate among their own eggs and the mimetic eggs of a parasite, with few exceptions host species do not appear to discriminate against parasitic nestlings (Payne, 1977; Redondo, 1993; Rothstein, 1990). It has been suggested that the costs of discrimination may prevent untempered rejection behavior in a host species (Redondo, 1993). Indeed, Davies and Brooke (1988: Table 16) show that mimicry is usually absent in parasite nestlings raised alone but is sometimes present in parasites raised together with host nestlings, increasing these costs of discrimination for hosts.
In a model based on the premise that host species learn to recognize their offspring (eggs and nestlings) through imprinting on first breeding (Lotem et al., 1992; Rothstein, 1978), Lotem (1993) showed that nestling recognition is unlikely to be adaptive for hosts of the common cuckoo when only the parasitic nestling remains in the nest. He argued that the cost of misimprinting (if the first brood is parasitized) exceeds the benefit of correct learning, and that recognition costs act as a stabilizing selection pressure against rejection. Therefore, the absence of nestling discrimination is the result of an evolutionary equilibrium maintained largely by the existence of these recognition costs (Lotem, 1993; Redondo, 1993).
However, some brood parasites, such as cowbirds, do not always eject the eggs and seldom, if ever, eject the nestlings of the host from the nest (but see Arcese et al., 1996). When the parasite nestling is reared along with host young, one might expect the benefits of discrimination to be greater (Davies and Brooke, 1988), because the costs of misimprinting or acceptance are lower (Lotem, 1993; Winfree, 1999). Nestling discrimination, if it occurs, would most likely evolve under these circumstances. In this article, we extend Lotem's (1993) misimprinting-equilibrium model to the nonevicting parasite case. We investigate the conditions under which an imprinting mechanism of nestling recognition may be expected to result in rejection of nonevicting brood parasites.
|
THE MODEL |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
Rationale
Several studies suggest that egg recognition in birds is learned by an imprinting process (Lotem et al., 1992
, 1995; Rothstein, 1974, 1975a, 1978), and there are no obvious reasons to explain why hosts should not employ an already existing set of mechanisms to reject both parasitic eggs and nestlings (Redondo, 1993; Rothstein, 1990). Beecher (1982) argues that there are two nestling recognition mechanisms. In both cases, character traits are used by the parent to identify its own nestlings. These traits are either compared with some model learned earlier (imprinted), for example, those individual-specific traits displayed by a particular nestling on first hatching, or some species-specific cue that the nestling mimics (type 1). On the other hand, the model may be learned from some individual other than a nestling (parent or relative), or recognition relies on a genetic mechanism which estimates the proportion of genes shared by parents and nestlings (type 2; cf. Beecher, 1982; Lotem, 1993; Redondo, 1993).
Like Lotem (1993), we assume an imprinting-like process of learned nestling recognition, in which hosts learn to recognize the whole range of variation within a brood on first breeding. We further infer that the parasite nestlings look different from the host's young, although altricial young can look remarkably similar (McLean and Maloney, 1998; Redondo, 1993; Starck, 1993). We argue that this results in subsequent rejection behavior that is a tempered response (i.e., influenced by the cost of making a recognition error) to variability within the clutch and/or brood (for cuckoo-host system, see Davies et al., 1996; Lotem et al., 1995). A cost-benefit equilibrium develops as a consequence of a signal threshold above or below which the host is likely to make errors of recognition (Beecher, 1988; Davies et al., 1996; Reeve, 1989). Nestlings whose traits fall within the range of previously learned variation will be accepted. Thus, a host, previously parasitized on first breeding, will not evict one of its own young in an unparasitized later brood and will always accept the parasite (Lotem, 1993). Hosts that display this serial-learning ability fit the imprinting type 1 model most closely (Beecher, 1982). For naïve "imprinters" (young breeders), this sort of recognition can be maladaptive, if parasitized on the first attempt, because parasites may be recognized as kin.
We assess the reproductive outcomes from acceptance and rejection of eggs and nestlings by hosts of nonevicting brood parasites. The model is based on decision or probability trees (Figure 1), the outcomes of which describe the fitness payoffs for acceptance and rejection. For nestling recognition and rejection to evolve and persist in a population, the payoff for a strategy of rejection must exceed that for acceptance of a parasite.
|
We use cowbird behavior as our exemplar of a nonevicting brood parasite. As a rule, cowbirds do not lay eggs mimetic of their host's egg (Payne, 1997
; Winfree, 1999). Nevertheless, many host species accept all cowbird eggs. Many species also reject all cowbird eggs, but there are very few species that show intermediate levels of rejection of cowbird eggs (Davies, 1999; Rothstein, 1990; Winfree, 1999). This bimodal rejection response by hosts to cowbird parasitism may be attributed to a lag in response that arises among host species that have been recently exposed to intense rates of parasitism (strong selection) for different lengths of time (evolutionary lag hypothesis; Rothstein, 1975b; Rothstein and Robinson, 1998; Winfree, 1999).
Although less favored, the cost-benefit equilibrium hypothesis may also explain the bimodal rejection response if acceptance and rejection costs vary among, but not within, cowbird host species (see Røskaft et al., 1993). Whether cowbird hosts lag in their response to parasitism or are constrained by the costs of rejection (either recognition errors or the risk of damaging an own egg in the process of rejecting), we assume that host individuals are characterized by two types of behavioral response to parasitism. Host individuals either accept the parasite and make no attempt to reject either parasite eggs or nestlings, or attempt rejection of both the parasite egg and nestling. Thus, in a strict sense, the host population comprises either "acceptors" or "rejecters." This is commensurate with the observed bimodal rejection response to parasitism among host species of cowbirds (Davies, 1999; Rothstein, 1990; Winfree, 1999) described above. We do not consider host individuals that adopt a mixed strategy of acceptance and rejection (although in our model, all hosts accept on their first breeding attempt, and we do model the outcome of egg and nestling rejection as separate strategies). Neither do we include the influence of nest-desertion or nest-burial in the model.
To account for the differing costs and selection pressures in the course of parasitism, costs of parasitism that occur when the cowbird removes or damages a host egg and lays its own, and those that occur only after the cowbird young has hatched, are distinguished. In this way, the schedule of costs and their influence on the effectiveness of removing the cowbird's egg or young are established.
Assumptions of the model
The parameters used in the model are listed in Table 1. We assume that the cowbird host lays one clutch per breeding cycle and, on average, that clutches contain x eggs. There is a probability, p (0 
p 1 and define q = 1 - p), of any given nest being parasitized by the end of the egg-laying period. This probability, p, is assumed not to vary from year to year. We do not consider situations in which a nest has been visited by more than one parasite (or more than once by the same parasite, see Robert et al., 1999).
|
The cost of parasitism to the host, at the very least, includes the loss of an egg to the parasite. Cowbirds remove and eat a host's egg or eggs before they lay their own (Payne, 1997
). Although cowbirds are puncture ejectors and may also occasionally damage some of the host's eggs in addition to those it removes from the nest (Røskaft et al., 1990), we assume that the parasite removes only one egg and replaces it with its own egg (Table 2).
|
Hosts are assumed to make their egg or nestling acceptance and rejection decisions according to constant probabilities (Figure 1). On its first attempt at breeding, a rejecter host imprints on the eggs and nestlings in its nest and has yet to develop the ability to differentiate between the parasite and its own eggs and nestlings. Because no rejection occurs in the first brood, the probability tree for acceptors and rejecters is identical for first broods (Figure 1). After the first brood, the host is assumed to be imprinted, meaning that any acceptance or rejection behavior it subsequently displays will be the same on all further broods. This subsequent behavior may involve some attempt at rejection of eggs and nestlings if either does not conform to the hosts imprinted memory of the variation seen in its first nest. Because all clutches are not the same (whether parasitized or not), interclutch and intraclutch variation may confuse the host, and it may reject one of its own eggs or nestlings by mistake when looking for the parasite. This is the "cost of recognition" and is modeled by introducing the following probabilities (Figure 1 and Table 1):
First, "
" (0 1). When a subsequent clutch does not contain a parasite, is the probability that the host will accept all its own eggs and nestlings. If is less than one, there is a slight chance that the host will reject one of its own eggs and nestlings even if there is no parasite present. It is irrelevant to the model as formulated whether or not this rejection decision is made at the egg or the nestling stage.
Second, "
" (0 1). When a subsequent clutch does contain a parasite, is the probability that the host individual will correctly find and reject the parasitic egg. If is less than one, there is a slight chance that the host will reject one of its own eggs by mistake. We assume that a rejecter host will never simply accept all the eggs it sees in the parasitized case.
Third, "r" (0 r 1). When a subsequent clutch does contain a parasite egg that is not detected, r is the probability that the host will correctly identify and reject the parasite nestling. If r is less than one, there is a slight chance that the host will reject one of its own nestlings by mistake. We assume that a rejecter host will never simply accept all the nestlings it sees in the parasitized case, and we also assume that if a rejecter host has already successfully rejected a parasitic egg, it will not attempt any further rejections at the nestling stage.
Finally, we must include the effect of the imprinting mechanism in the model. The above probabilities (, , and r) will depend on whether or not the host was parasitized on its first clutch. By using subscripts f and c to suggest a falsely imprinted and correctly imprinted host, respectively, we assume that
- if parasitism occurred on the host's first clutch, it will subsequently reject in the manner described above with probabilities = f, = f, and r = rf; and
- if parasitism did not occur on the host's first clutch, it will subsequently reject in the manner described above with probabilities = c, = c and r = rc (and we shall expect to find f c, f c, and rf rc).
It is not clear what values the parameters , , and r should take. Estimates from the literature of the parameters , , and r are given in Table 1. Lotem (1993) models an evicting parasite situation and takes f = c = 1. He also assumes x = 4 and f = 0 but is not specific about c. Davies and Brooke (1989b) also model an evicting parasite situation but do not include imprinting so f = c, f = c, and rf = rc, and they take f = c = 0. However, Davies et al. (1996) and Davies (1999), referring to the reed warbler–common cuckoo system, take f = c = 0.7 and f = c = 0.7.
In the absence of data for chick rejection rates in the cowbird system, we estimate the r values for the nonevicting parasite case. Because the worst a rejecter host can possibly do is have an even chance of correctly rejecting the parasite, we initially set f = c = (1/x), and rf = rc = (1/x) in the chick rejecter case and rf = rc = (1/(x - 1)) in the case of a full rejecter (see later; Figure 1). This worst-case scenario for values of r is compared with improved scenarios in which rc approaches one.
The costs of ejection are sometimes included as part of the costs of rejection (see Davies and Brooke, 1988; Davies et al., 1996). These costs arise when the hosts sometimes damage their own eggs while ejecting the parasite's egg (Rohwer et al., 1989; Røskaft et al., 1993; Rothstein, 1976) and are relatively small (approximately 0.3–0.5 of own eggs lost per parasite egg; Davies, 1999; Davies and Brooke, 1988). From a sensitivity analysis of ejection costs on fitness payoffs to rejecters, we conclude that these small costs are not significant, unless host clutch size is unrealistically small (less than two eggs). Cowbird hosts have a mean clutch size of 3.55 ± 0.27 eggs (mean ± 1 SE, n = 21 host species; Table 2). If we include in the model a cost of 0.5 eggs broken or trampled on average during the process of evicting a parasite's egg (Davies, 1999), using the egg rejecter tree, we can show that the advantage of the egg rejection strategy would be reduced by p x x 0.5. This is a reduction of the order of 0.06 x 0.7 x 0.5 = 0.021 own eggs per cycle in the cuckoo case, or 0.36 x 0.28 x 0.5 = 0.05 (see Table 1 for derivation of p =.36) own eggs in the cowbird case (parameter estimates from Table 1 and estimating = 1/[mean clutch size] = 1/3.55 for the cowbird case). Such values are not sufficiently large to significantly change the shape of the advantage curves, and thus, we do not include the effects of such ejection costs in this model.
If the parasitic egg is allowed to hatch in a nest, then some proportion of the host's nestlings are likely to be lost through the actions of the parasite nestling, either owing to smothering or being outcompeted (Payne, 1997: Table 16.1; Redondo, 1993; Sedgewick and Knopf, 1988). This is modeled by introducing a survival probability, s(0 s 1), for each of the host's nestlings in a parasitized nest. For example, for a parasitized clutch of x eggs with an acceptor host, (x - 1) of the host's eggs will hatch; however, because of smothering, only s(x - 1) of the host's nestlings will survive to fledging (Figure 1; see acceptor tree). An evicting parasite may be modeled by setting s to zero. The parameter s can be estimated by taking the apparent survival probability of nestlings in the parasitized case (given by Y/y, where Y is the mean number of nestlings that fledge in parasitized nests, and y is the mean number of eggs that hatch in parasitized nests) and dividing it by the apparent survival probability of nestlings in the unparasitized case (given by X/x, where X is the mean number of nestlings that fledge in unparasitized nests, and x is the mean number of eggs that hatch in unparasitized nests), so that s = (Y/y)/(X/x) = Yx/Xy (Røskaft and Moksnes, 1998).
The survival of host nestlings beside a nonevicting parasite's nestling may be dependent on the size of the brood. If smothering of host nestlings is less severe when there are fewer nestlings in the nest, s should increase as x decreases. On the other hand, the parasite may more easily dominate a nest with a small number of host nestlings so that s will decrease as x decreases. As a compromise between these two arguments, we assume that s, as calculated above, is constant and not dependent on the value of x at all.
In this model, we assume that nestling rejection happens before smothering, thereby ignoring any smothering effects that might occur between hatching and nestling rejection. This is only unreasonable in the case of an evicting parasite nestling that generally evicts before the host has a chance to attempt nestling rejection. Therefore, to model the case of an evicting parasite with a rejecter host, we must not only take s = 0, but r = 0 as well.
Smothering of a sort may occur also at the egg stage; for example, Wiley (1985) found that when the host's eggs were much smaller than the parasite's eggs, the host's eggs did not adequately touch the adult's body, resulting in inadequate incubation. Furthermore, Røskaft et al. (1990) refer to impact with cowbird eggs during incubation as a possible cause of lower host-egg hatching rates beside a cowbird egg. We do not consider such effects in this model.
Although it is known that nests are sometimes depredated, and in very special cases, cowbirds may increase host nestling survival if they reduce the incidence of disease-related mortality by eating ectoparasites such as the larvae of botfly (Smith, 1968), for simplicity, it is assumed that none of the hosts eggs or nestlings die from any other cause such as predation or disease: all nestling deaths are owing to either the host's rejection attempts or the parasite in some way.
Calculating fitness advantage
Using the probability trees (Figure 1) and taking "payoff" to mean the average number of eggs laid by a host that will become chicks and survive to fledging, use Pn for the payoff from the nth clutch and N for the average number of breeding cycles over the lifetime of the host (which equals the total number of clutches because we are assuming one clutch per cycle). Define P to be the payoff over the whole life cycle of the host, that is, P = (P1 + P2 + ... + PN).
Acceptors
Summing probability-times-outcome for all the branches of the first tree in Figure 1, we find that
|
|
1, so the total payoff of an acceptance strategy is
|
|
Full rejecters
For full rejecters that reject both eggs and nestlings
|
|
|
|
Defining T = (pf + qc), U = (pf + qc) and V = (prf + qrc) and W = (prff + qrcc), then
|
|
2 (all subsequent clutches are assumed to be independent), so the total payoff for a full rejection strategy is
|
|
Egg rejecters
|
|
2
|
|
|
|
|
|
Nestling rejecters
Noting the similarities between the third and fourth trees in Figure 1, we can see that the payoff for a nestling rejection strategy is the same for an egg rejection strategy, except that the 's become r's, and
|
|
Fitness advantage
The fitness advantage of a rejection strategy over an acceptance strategy may be defined as,
|
|
Because N (the number of clutches per lifetime) only enters these equations as a multiplicative factor and we are only interested in inequalities involving A (specifically whether or not A is greater than zero), we may restrict ourselves without loss of generality to consideration of the fitness advantage measured as the number of extra host nestlings surviving to fledging per postimprinting clutch, and N becomes inconsequential to the model as formulated.
For small values of x (i.e., x < 3), the payoff expressions Pacceptor, Pfull, Pegg, and Pnestling give negative values. This is a consequence of the fact that the probability trees in Figure 1 simplify for low values of x (a host with a clutch size of x = 1 egg cannot result in [x - 2] fledglings, and its fecundity will realistically be zero). This simplification effectively alters the equations for the various forms of A, and for completeness, these different forms are given in full below for different values of x:
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
Cuckoo-host systems: the evicting parasite
As a first step toward determining whether or not it is adaptive for a host to learn to recognize a parasite egg or nestling, we consider the case of an evicting parasite, exemplified by the common cuckoo, Cuculus canorus. In doing so, we are essentially comparing our model with Lotem's (1993)
misimprinting-equilibrium model.
Egg rejection
Because cuckoos are evicting parasites, we take s = rf = rc = 0 and also assume, as does Lotem (1993), that f = 0 and f = c = 1. From the expression for Aegg, the condition for advantage (Aegg > 0) reduces to
|
|
c(x - 1), we obtain Lotem's (1993) outcome for egg rejection
|
|
f = c = c = 0.7 (Table 1), or estimates of the probabilities that a rejecter host makes an error of recognition. Keeping Lotem's (1993) basic assumptions (essentially s = f = 0 and x = 4), we show that egg rejection behavior is likely to be favored by hosts only if parasitism rates are above approximately 14% (Figure 2). Davies et al. (1996) also concluded that recognition errors are costly enough to make it adaptive for reed warblers (Arcocephalus scirpaceus) to accept rather than reject cuckoo eggs when parasitism is below a critical level (14% for nonmimetic eggs, and 19% for mimetic eggs) or absent altogether.
|
Nestling rejection
In the case of nestling recognition in which the parasite is an evictor, Lotem (1993)
showed for the cuckoo-host system that rejecters do better than acceptors only if the benefit of nestling rejection (b) is greater than the reproductive value of unparasitized nests (x) or b > x. Substituting b = c(x - 1) into the latter, this condition becomes c(x - 1) > x or x < c/(c - 1). Because c is a probability (hence, 0 c 1), the quantity c/(c - 1) is negative, and therefore, learning to recognize the cuckoo nestling is always maladaptive. In other words, a nestling rejection strategy confers no advantage on cuckoo hosts when the cuckoo nestling remains alone in the nest.
Cowbird-host systems: the nonevicting parasite
Can a nestling rejection strategy become established in a typical cowbird host population in which the parasite nestling grows alongside those of the host? We consider three rejection scenarios by a host of a nonevicting parasite. First is a scenario in which the host rejects only the cowbird egg (using the advantage function Aegg). This scenario allows us to examine the payoff advantage that exists for potential rejection circumstances in hosts of cowbirds (where s > 0). In these scenarios, we use s = 0.4, which is a conservative value (actual s = 0.63; Table 3), but permits the useful illustration of payoff advantage for rejection behavior in which the cowbird seriously affects nestling survival (Figure 3a). Second, we examine a scenario in which the host exhibits only nestling rejection (using the advantage function Anestling) (Figure 3b). Last, we model a serial rejecter who rejects both egg and nestlings (this is our full rejecter, and we use the advantage function Afull) (Figure 3c).
|
|
Taking the parameter values
f = 0, c = 0.7, and c = 0.7 and assuming that the host has only an even chance of correctly rejecting the parasite (i.e., f = [1/x]; rf = rc = [1/x] in the nestling rejecter case, and rf = rc = [1/(x - 1)] in the full rejecter case) and by comparing scenarios, it is possible to show that an imprinting mechanism of recognition could result in discrimination of the nonevicting brood parasite in the nest, but only under strict conditions. Namely, for parasites that are harmful to hosts (low s; at least < 0.45; here we apply s = 0.4), when parasitism rates are moderately high (p > 0.32) in the egg or full rejection cases, or very high (p 0.85, when x 5) in the nestling rejection case, and among those hosts with large clutch sizes (x 3) (Figure 3).
The parameter values used to create Figure 3 are deliberately conservative about the likelihood of nestling rejection because they assume that nestling recognition is more difficult (i.e., rc < c) than egg recognition, and the payoffs represent a worst-case scenario. As we increase the rc value that controls the likelihood of nestling rejection in the correctly imprinted case so that rc = c, we find that the fitness advantage of nestling rejection increases to match that of the egg rejection case, but does not exceed it. Increasing rc further still to the (perhaps unlikely) best-case scenario value rc = 1 (so nestling rejection is now easier for the host than egg rejection, and rc >> rc), we find that the threshold parasitism rate (p = 0.35) above which rejection becomes adaptive is still moderately high for an average cowbird host with x = 3.55 eggs (Figure 4a).
|
Although very different nestling rejection fitness advantage curves result for the at-worst (Figure 3b) and at-best scenarios (Figure 4a), curiously, even though nestling recognition is highly constrained in the at-worst scenario, the full rejection fitness advantage (Figure 3c) is only slightly less than in the at-best scenario for all clutch sizes considered (Figure 4b). In addition, the critical threshold level of parasitism for promoting rejection behavior is lowered slightly when there is absolute certainty that the host can recognize and reject the parasite nestling. These results suggest that because the outcomes of the at-worst and at-best scenarios for full rejection are so similar, the ease with which a correctly imprinted host can recognize and reject nestlings relative to eggs is not an important factor in determining whether or not nestling rejection behavior will evolve. The model does show that for the most part, payoffs for egg rejection are greater than for nestling rejection, and thus, egg rejection should be favored by natural selection over nestling rejection even in a nonevicting system in which parasite nestlings are raised alongside those of the host's. Nevertheless, it is likely that the costs of rejection and the cost of misimprinting on the parasite nestling are among the important factors causing the lack of observed nestling rejection behavior in nonevicting avian brood parasitism systems.
Nestling rejection behavior is sensitive to changes in s and p (Figure 5). By using the mean parameter values for x (x = 3.6; Table 2) and s (s = 0.63; Table 3), we can see that if the parasite causes more than 50% likelihood of mortality (1 - s) of a host's nestlings, the cowbird host would have higher fitness, on average, by rejecting the parasite's nestling (Figure 5a). The current lack of nestling recognition may be attributed in part to the fact that, on average, the cowbird nestling simply does not cause sufficient harm to the host to outweigh the costs of recognition and rejection errors (Figure 5b). Thus, nestling rejection behavior is only likely to arise in hosts of nonevicting parasites in which nestling mortality and rates of parasitism are high (Figure 5c). Because neither of these conditions (note the mean p = 0.36; Table 2) is met in natural populations, cowbird hosts tend to be nestling acceptors.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
By using a cost-benefit equilibrium model in which we assume that both rejection ("recognition costs") and acceptance ("misimprinting costs") entail costs, we show that nestling recognition can be adaptive for hosts of cowbirds when some of the hosts nestlings survive alongside the parasitic nestling to fledging. In cuckoo-host systems in which only the parasitic nestling remains in the nest, nestling recognition is maladaptive because the cost of misimprinting (if the first brood is parasitized) exceeds the benefit of correct learning, and recognition costs act as a stabilizing selection pressure against rejection (Lotem, 1993
). Thus, the absence of nestling discrimination in cuckoo-host systems may be the result of an evolutionary equilibrium maintained largely by the existence of these recognition costs (Lotem, 1993; Redondo, 1993). It is thought that in hosts of nonevicting parasites (i.e., cowbird-host systems) recognition costs are likely to be much reduced, because hosts have their own young for comparison. In addition, the benefits of nestling discrimination are high (and the cost of misprinting low) because host parents may save most of their present brood (Davies and Brooke, 1988; Lotem, 1993; Redondo, 1993; Winfree, 1999). Why then is there a general absence of nestling rejection behavior among hosts of nonevicting parasites? In this analysis, we argue that cowbird hosts are most likely to accept parasite nestlings because the costs of recognition, although potentially less than in cuckoo-host systems, nevertheless outweigh their benefits when most of the host's offspring survive parasitism. An evolutionary equilibrium is reached (see Lotem, 1993; Lotem and Nakamura, 1998) as long as the parasite nestling does not sufficiently depress host nestling survival, which would increase the benefits of nestling recognition behavior. Indeed, the mean probability of host nestling survival alongside the cowbird nestling is relatively high (s = 0.63). Furthermore, the mean rate of parasitism by nonevicting parasites (p = 0.36) falls below levels required to justify nestling rejection behavior. We conclude that the parasite does insufficient harm to the host's offspring for nestling rejection behavior to be adaptive.
This model is based on the premise that host species learn to recognize their offspring (eggs and nestlings) through imprinting on first breeding (Lotem et al., 1992; Rothstein, 1978). Recognition of own eggs by imprinting on the variability in the first clutch is a widely distributed behavior among hosts of brood parasites (Davies, 2000; Lotem, 1993; Rothstein, 1974, 1975a,c, 1990). Thus, hosts of nonevicting parasites could use an already existing set of mechanisms to reject both parasite eggs and nestlings, learning to recognize their offspring and paying a relatively low cost of misimprinting (Lichtenstein, 2001; Lotem, 1993; Rothstein, 1990). In addition, adult birds are raised alongside siblings and potentially have the opportunity as nestlings to learn about the appearance of nestlings of their own species (McLean and Maloney, 1998).
There is, of course, the possibility that hosts do not recognize nestlings by imprinting on offspring signatures (McLean and Maloney, 1998; Redondo, 1993). If this were the case, the explanatory power of a cost-benefit equilibrium model such as this would be weak. However, the near resemblance between the nestlings of some host-specific, nonevicting parasites and their closely related hosts (e.g., the screaming cowbird, Molothrus rufoaxillaris, and the baywinged cowbird, Molothrus badius, Fraga, 1998; parasitic Vidua nestlings and their estrildid finch hosts, Payne, 1997) suggests that some hosts may discriminate among nestlings. Further evidence from a recent study suggests that parental discrimination of the cowbird nestling is not limited to related groups (Lichtenstein, 2001).
The equilibrium model may provide insight to why egg rejection, but not nestling rejection, has evolved among some cowbird hosts, and also why most cowbird hosts are acceptor species (Rothstein, 1992). Theoretically, the earlier in the reproductive cycle that the host correctly identifies the parasite and rejects it, the greater the benefit of rejection to the host. Thus, selection pressure on, and the benefit of, egg rejection should be greater than for nestling rejection (Røskaft et al., 1993; Rothstein, 1990). Our model of the cowbird-host system conforms to this prediction, and the advantage to a host of rejecting eggs (Figure 3a) is greater than the advantage to be gained by rejecting only nestlings (Figure 3b). This result is, however, a direct consequence of incorporating greater costs of recognizing nestlings than eggs (cf. parameter values for rc and c) in our model. These greater costs are likely for two reasons: (1) the cowbird egg is nonmimetic, and in fact, costs of egg-recognition may be less than modeled in the present study (Rothstein, 1975b); and (2) recognition of nestlings is probably more difficult than the egg because the nestling presents a dynamic set of cues that change as it grows, increasing the risk of misimprinting (McLean and Maloney, 1998; Redondo, 1993). Furthermore, because newly hatched young of altricial birds do look similar (McLean and Maloney, 1998; Starck, 1993), by the time specific differences between parasite and host nestling become obvious, the parasite nestling may be too large to evict easily or without the host damaging its own nestlings.
The latter arguments notwithstanding, the relative ease with which a correctly imprinted host recognizes the parasite egg and nestling does not appear to be a critical factor in the model determining whether or not chick rejection behavior evolves. Even when we made the likelihood of recognizing nestlings equal to, or much greater than, eggs in the model, the restrictive conditions necessary for nestling rejection behavior to evolve in the average host remained largely unchanged (i.e., there was still no fitness advantage to average hosts of nestling rejection). The latter was true for all but the most unlikely scenario, in which probability of nestling recognition by a correctly imprinted host approached 100%. We conclude that the costs associated with rejecting the parasite nestling and the costs of misimprinting are more important causes of the observed lack of nestling rejection than the relative ease of nestling and egg recognition.
The selective forces favoring nestling rejection by hosts of cowbird nestlings are probably not as great as those favoring rejection of parasite eggs (see Røskaft et al., 1993). Nevertheless, McLean and Maloney (1998) suggest that species that have not evolved egg rejection could still have evolved nestling rejection. Like us, they argue that nestling rejection may arise in circumstances in which the survival of host nestlings is low as a consequence of the actions of the parasite nestling, and nestling rejection is less costly than egg rejection because egg rejection results in damage to other eggs in the nest (Rohwer et al., 1989). However, our cost-benefit equilibrium model and the available breeding data suggest that nonevicting parasites cause insufficient harm to their hosts to outweigh the risks of rejection behavior and for nestling rejection behavior to be adaptive.
In a test of the "insufficient harm" hypothesis, a comparison of the nestling survival of relatively smaller and larger hosts of the cowbird may be instructive. Our model does not consider the relative size of the host. Host nestling survival may be dependent on the size difference between host and parasite. For example, relatively small hosts of cowbirds can lose all of their nestlings in competition with the cowbird nestling (Davies, 2000: 153; Lichtenstein and Sealy, 1998). For small hosts, having a cowbird nestling can be as bad as having a cuckoo in the nest (see Dearborn, 1996). Faced by these high costs of accepting the parasite nestling, why do small hosts not discriminate against a larger parasitic nestling? One possibility is that large nestlings are a supernormal stimulus (Redondo, 1993). Another suggestion is that host nestlings are lost so soon after hatching that nestling recognition does not have time to operate. If all other costs are equal, discrimination against the cowbird nestling(s) is more likely to evolve when the cost of recognition error is low, such as when the parasitic nestling is notably smaller than the host's, and the foster parents can save most of their nestlings after rejecting the parasite (Davies and Brooke, 1988). It should be noted, however, that Røskaft and Moksnes (1998) found weak support for the prediction that smaller-sized hosts suffered higher overall costs.
On the other hand, the cowbird may be ahead of the host in the evolutionary arms race. Cowbird nestlings may not be evicted from the host nest because they beg more vigorously than the host young do (Gochfeld, 1979). The adaptive value of a positive response toward a cowbird nestling begging for food may be so great that host adaptations based on rejection behaviors (eviction, neglect) are largely precluded by the supernormal begging stimulus (Hamilton and Orians, 1965; Redondo, 1993; Røskaft and Moksnes, 1998). However, this begging behavior is not necessarily a consequence of an evolutionary arms race. The parasite may beg more vigorously than does the host young because it has no genetic interest in the survival of its nest mates or in the parent's future reproduction; that is, because it has no kinship with the host, there are no genetic costs to selfish begging behavior (Godfray, 1995; Kilner and Davies, 1999). Thus, although the host will gain the greatest fitness benefit by being a serial rejecter and rejecting both eggs and nestlings (Figure 3c), it is likely that, for the most part, the dynamic nature of the mechanism of nestling recognition (Redondo, 1993) and possibly the cognitive limitations of the hosts (McLean and Maloney, 1998), impose costs on the recognition of nestlings (evolutionary equilibrium hypothesis) and has restricted the evolution of rejection behavior to the egg phase.
Conclusion
Because parasitism reduces host fitness, rejecting the parasitic egg and nestlings is expected to be advantageous. However, both egg and nestling rejection incur several costs that reduce the selective advantage of rejection, resulting in a slower evolutionary rate (Davies and Brooke, 1989a). In this analysis and by using normative parameter values, the sum of all the costs of egg and nestling rejection exceeds the benefits, showing that for the most part the acceptance of the parasitic nestling is adaptive. This model of acceptance behavior can be explained as an evolutionary equilibrium based on the costs of recognizing and rejecting the parasitic nestling, as well as the cost of misimprinting on the parasitic nestling (Lotem and Nakamura, 1998). The dynamic nature of the mechanism of nestling recognition imposes costs on the recognition of nestlings so that even when the host's nestlings are raised alongside the parasite, unless the parasite causes the loss of most of the host's offspring (the "insufficient harm hypothesis"), the costs of recognition and rejection outweigh the benefits resulting in nestling acceptance. Finally, it is not necessary to invoke evolutionary lag in explaining this outcome.
|
ACKNOWLEDGEMENTS |
|---|
We wish to thank Malte Andersson, Nick Davies, David Lathi, and Justin Schuetz for their valuable advice and comments on earlier and incomplete drafts of this article. Anna Lindholm was instrumental in sowing the seeds of the idea pursued in this article; we thank her for her support. Our thanks go to Dale Forbes and Hallam Payne for helping to collate Tables 2 and 3. We are also grateful to Rudi van Aarde and Theo Wassenaar (Zoology, University of Pretoria) for giving us access to computing facilities and deferring their own deadlines so that we could complete the revisions to this article. M.J.L. gratefully acknowledges the financial support from the University of Natal Research Fund (URF).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MODELRESULTSDISCUSSIONREFERENCES |
|---|
Arcese P, Smith JNM, Harch MI, 1996. Nest predation by cowbirds and its consequences for passerine demography. Proc Natl Acad Sci, USA 93:4608-4611.
Beecher MD, 1982. Signature systems and kin recognition. Am Zool 22:477-490.
Beecher MD, 1988. Kin recognition in birds. Behav Genet 18:465-482.[CrossRef][Web of Science][Medline]
Braden GT, McKernan RL, Powell SM, 1997. Effects of nest parasitism by the brown-headed cowbird on nesting success of the California gnatcatcher. Condor 99:858-865.[CrossRef][Web of Science]
Briskie JV, Sealy SG, Hobson KA, 1990. Differential parasitism of least flycatchers and yellow warblers by the brown-headed cowbird. Behav Ecol Sociobiol 27:403-410.[CrossRef][Web of Science]
Burgham MCJ, Picman J, 1989. Effect of brown-headed cowbirds on the evolution of yellow warbler anti-parasite strategies. Anim Behav 38:298-308.
Burhans DE, 1997. Habitat and microhabitat features associated with cowbird parasitism in two forest edge cowbird hosts. Condor 99:866-872.[CrossRef][Web of Science]
Clark KL, Robertson RJ, 1979. Spatial and temporal multi-species nesting aggregations in birds as anti-parasite and anti-predator defenses. Behav Ecol Sociobiol 5:359-371.[CrossRef][Web of Science]
Clark KL, Robertson RJ, 1981. Cowbird parasitism and evolution of anti-parasite strategies in the yellow warbler. Wilson Bull 92:244-258.
Cruz A, Wiley JW, 1989. The decline of an adaptation in the absence of a presumed selection pressure. Evolution 43:55-62.
Davies NB, 1999. Cuckoos and cowbirds versus hosts: co-evolutionary lag and equilibrium. Ostrich 70:71-79.[Web of Science]
Davies NB, 2000. Cuckoos, cowbirds and other cheats. London: Poyser.
Davies NB, Brooke MdL, 1988. Cuckoos versus reed warblers: adaptations and counteradaptations. Anim Behav 36:262-284.[CrossRef][Web of Science]
Davies NB, Brooke MdL, 1989a. An experimental study of co-evolution between the cuckoo, Cuculus canorus, and its hosts, I: host egg discrimination. J Anim Ecol 58:207-224.[CrossRef]
Davies NB, Brooke MdL, 1989b. An experimental study of co-evolution between the cuckoo, Cuculus canorus, and its hosts, II: host egg markings, chick discrimination and general discussion. J Anim Ecol 58:225-236.[CrossRef]
Davies NB, de L., Brooke M, Kacelnik A, 1996. Recognition errors and probability of parasitism determine whether reed warblers should accept or reject mimetic cuckoo eggs. Proc R Soc Lond, B 263:925-931.
Dearborn DC, 1996. Video documentation of a brown-headed cowbird nestling ejecting an indigo bunting nestling from the nest. Condor 98:645-649.[CrossRef][Web of Science]
Donovan TM, Thompson FR, Faaborg J, Probst JR, 1995. Reproductive success of migratory birds in habitat sources and sinks. Conserv Biol 9:1380-1395.[CrossRef]
Eckerle KP, Breitwisch R, 1997. Reproductive success of the northern cardinal, a large host of brown-headed cowbirds. Condor 99:169-178.
Elliott PF, 1978. Cowbird parasitism in the Kansas tallgrass prairie. Auk 95:161-167.[Web of Science]
Fauth PT, 2000. Reproductive success of wood thrushes in forest fragments in northern Indiana. Auk 117:194-204.[CrossRef][Web of Science]
Flaspohler DJ, Temple SA, Rosenfield RN, 2001. Species-specific edge effects on nest success and breeding bird density in a forested landscape. Ecol Appl 11:32-46.[CrossRef]
Fraga RM, 1978. The rufous-collared sparrow as a host of the shiny cowbird. Wilson Bull 90:271-284.
Fraga RM, 1998. Interactions of the parasitic screaming cowbirds (Molothrus rufoaxillaris and M. bonaerensis) with a shared host, the bay-winged cowbird (Molothrus badiu). In: Parasitic birds and their hosts: studies in coevolution (Rothstein SI, Robinson SK, eds). New York: Oxford University Press; 173–193.
Gochfeld M, 1979. Brood parasite and host coevolution: interactions between shiny cowbirds and two species of meadowlarks. Am Nat 113:855-870.[CrossRef][Web of Science]
Godfray HCJ, 1995. Signaling of need between parents and young: parent-offspring conflict and sibling rivalry. Am Nat 146:1-24.[CrossRef][Web of Science]
Goertz JW, 1977. Additional records of brown-headed cowbird nest parasitism in Louisiana. Auk 94:386-389.[Web of Science]
Goguen CB, Mathews NE, 2000. Local gradients of cowbird abundance and parasitism relative to livestock grazing in a western landscape. Conserv Biol 14:1862-1869.[CrossRef]
Hamilton WJ, Orians GH, 1965. The evolution of brood parasitism in altricial birds. Condor 67:361-382.[CrossRef]
Hanski IK, Fenske TJ, Niemi GJ, 1996. Lack of edge effect in nesting success of breeding birds in managed forest landscapes. Auk 113:578-585.[Web of Science]
Hill DP, Sealy SG, 1994. Desertion of nests parasitized by cowbirds: have clay-colored sparrows evolved an antiparasite defense. Anim Behav 48:1063-1070.[CrossRef]
Kattan GH, 1998. Impact of brood parasitism: why do house wrens accept shiny cowbird eggs? In: Parasitic birds and their hosts: studies in coevolution (Rothstein SI, Robinson SK, eds). New York: Oxford University Press; 212–220.
Kelly C, 1987. A model to explore the rate of spread of mimicry and rejection in hypothetical populations of cuckoos and their hosts. J Theor Biol 125:283-299.[CrossRef]
Kilner RM, Davies NB, 1999. How selfish is a cuckoo chick? Anim Behav 58:797-808.[CrossRef][Web of Science][Medline]
King JR, 1973. Reproductive relationships of the rufous-collared sparrow and the shiny cowbird. Auk 90:19-34.[Web of Science]
Lichtenstein G, 2001. Low success of shiny cowbird chicks parasitizing rufous-bellied thrushes: chick-chick competition or parental discrimination? Anim Behav 61:401-413.[CrossRef]
Lichtenstein G, Sealy SG, 1998. Nestling competition, rather than supernormal stimulus, explains the success of parasitic brown-headed cowbird chicks in yellow warbler nests. Proc R Soc Lond, B 265:249-254.
Linz GM, Bolin SB, 1982. Incidence of brown-headed cowbird parasitism on red-winged blackbirds. Wilson Bull 94:93-95.
Lotem A, 1993. Learning to recognize nestlings is maladaptive for cuckoo Cuculus canorus hosts. Nature 362:743-745.[CrossRef]
Lotem A, Nakamura H, 1998. Evolutionary equilibria in avian brood parasitism: an alternative to the "arms race-evolutionary lag" concept. In: Parasitic birds and their hosts: studies in coevolution (Rothstein SI, Robinson SK, eds). New York: Oxford University Press; 223–235.
Lotem A, Nakamura H, Zahavi A, 1992. Rejection of cuckoo eggs in relation to host age: a possible evolutionary equilibrium. Behav Ecol 3:128-132.
Lotem A, Nakamura H, Zahavi A, 1995. Constraints on egg discrimination and cuckoo-host co-evolution. Anim Behav 49:1185-1209.[CrossRef]
Marvil RE, Cruz A, 1989. Impact of brown-headed cowbird parasitism on the reproductive success of the solitary vireo. Auk 106:476-480.[Web of Science]
Massoni V, Reboreda JC, 1998. Costs of brood parasitism and the lack of defenses on the yellow-winged blackbird shiny cowbird system. Behav Ecol Sociobiol 42:273-280.[CrossRef][Web of Science]
Mayfield H, 1965. Chance distribution of cowbird eggs. Condor 67:257-263.[CrossRef]
McLaren CM, Sealy SG, 2000. Are nest predation and brood parasitism correlated in yellow warblers: a test of the cowbird predation Hypothesis. Auk 117:1056-1060.[CrossRef][Web of Science]
McLean IG, Maloney RF, 1998. Brood parasitism. recognition, and response: the options. In: Parasitic birds and their hosts: studies in coevolution (Rothstein SI, Robinson SK, eds). New York: Oxford University Press; 255–269.
McMaster DG, Sealy SG, Gill SA, Neudorf DL, 1999. Timing of egg-laying in yellow warblers. Auk 116:236-240.[Web of Science]
Mermoz ME, Reboreda JC, 1996. New host for a specialized brood parasite, the screaming cowbird. Condor 98:630-632.[CrossRef][Web of Science]
Morse SF, Robinson SK, 1999. Nesting success of a neotropical migrant in a multiple-use, forested landscape. Conserv Biol 13:327-337.[CrossRef]
Murphy MT, 1986. Brood parasitism of eastern kingbirds by brown-headed cowbirds. Auk 103:626-628.[Web of Science]
Neudorf DL, Sealy SG, 1992. Reactions of four passerine species to threats of predation and cowbird parasitism: enemy recognition or generalised responses? Behaviour 123:84-105.[CrossRef][Web of Science]
Neudorf DL, Sealy SG, 1994. Sunrise nest attentiveness in cowbird hosts. Condor 96:162-169.[CrossRef][Web of Science]
Nice MM, 1937. Studies in the life history of the song sparrow, part 1: transactions of the Linnean Society, NY. Biol J Linn Soc 4:1-247.
Nolan V, 1978. The ecology and behavior of the prairie warbler Dendroica discolor. Ornith Monogr 26:1-595.
Orians GH, Røskaft E, Beletsky LD, 1989. Do brown-headed cowbirds lay their eggs at random in the nests of red-winged blackbirds? Wilson Bull 101:599-605.
Payne RB, 1977. The ecology of brood parasitism in birds. Annu Rev Ecol Syst 8:1-28.[CrossRef][Web of Science]
Payne RB, 1997. Avian brood parasitism. In: Host-parasite evolution: general principles and avian models (Clayton DH, Moore J, eds). Oxford: Oxford University Press; 338–369.
Peer BD, Bollinger EK, 1997. Explanations for the infrequent cowbird parasitism on common crackles. Condor 99:151-161.[CrossRef][Web of Science]
Peer BD, Robinson SK, Herkert JR, 2000. Egg rejection by cowbird hosts in grasslands. Auk 117:892-901.[CrossRef][Web of Science]
Petit LJ, 1991. Adaptive tolerance of cowbird parasitism by prothonotary warblers: a consequence of nest-site limitation? Anim Behav 41:425-432.[CrossRef]
Piper WH, 1994. Courtship, copulation, nesting behaviour and brood parasitism in the Venezuelan stripe-backed wren. Condor 96:654-671.[CrossRef][Web of Science]
Porneluzi PA, Faaborg J, 1999. Season-long fecundity, survival, and viability of ovenbirds in fagmented and unfragmented landscapes. Conserv Biol 13:1151-1161.[CrossRef]
Post W, Nakamura TK, Cruz A, 1990. Patterns of shiny cowbird parasitism in St. Lucia and southwestern Puerto Rico. Condor 92:461-469.[CrossRef][Web of Science]
Post W, Wiley JW, 1977. Reproductive interactions of the shiny cowbird and the yellow-shouldered blackbird. Condor 79:176-184.[CrossRef][Web of Science]
Redondo T, 1993. Exploitation of host mechanisms for parental care by avian brood parasites. Etología 3:235-297.
Reeve HK, 1989. The evolution of conspecific acceptance thresholds. Am Nat 133:407-435.[CrossRef][Web of Science]
Robert M, Sorci G, Moller AP, Hochberg ME, Pomiankowski A, Pagel M, 1999. Retaliatory cuckoos and the evolution of host resistance to brood parasites. Anim Behav 58:817-824.[CrossRef][Web of Science][Medline]
Rogers CM, Taitt MJ, Smith JNM, Jongejan G, 1997. Nest predation and cowbird parasitism create a demographic sink in wetland-breeding song sparrows. Condor 99:622-633.[CrossRef][Web of Science]
Rohwer S, Spaw CD, Røskaft E, 1989. Costs to northern orioles of puncture-ejecting parasitic cowbird eggs from their nests. Auk 106:734-738.[Web of Science]
Røskaft E, Moksnes A, 1998. Coevolution between brood parasites and their hosts: an optimality theory approach. In: Parasitic birds and their hosts: studies in coevolution. (Rothstein SI, Robinson SK, eds). New York: Oxford University Press; 236–254.
Røskaft E, Orians GH, Beletsky LD, 1990. Why do red-winged blackbirds accept eggs of brown-headed cowbirds? Evol Ecol 4:35-42.
Røskaft E, Rohwer S, Spaw CD, 1993. Costs of puncture ejection compared with costs of rearing cowbird chicks for northern orioles. Ornis Scand 24:28-32.[CrossRef]
Rothstein SI, 1974. Mechanisms of avian egg recognition: possible learned and innate factors. Auk 91:796-807.[Web of Science]
Rothstein SI, 1975a. Evolutionary rates and host defenses against avian brood parasitism. Am Nat 109:161-176.[CrossRef][Web of Science]
Rothstein SI, 1975b. An experimental and teleonomic investigation of avian brood parasitism. Condor 77:250-271.[CrossRef][Web of Science]
Rothstein SI, 1975c. Mechanisms of avian egg recognition: do birds know their own eggs? Anim Behav 23:268-278.[CrossRef]
Rothstein SI, 1976. Experiments on defenses cedar waxwings use against cowbird parasitism. Auk 93:675-691.[Web of Science]
Rothstein SI, 1978. Mechanisms of avian egg-recognition: additional evidence for learned components. Anim Behav 26:671-677.[CrossRef]
Rothstein SI, 1990. A model system for coevolution: avian brood parasitism. Annu Rev Ecol Syst 21:481-508.[CrossRef][Web of Science]
Rothstein SI, 1992. Brood parasitism, the importance of experiments and host defences of avifaunas on different continents. Proc 7th Pan-African Ornithol Congr 521–535.
Rothstein SI, Robinson SK, 1998. Parasitic birds and their hosts: studies in coevolution. New York: Oxford University Press.
Scott DM, Lemon RE, 1996. Differential reproductive success of brown-headed cowbirds with northern cardinals and three other hosts. Condor 98:259-271.[CrossRef][Web of Science]
Sealy SG, 1992. Removal of yellow warbler eggs in association with cowbird parasitism. Condor 94:40-54.[CrossRef][Web of Science]
Sealy SG, Neudorf DL, 1995. Male northern orioles eject cowbird eggs: implications for the evolution of rejection behavior. Condor 97:369-375.[CrossRef][Web of Science]
Sedgewick JA, Knopf FL, 1988. A high incidence of brown-headed cowbird parasitism of willow flycatchers. Condor 90:253-256.[CrossRef][Web of Science]
Sick H, Ottow J, 1958. Vom brasilianischen Kuhvogel, Molothrus bonariensis, und seinen Wirten, besonders dem Ammerfinken, Zonotrichia capensis. Bonner Zoologische Beiträge 9:40-62.
Smith JNM, 1981. Cowbird parasitism, host fitness, and age of the host female in an island song sparrow population. Condor 83:152-161.[CrossRef][Web of Science]
Smith JNM, Arcese P, 1994. Brown-headed cowbirds and an island population of song sparrows: a 16-year study. Condor 96:916-934.[CrossRef][Web of Science]
Smith NG, 1968. The advantage of being parasitized. Nature 219:690-694.[CrossRef][Medline]
Starck JM, 1993. Evolution of avian ontogenies. Curr Ornithol 10:275-366.
Strausberger BM, 1998. Temporal patterns of host availability, brown-headed cowbird brood parasitism, and parasite egg mass. Oecologia 116:267-274.[CrossRef][Web of Science]
Takasu F, Kawasaki K, Nakamura H, Cohen JE, Shigesada N, 1993. Modeling the population dynamics of a cuckoo-host association and the evolution of host defenses. Am Nat 142:819-839.[CrossRef][Web of Science]
Tewksbury JJ, Hejl SJ, Martin TE, 1998. Breeding productivity does not decline with increasing fragmentation in a western landscape. Ecology 79:2890-2903.[CrossRef][Web of Science]
Trine CL, 1998. Wood thrush population sinks and implications for the scale of regional conservation strategies. Conserv Biol 12:576-585.[CrossRef]
Ward D, Smith JNM, 2000. Brown-headed cowbird parasitism results in a sink population in warbling vireos. Auk 117:337-344.[CrossRef][Web of Science]
Weatherhead PJ, 1989. Sex ratios, host-specific reproductive success, and impact of brown-headed cowbirds. Auk 106:358-366.[Web of Science]
Wiley JW, 1985. Shiny cowbird parasitism in two avian communities in Puerto Rico. Condor 87:165-176.[CrossRef][Web of Science]
Winfree R, 1999. Cuckoos, cowbirds and the persistence of brood parasitism. Trends Ecol Evol 14:338-342.[CrossRef][Medline]
Wolf L, 1987. Host-parasite interactions of brown-headed cowbirds and dark-eyed juncos in Virginia. Wilson Bull 99:338-350.
Woodworth BL, 1997. Brood parasitism, nest predation, and season-long reproductive success of a tropical island endemic. Condor 99:605-621.[CrossRef][Web of Science]
Zimmerman JL, 1983. Cowbird parasitism of dickcissels in different habitats and at different nest densities. Wilson Bull 95:7-22.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. E. Langmore, A. Cockburn, A. F. Russell, and R. M. Kilner Flexible cuckoo chick-rejection rules in the superb fairy-wren Behav. Ecol., September 1, 2009; 20(5): 978 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Hauber and C. Moskat Shared parental care is costly for nestlings of common cuckoos and their great reed warbler hosts Behav. Ecol., January 1, 2008; 19(1): 79 - 86. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Hoover and S. K. Robinson Retaliatory mafia behavior by a parasitic cowbird favors host acceptance of parasitic eggs PNAS, March 13, 2007; 104(11): 4479 - 4483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||