Spring and Autumn Territoriality in Song Sparrows: Same Behavior, Different Mechanisms? 1

Transcription

1 INTEG. AND COMP. BIOL., 42:11 20 (2002) Spring and Autumn Territoriality in Song Sparrows: Same Behavior, Different Mechanisms? 1 JOHN C. WINGFIELD 2, * AND KIRAN K. SOMA*, *Department of Zoology, Box , University of Washington, Seattle, Washington Department of Physiological Science, University of California, Los Angeles, California SYNOPSIS. Vertebrates show a diverse array of social behaviors associated with territoriality. Field and laboratory experiments indicate that underlying themes including mechanisms may exist. For example in birds, extensive evidence over many decades has implicated a role for testosterone in the activation of territorial aggression in reproductive contexts. Territoriality at other times of the year appeared to be independent of gonadal hormone control. One obvious question is why this diversity of control mechanisms for an apparently similar behavior? Control of testosterone secretion during the breeding season must balance the need to compete with other males (that tends to increase testosterone secretion), and the need to provide parental care (that requires lower testosterone concentrations). Regulation of aggressive behaviors by testosterone in the non-breeding season may incur substantial costs. A series of experiments on the male song sparrow, Melospiza melodia morphna, of western Washington State have revealed possible mechanisms to avoid these costs. Song sparrows are sedentary and defend territories in both breeding and non-breeding seasons. Dominance interactions, territorial aggression and song during the non-breeding season are essentially identical to those during the breeding season. Although in the non-breeding season plasma testosterone and estradiol levels are very low, treatment with an aromatase inhibitor decreases aggression and simultaneous implantation of estradiol restores territorial behavior. These data suggest that the mechanism by which testosterone regulates territorial behavior at the neural level remains intact throughout the year. How the hormonal message to activate such behavior gets to the brain in different season does, however, appear to be different. INTRODUCTION The ecology and behavior of vertebrates changes dramatically throughout the year, particularly in habitats with pronounced seasonality. Because the vast majority of vertebrates live at least one year, they must adjust their morphology, physiology and behavior to maximize fitness over the life cycle as a whole. To provide a framework for investigating how the endocrine system may regulate these changes, we have proposed that the life cycle of vertebrates is made up of distinct life history stages (LHSs) each with a unique set of sub-stages (Fig. 1, Jacobs and Wingfield, 2000). The temporal sequence of LHSs is timed by environmental cues such as photoperiod, temperature and rainfall on a schedule that is usually repeatable from year to year (Wingfield et al., 2000). In the song sparrow, Melospiza melodia morphna, of western Washington State, there are three LHSs comprising a breeding stage, pre-basic molt and a non-breeding (wintering) stage (Fig. 1). This population is particularly interesting because it is sedentary and free-living individuals can be followed throughout the year. This allows us to investigate how behavioral traits such as territorial aggression are expressed in different LHSs and then go on to determine the neuroendocrine and endocrine mechanisms that underlie these changes. Although the endocrine control of reproductive behavior, including territoriality, has received consider- 1 From the Symposium Taking Physiology to the Field: Advances in Investigating Physiological Function in Free-Living Vertebrates presented at the Annual Meeting of the Society for Integrative and Comparative Biology, 3 7 January 2001, at Chicago, Illinois. 2 jwingfie@u.washington.edu able attention in the breeding LHS, the regulation of aggression in non-reproductive contexts is much less well known (Becker et al., 1992; Nelson, 1999). Many vertebrates living at mid to high latitudes have highly specialized morphological, physiological and behavioral changes in the non-breeding LHS. For example, some species molt into a white plumage or pelage in winter, and their physiology is adjusted to allow accumulation of fat for thermoregulation during long cold nights. Others may take advantage of shelters, even snow caves (e.g., Andreev, 1999). Although there is a growing literature on wintering strategies in birds, most of the hormone mechanisms remain entirely unknown (Silverin, 1992). In this chapter we will focus on the regulation of territorial aggression in the song sparrow throughout different LHSs in the life cycle (Fig. 1), including winter, and how the endocrine control mechanisms may vary. SAME BEHAVIOR IN DIFFERENT LIFE HISTORY STAGES The song sparrow of western Washington State and southwestern British Columbia is territorial year round and in some populations breeding pairs may stay on their territory for more than one year (e.g., Arcese, 1989; Nordby et al., 1999). To test whether territorial aggression is similar at all times of year, the responses of free-living male song sparrows were tested by simulated territorial intrusions (STIs). This involved placing a live male decoy in a cage on a territory and playing tape-recorded songs through a speaker placed alongside. The resident male typically responds by singing back at the intruder, then approaches the decoy, tends to display with threats, wing flutters, flights around the decoy and also attacks. The STI pro- 11

2 12 J. C. WINGFIELD AND K. K. SOMA FIG. 1. Sequence of life history stages (LHSs) for the rufous song sparrow, Melospiza melodia morphna, in western Washington State. Each box represents a specific LHS with characteristic sub-stages denoted inside each box. Only those sub-stages relating to territorial aggression and social interactions are shown. The temporal sequence of LHSs progresses in one direction on a schedule determined by seasons (Jacobs and Wingfield, 2000). Note that territorial aggression is expressed in the breeding and non-breeding LHSs but less so in the prebasic molt LHS. tocol has the advantage of bringing the bird out into the open where its aggressive responses can be quantified. These include numbers of songs, closest approach to the decoy, time spent within five meters of the decoy and numbers of flights (Wingfield, 1985; Wingfield and Hahn, 1994). By placing a mist net close to the decoy, the responding male can be captured for blood sampling (hormone measurements), or then treated with hormones or their inhibitors (Soma and Wingfield, 1999). Free-living male song sparrows showed a marked increase in numbers of songs and flights in response to STI in spring as the breeding LHS began (Fig. 2). They tend to come closer to the decoy and spent more time with five meters than males during the winter although some of the latter show high levels of aggression also at some localities (Wingfield and Hahn, 1994; Soma et al., 1999a, 2000a). These high levels of territorial response to STI were maintained throughout the breeding LHS but declined markedly when they were in molt (Fig. 2). Males (identified by their unique color band combinations) were still on their territories, but did not respond to STI. It has been suggested that this decline in territorial aggression may be because large flight and tail feathers were growing at this time and would be easily damaged in a fight (Wingfield and Hahn, 1994). During feather growth the sheath was soft, highly vascularized and could be severely deformed if subjected to trauma. However, after the molt LHS was terminated, feathers were keratinized and no longer vulnerable. Males began to sing and defend territories vigorously in late September and October, the beginning of the non-breeding LHS (Arcese, 1989). At this time there was a resur- FIG. 2. Seasonal changes in aggressive responses to simulated territorial intrusion (STI) in free-living male rufous song sparrows, Melospiza melodia morphna. Note that there was an increase in number of songs and flights, males approached more closely to the STI and spent more time within five meters of the decoy in spring as breeding got underway. A high level of response was maintained through the breeding season and declined to a low when in molt (August and September). In autumn, as the non-breeding LHS got underway, males showed a second increase in aggressive response to STI. There are two periods of heightened territorial aggression in two different LHSs. From Wingfield and Hahn (1994).

3 CONTROL OF TERRITORIALITY 13 FIG. 3. Biological actions of the steroid hormone testosterone. The morphological, physiological and behavioral actions of testosterone that are essential for male reproductive function are given on the right hand and lower sides of the figure. The costs of prolonged high levels of testosterone are given on the left hand side in italics. The patterns of plasma testosterone levels may be a function of secretion patterns to maintain male reproductive function, and costs of testosterone that require that plasma levels be low. From Wingfield et al. (2000). gence of territorial aggression in response to STI that is identical to that in spring (Fig. 2) indicating that territorial aggression of males song sparrows in spring and autumn were similar. There is evidence, however, that song stereotypy in autumn may be less than in spring (Smith et al., 1997), although all other measures of aggression during STI appear to be the same (Wingfield and Hahn, 1994). CONTEXT OF SPRING AUTUMN TERRITORIAL BEHAVIOR By following the behavior and location of colorbanded song sparrows throughout the year, it was found that in some areas, both males and females moved territories between breeding and non-breeding LHSs (Wingfield and Monk, 1992; Wingfield, 1994a). For example, a territory at the high tide mark by a beach may be a high quality location for breeding, but it is exposed to inclement weather in winter. Song sparrows moved to more sheltered locations after breeding. This movement may be a few meters (perhaps overlap with the breeding territory), whereas others may move 200 m or more to a completely separate winter territory. Although many pairs of song sparrows may stay together on the same territory throuhout the year, many others appear to form an alliance with another individual. Most breeding territories have a male and a female whereas the remainder usually have a single, unmated male. In the non-breeding LHS, only 30% of pairs are male-female. The rest may be single birds of either sex, male-male pairs or associations of three and occasionally more birds on a territory (Wingfield and Monk, 1992; Wingfield, 1994a). Furthermore, of the male-female pairs on wintering territories in some locations, very few were breeding pairs. In many cases the male moved to a separate breeding territory in spring and paired with a different female (Wingfield and Monk, 1992; Wingfield, 1994a). The conclusion from these studies was that north-western song sparrows form an alliance with one or more birds in autumn, and do not necessarily remain as a breeding pair the next spring. In winter these birds are apparently defending resources to survive and can form an alliance with another regardless of gender. In spring, mated pairs are defending territories on which they breed and thus necessarily must be male and female. Although the expression of territorial aggression appears identical in breeding and non-breeding LHSs, the context is not the same. CORRELATES WITH TESTOSTERONE AND SOCIAL INTERACTIONS It is well known that testosterone activates aggression associated with male-male competition over territories and mates (Balthazart, 1983; Harding, 1983), although the correlation of plasma levels of testosterone with expression of territorial aggression when breeding are frequently unclear (Soma et al., 2000a). It is thought that baseline levels of testosterone during the breeding season result in development and maintenance of morphological, physiological and behavioral components of the male reproductive system (Fig. 3). However, these baseline levels do not necessarily correlate with actual expression of territorial aggres-

4 14 J. C. WINGFIELD AND K. K. SOMA FIG. 4. The pattern of testosterone secretion in free-living populations of the rufous song sparrow, Melospiza melodia morphna. Plasma levels peak in April and May as the breeding LHS got underway and then were maintained at a lower breeding baseline during the rest of the breeding season. As prebasic molt ensued, plasma levels of testosterone were basal and remained so throughout autumn and winter. Compare this pattern with changes in expression of territorial aggression in Figure 1. Testosterone concentrations in blood peaked during the spring increase in territoriality, but not in autumn. From Wingfield and Hahn (1994). sion. Superimposed on this breeding baseline of circulating testosterone level are transient surges to much higher concentrations that are tightly correlated with periods of heightened male-male competition, especially when establishing a territory, being challenged by another male, or when mate guarding. This is the challenge hypothesis stating that high plasma levels of testosterone occur during periods of social instability in the breeding season, but are at a lower breeding baseline in stable social conditions (Wingfield et al., 1990, 2000). Because territories in the non-breeding LHS are defended and maintained by aggressive behavior that appears identical to that during in the breeding LHS, it is tempting to assume that testosterone plays a role in both LHSs. This view is supported by investigations of another apparently testosterone-dependent behavior, autumn sexuality. This phenomenon also includes male-male competition for different resources in some avian species. It is well known in Europe that the rook, Corvus frugilegus, returns to breeding colonies in autumn. In some years eggs were laid, although these breeding attempts were never successful (Marshall and Coombs, 1957). However, in most individuals in the population the gonads were not as developed as in spring Marshall and Coombs (1957). In contrast, Lincoln et al. (1980) could not find any increase in testis size during autumn (October and November) in a population of rooks in the United Kingdom. They did measure a pronounced peak in LH, but not testosterone, despite expression of courtship behavior. Similar transient surges of LH and/or testosterone in autumn have been measured in herring gulls, Larus argentatus, (Scanes et al., 1974), red grouse, Lagopus lagopus scoticus, (Sharp et al., 1974), mallards, Anas platyrhynchos, (Haase et al., 1975), eider duck, Somateria mollissima, (Gorman, 1974; Spurr and Milne, 1976), and Parus sp., (Röhss and Silverin, 1983; Silverin et al., 1984, 1989). It has been argued that all these species show some form of reproductively relevant behavior in autumn such as courtship and pair formation, or establishment of a territory on which the individual will eventually breed (Wingfield et al., 1997). This may mark the beginning of a potential breeding LHS, but further development and breeding is suppressed until spring. In these cases the breeding and nonbreeding LHSs must overlap considerably (Wingfield et al., 1997). The next question is, do those species in which there are no reproductive contexts to autumn territoriality also show an increase in plasma testosterone levels at this time? Avian species that establish autumn territories in non-reproductive contexts apparently do not show similar surges in testosterone and/or LH. Plasma levels of testosterone were high in lesser sheathbills, Chionis minor, when on breeding territory but not on the nonbreeding territory, even though the behavior expressed appeared similar in both LHSs (Burger and Millar, 1980). Similarly in European robins, Erithacus rubecula, (Schwabl and Kriner, 1991) there were no changes in reproductive hormones levels in autumn when both sexes established independent winter feeding territories. Migratory stonechats, Saxicola torquata, established territories as apparent male and female pairs on their wintering grounds in Israel. These pairs did not leave on spring migration together and pair bonds were not always stable in winter. Both sexes showed territorial aggression but did not have elevated testosterone levels during this period (Gwinner et al., 1994). The sedentary northern mockingbird, Mimus polyglottos, may be an exception because they show no autumn peaks of LH or testosterone yet their autumn/ winter territories were used later as breeding territories (Logan and Wingfield, 1990). Tropical birds that were territorial throughout their reproductive life (perhaps many years) may also be an exception (Wikelski et al., 1999). This may be similar to the situation in song sparrows in which some winter territories may also be breeding territories whereas others may not (Wingfield 1994a, b). Plasma levels of testosterone were highest early in the breeding LHS when males were on breeding territories. Circulating levels remained at a breeding baseline throughout the rest of the breeding season, and then declined to undetectable levels when the prebasic molt began and expression of territorial behavior was lowest (Fig. 4, Wingfield and Hahn, 1994). After the prebasic molt, when the non-breeding LHS began in autumn, there was a resurgence of territorial aggression but no change in testosterone levels (Fig. 4, Wingfield and Hahn, 1994). Despite low testosterone concentrations in autumn it is possible that transitory peaks were missed. We know that in the breeding

5 CONTROL OF TERRITORIALITY 15 LHS, high plasma testosterone levels were correlated with increased territorial aggression only during periods of social instability the challenge hypothesis. It is possible that this could be in effect in autumn also. To test this, free-living male song sparrows were removed from their territories in autumn. Other floater males take over, and a period of social instability follows as the replacement male interacts with neighbors to establish a territory. These replacement males and their neighbors have significantly elevated plasma levels of testosterone when the experiment is conducted in spring (Wingfield, 1985) but not in autumn (Wingfield, 1994a, b). Similarly, it is well established that plasma testosterone levels increase if males of socially monogamous species are challenged by, for example, STI during the breeding season (Wingfield, 1985; Wingfield et al., 1990). However, STIs in song sparrows in autumn have no effect on LH and testosterone levels (Wingfield, 1994a, b; Wingfield and Hahn, 1994; Soma and Wingfield, 2001). Another potent stimulus for LH and testosterone secretion in male socially monogamous birds during the breeding LHS is sexual behavior of females (e.g., Moore, 1982, 1983). Free living female song sparrows implanted with estradiol in autumn became sexually receptive and solicited copulations from nearby males, but their behavior had no effect on territorial aggression in response to STI in these males, and had no effect on plasma testosterone levels (Wingfield and Monk, 1994). It is possible that estradiol treatment of these females in autumn did not activate the full repertoire of female courtship behavior and thus were not effective to elicit a hormone response in males. However, by February, as day length increased and males began to show some reproductive development, plasma levels of testosterone were higher in male associated with estrogen treated females compared with males associated with controls (Wingfield and Monk, 1994). Finally, free-living males that were castrated continued to defend territories and respond to STI in autumn equally well as sham operated males (Wingfield, 1994a, b). Clearly, if testosterone is involved in the expression of autumn territoriality it must do so through a different mechanism than during spring. This concept is bolstered further by the observation that independent fledgling male song sparrows from the first brood established territories and sing full songs to defend them in August and September (Nordby et al., 1999). At this time adult males were in molt and much less territorial. This aggression by first year males was not accompanied by elevated plasma levels of LH or testosterone, and the testes remained in a juvenile state (Wingfield, 1994a, b; Wingfield and Hahn, 1994). Given that even castrated males can defend a territory perfectly well, and that song sparrows are territorial almost year round, do high levels of testosterone in spring have any role in territorial aggression? The response of male song sparrows to STI during the breeding LHS typically involved high levels of aggression directed toward the decoy (Fig. 2). However, we also found that after the decoy and speaker (i.e., source of the STI) were removed, males continued to show high levels of spontaneous aggression for many hours or even up to two days later (Wingfield, 1994a, b). In contrast, in autumn, aggressive responses of male song sparrows to STI were similar to those in the breeding LHS, but after removal of the decoy and speaker territorial aggression declined within minutes. Implants of testosterone into male song sparrows in autumn increased the level of aggression expressed in response to STI and also maintained a high level of spontaneous aggression after the decoy and speaker were removed (Wingfield, 1994a, b). Thus in song sparrows, testosterone appears to increase persistence of aggression following an intrusion rather than activate aggression per se. This would be highly adaptive in the breeding season when reproductive success is at stake, but would not be adaptive in autumn when other strategies (switch territories, float) are possible (Wingfield, 1994a, b). It is possible that territorial aggression mediated via high testosterone levels may be important during periods of social instability in the breeding LHS but not in other LHSs. Given that testosterone also has marked effects on sexual behavior as well as morphology of reproductive accessory organs etc. (see Fig. 3), secretion of this hormone in an LHS other than breeding would be inappropriate (Wingfield et al., 1997, 2000). SAME BEHAVIOR IN SPRING AND AUTUMN, DIFFERENT MECHANISMS? Results of investigations presented thus far lead us to the hypothesis that territorial aggression in the nonbreeding season is independent of sex steroid hormones. The observation that experimental elevation of testosterone in male song sparrows in autumn made them more aggressive (Wingfield, 1994b) at first appears incompatible with the hypothesis. To examine a possible role of endogenous sex steroids on territorial behavior in non-breeding song sparrows further, field experiments using pharmacological agents to block receptors and modify metabolism of testosterone in target cells were conducted. During the breeding LHS it has been established that territorial aggression displayed is dependent upon the rate by which neurons in the avian brain aromatize testosterone to estradiol (Schlinger and Callard, 1990; Foidart et al., 1998; Balthazart et al., 1999; Silverin et al., 1999). Androgens are necessary precursors of estrogens (Fig. 5). This may also be true for territoriality in the non-breeding LHS. Treatment of free-living European robins in autumn and winter with an antiandrogen (flutamide that prevents androgen from binding to its receptor) only, did not affect territorial aggression suggesting that androgen receptors are not primarily involved at this time (Schwabl and Kriner, 1991). Field experiments in territorial male song sparrows using both flutamide and an aromatase inhibitor

6 16 J. C. WINGFIELD AND K. K. SOMA FIG. 5. A simplified diagram of androgen and estrogen synthesis in a typical vertebrate. There is growing evidence that most or all of these enzymes are present in the brain of song birds (Tsustui and Yamakazi, 1995; Vanson et al., 1996; Schlinger et al., 1999; Ukeda et al., 1999). (ATD) significantly reduced territorial behavior both during and after STI in autumn and winter (Soma et al., 1999a; Soma and Wingfield, 1999). The same result was obtained when a potent aromatase inhibitor (Fadrozole) was used alone (Soma et al., 2000a, b). The effects in autumn were repeatable both by acute treatment with Fadrozole as well as more long-term administration by osmotic mini-pumps (Soma et al., 2000b). In all these experiments there were no effects of inhibitors on body condition suggesting that the experimental birds were not debilitated and thus showed less aggression (Soma et al., 1999a, 2000a, b). By blocking the aromatase activity, but not by just blocking androgen receptors, autumn territorial aggressive behavior was reduced. These results strongly suggest that estrogens (presumably aromatized from androgen in the brain, Fig. 5) were involved in the regulation of territorial behavior outside the breeding season. Furthermore, a very important experiment in which song sparrows were given Fadrozole in combination with estradiol implants as replacement, completely restored territorial aggression in response to STI (Soma et al., 2000a). These data emphasize that estrogen, resulting from aromatization of androgen, and acting through estrogen receptors is critical for the expression of territorial aggression in autumn, whereas androgen receptor-mediated mechanisms appear less important. These experiments clearly point to the conversion of testosterone to estradiol as being important for the expression of territorial aggression in autumn. They do not support the hypothesis that winter territoriality is independent of sex steroids. If this is true the next questions are: 1) Why should plasma levels of testosterone be undetectable during the non-breeding LHS? 2) Castration also did not affect territorial behavior why? During the breeding LHS it is important to remember that testosterone regulates a number of behavioral traits such as sexual displays, song and aggressive behavior. It also regulates many morphological events such as development of secondary sex characters and accessory organs, affects spermatogenesis and muscle hypertrophy. Extensive field studies have indicated that prolonged elevation of testosterone levels during breeding may incur costs such as increased rate of injury and depredation, reduced fat stores etc. (Fig. 3, Dufty, 1989; Ketterson et al., 1992, 1996; Beletsky et al., 1995). Furthermore, high testosterone levels interfere with parental care (Silverin, 1980; Hegner and Wingfield, 1987), and in some species may impair the immune system (Hillgarth and Wingfield, 1997). Taken together these ecological constraints and the costs associated with prolonged high levels of testosterone may have had a profound influence on the evolution of hormone-behavior mechanisms (Wingfield et al., 1997b, 1999). It is therefore important to acknowledge the concept that testosterone secretion must be decreased whenever possible, or that the potential costs and inappropriate actions (such as in a nonbreeding LHS) be avoided by some mechanism. Thus, although it appears that the hormone dependent mechanisms controlling spring and autumn territoriality may remain the same at cell and molecular levels (receptors and androgen metabolizing enzymes), the mode of hormone delivery to the target cells in the brain differs. The expressions of the enzymes involved in the metabolism of testosterone may hold the key to mechanisms underlying strategies in behavioral ecology. AVOIDING THE COSTS OF TESTOSTERONE We have three hypotheses to explain how the potential costs of testosterone secreted into blood in nonbreeding LHSs could be ameliorated (Soma and Wingfield, 2000). (1) Target neurons in the central nervous system (CNS) may become highly sensitized to low testosterone levels the increased sensitivity hypothesis (Soma and Wingfield, 1999). (2) There is growing evidence that steroid synthesis may occur de novo from cholesterol within the brain the neurosteroid hypothesis. (3) A biologically inactive androgen precursor, such as dehydroepiandrosterone (DHEA) may be secreted by, for example, the adrenals. This could then be converted to the active hormone within the brain the circulating precursor hypothesis. These hypotheses are not mutually exclusive, and others may come to light in the future (Soma and Wingfield, 1999). Increased sensitivity hypothesis Increased sensitivity of the brain to low levels of sex steroids during the non-breeding LHS is a possible but unlikely hypothesis. If this is true in song sparrows, then sex steroids must originate from non-gonadal sites (e.g., the adrenals) because castration in autumn had no effect on territorial aggression (Wingfield, 1994b). The brain could become more sensitive to steroids in autumn by increased hormone receptors, elevated expression of aromatase, or decreased 5 -reductase (a deactivation shunt that coverts testoster-

7 CONTROL OF TERRITORIALITY 17 one to 5 -dihydrotestosterone, a metabolite that does not bind to the androgen receptor with high affinity). This hypothesis is unlikely based on behavioral and neurobiological studies. Several different lines of evidence indicate that birds have a decreased sensitivity to testosterone outside the breeding LHS. These include a decrease in the number of androgen receptors (AR), the efficacy of testosterone to activate postbreeding singing is less, and hypothalamic aromatase activity is reduced during the non-breeding period (e.g., Hutchison et al., 1986; Nowicki and Ball, 1989; Schlinger and Callard, 1990; Silverin and Deviche, 1991; Soma et al., 1999b; Gahr and Metzdorf, 1997; Ball, 1999). In the canary, Serinus canarius, expression levels of estrogen receptor (ER) in the telencephalon were higher in November (lower levels of sex steroids and rates of singing) than in April, but aromatase-mrna levels were higher in November. Levels of AR-nRNA were similar in November and April (Fusani et al., 1999). Clearly more studies are needed, including investigations of expression of ER and ER, before this hypothesis is fully tested. Neurosteroid hypothesis The relatively new concept of the brain being able to synthesize steroids de novo (Baulieu, 1998) raises an attractive possibility for regulation of autumn territorial aggression. Evidence that the brain can synthesize sex steroids from cholesterol de novo was supported by high local concentrations of steroids in brain that were not paralleled by circulating levels in plasma (Robel and Baulieu, 1995; Tsutsui and Yamakazi, 1995; Mensah-Nyagan et al., 1996; Baulieu, 1998; Ukena et al., 1999). Evidence for steroidogenic enzymes (protein and mrna activity) has been found in brain. The presence of P450c17 in adult brain remains unclear in mammals (Compagnone and Mellon, 2000), but recent reports suggest that it may be present in brains of adult birds (Nomura et al., 1998). More recent studies give specific localization of key enzymes, protein and mrna, in specific areas of the brain (Fig. 5, Vanson et al., 1996; Schlinger et al., 1999; Soma et al., 1999c; Ukena et al., 1999). These enzymes could create high levels of sex steroids in brain, independent of the gonads and other peripheral tissues. A critical question then becomes: what regulates these enzymes in brain? Circulating precursor hypothesis An alternate possibility also involving steroidogenic enzymes in the brain is the circulating precursor hypothesis stating that there is peripheral production of a biologically inert sex steroid precursor that is converted to an active hormone in the brain (Labrie et al., 1995). There are several possibilities, but dehydroepiandrosterone (DHEA) is a possible candidate and enzymatic activity in brain could convert this steroid to androstenedione that in turn, could be converted to testosterone or aromatised to estradiol (Labrie et al., 1995; Vanson et al., 1996; Ukena et al., 1999). Such precursors may be of extra-gonadal origin such as the adrenal (Labrie et al., 1995; Schlinger et al., 1999; Soma and Wingfield, 2001). Very recent data show that plasma levels of DHEA in free-living song sparrows were elevated during the breeding LHS, declined in the prebasic molt LHS, when territorial aggression was lowest, and then increased again in autumn coincident with heightened territorial aggression (Soma and Wingfield, 2001). Implants of DHEA into male song sparrows increased singing, one component of territorial aggression, and also resulted in growth of HVc, a song control nucleus in the telencephalon (Soma et al., 2000c). The key experiment in which DHEA levels are reduced has yet to be conducted. It should also be noted that the neurosteroid and circulating precursor hypotheses are not be mutually exclusive. CONCLUSIONS DIVERSE MECHANISMS IN ENVIRONMENTAL ENDOCRINOLOGY Finite state machine theory describes the organization of life history stages (LHSs) in the life cycle of individuals. We can use this organization to explore how the endocrine system regulates changes in morphology, physiology and behavior characteristic of each LHS (Jacobs and Wingfield, 2000). One fascinating aspect of this approach is that certain traits may occur in different LHSs. Does this mean they are regulated by the same hormone systems, or are they different? Many physiological traits are common to most if not all LHSs. For example, glucose storage and mobilization is a constant balance that is orchestrated very finely by hormones such as epinephrine, insulin and glucagon (and others, e.g., Norris, 1999). These hormones act throughout the individual s life cycle. However, other traits expressed in two or more LHSs and regulated by hormones specialized for a specific LHS (such as breeding) may pose considerable problems. In spring and autumn territoriality, behavioral traits expressed in the establishment and maintenance of a territory appear to be identical in the breeding and nonbreeding LHSs although the context of that behavior may be different. There is extensive evidence that testosterone is involved in the activation and maintenance of aggression in reproductive contexts, but its role in expression of territorial aggression at other times in the life cycle is unclear and problematic (Wingfield et al., 1997). It is now abundantly clear that prolonged high levels of testosterone can incur costs that lead to reduced fitness (Ketterson et al., 1996; Wingfield et al., 2000). Additionally, the many morphological, physiological and behavioral actions of testosterone essential for male reproductive function would not be appropriate in the non-breeding LHSs. The hypothesis then can be raised, if a behavioral trait is regulated by a hormone in one LHS, then the identical behavior in another LHS may be regulated by that same hormone if the costs associated with it are somehow avoided thus maximizing fitness. For the example of testosterone

8 18 J. C. WINGFIELD AND K. K. SOMA and spring and autumn territoriality, it appears that cellular mechanisms of hormone control appear to be similar in spring and autumn, but the mode of delivery is different. In other words, circulating testosterone levels remain very low (below detectable limits) in autumn, but at the target neuron level, sex steroid receptors remain intact (Soma and Wingfield, 1999; Soma et al., 2000a). In this way the additional effects of testosterone relating to male reproductive function are avoided in autumn. Three hypotheses are presented to explain how these costs of high plasma testosterone could be avoided. They include increased brain sensitivity to testosterone, synthesis of sex steroids de novo from cholesterol in the brain (neurosteroid hypothesis), and secretion of a biologically inactive precursor (e.g., DHEA) that can then be metabolized to an active form in target tissues (circulating precursor hypothesis). Current research is exploring these possibilities. It is also important to recognize here that field and laboratory investigations are essential to determine how animals use the endocrine system to orchestrate their life cycles. Investigations in the field have provided us with a perspective on ecological bases of patterns of behavior and the costs associated with prolonged high levels of sex steroids. It is unlikely that these perspectives could have been realized in laboratory experiments on conventional animal models alone. We have discussed one hormone-behavior interaction here, but it is possible that many more examples exist throughout vertebrate taxa. They will only be revealed by taking endocrinology to the field in animals interacting with their real world. Ecological bases of hormone actions may have had a strong influence on the evolution of mechanisms including, paradoxically, ways to preserve one action of a hormone in different LHSs but avoid potential costs that may accompany it. We can then return to the laboratory to pursue these mechanisms at the cell and molecular levels. Additionally, given the diversity of morphological, physiological and behavioral combinations expressed even within a population from season to season, as well as among different populations, we can exploit them as natural experiments to determine those mechanisms. ACKNOWLEDGMENTS JCW is grateful for a series of grants from the Division of Intergrative Biology and Neuroscience, National Science Foundation, a John Simon Guggenheim Foundation Fellowship and the Russell F. Stark University Professorship from the University of Washington. KKS received a Howard Hughes Medical Institute Predoctoral Fellowship, and currently is a National Institutes of Health Postdoctoral Fellow. REFERENCES Andreev, A. V Energetics and survival of birds in extreme environments. Ostrich 70: Arcese, P Territory acquisition and loss in male song sparrows. Anim. Behav. 37: Ball, G Neuroendocrine basis of seasonal changes in vocal behavior among songbirds. In M. Hauser and M. Konishi (eds.), The design of communication, pp MIT Press Cambridge, Massachusetts. Balthazart, J Hormonal correlates of behavior. In D. S. Farner, J. R. King, and K. C. Parkes (eds.), Avian biology, Vol. 7, pp Academic Press, New York. Balthazart, J., A. Foidart, M. Baillien, and B. Silverin Brain aromatase in laboratory and free-living songbirds: Relationships with reproductive behaviour. In N. J. Adams and R. H. Slotow (eds.), pp Proc. 22 Int. Ornithol. Congr., Durban. Baulieu, E Neurosteroids: A novel function of the brain. Psychoneuroendocrinology 23: Becker, J. B., S. M. Breedlove, and D. Crews. (eds.) Behavioral endocrinology. Massachusetts Institute of Technology Press, Cambridge. Beletsky, L. D., D. F. Gori, S. Freeman, and J. C. Wingfield Testosterone and polygyny in birds. Curr. Ornithol. 12:1 41. Burger, A. E. and R. P. Millar Seasonal changes of sexual and territorial behavior and plasma testosterone levels in male lesser sheathbills (Chionis minor). Z. Tierpsychol. 52: Dufty, A. M Testosterone and survival. Horm. Behav. 23: Foidart, A., B. Silverin, M. Baillien, N. Harada, and J. Balthazart Neuroanatomical distribution and variations across the reproductive cycle of aromatase activity and aromatase-immunoreactive cells in the pied flycatchers (Ficedula hypoleuca). Horm. Behav. 33: Fusani, L., T. Van t Hof, J. B. Hutchison, and M. Gahr Seasonal expression of androgen receptors, estrogen receptors, and aromatase in the canary brain in relation to circulating androgens and estrogens. J. Neurobiol. 43: Gahr, M. and R. Metzdorf Distribution and dynamics in the expression of androgen and estrogen receptors in vocal control systems of songbirds. Brain Res. Bull. 44: Gorman, M. L The endocrine basis of pair-formation behaviour in the male eider Somateria mollissima. Ibis 116: Gwinner, E., T. Rödl, and H. Schwabl Pair territoriality of wintering stonechats: behaviour, function and hormones. Behav. Ecol. Sociobiol. 34: Haase, E., P. J. Sharp, and E. Paulke Annual cycle of plasma lueinizing hormone concentrations in wild mallard drakes. J. Exp. Zool. 194: Harding, C. F Social modulation of circulating hormone levels in the male. Amer. Zool. 21: Hegner, R. E. and J. C. Wingfield Effects of experimental manipulation of testosterone levels on parental investment and breeding success in male house sparrows. Auk 104: Hillgarth, N. and J. C. Wingfield Parasite-mediated sexual selection: Endocrine aspects. In D. H. Clayton and J. Moore (eds.), Host-parasite evolution, pp Oxford University Press, Oxford. Hutchison, J., T. Steimer, and D. Jaggard Effects of photoperiod on formation of oestradiol-17 in the dove brain. J. Endocrinol. 109: Jacobs, J. D. and J. C. Wingfield Endocrine control of lifecycle stages: A constraint on response to the environment? Condor 102: Ketterson, E. D., V. Nolan, Jr., L. Wolf, and C. Ziegenfus Testosterone and avian life histories: Effects of experimentally elevated testosterone on behavior and correlates of fitness in the dark-eyed junco (Junco hyemalis). Am. Nat. 140: Ketterson, E. D., V. Nolan, Jr., M. J. Cawthorn, P. G. Parker, and C. Ziegenfus Phenotypic engineering: Using hormones to explore the mechanistic and functional bases of phenotypic variation in nature. Ibis 138: Labrie, F., A. Belanger, J. Simard, V. Luu-the, and C. Labrie DHEA and peripheral androgen and estrogen formation: Intracrinology. Ann. N.Y. Acad. Sci. 774:16 28.

9 CONTROL OF TERRITORIALITY 19 Lincoln, G. A., P. A. Racey, P. J. Sharp, and H. Klandorf Endocrine changes associated with spring and autumn sexuality of the rook Corvus frugilegus. J. Zool. 190: Logan, C. A. and J. C. Wingfield Autumnal territorial aggression is independent of plasma testosterone in mockingbirds. Horm. Behav. 24: Marshall, A. J. and C. J. F. Coombs The interaction of environmental, internal and behavioural factors in the rook, Corvus frugilegus L. Proc. Zool. Soc. London 128: Moore, M. C Hormonal responses of free-living male whitecrowned sparrows to experimental manipulation of female sexual behavior. Horm. Behav. 16: Moore, M. C Effect of female sexual displays on the endocrine physiology and behavior of male white-crowned sparrows, Zonotrichia leucophrys gambelii. J. Zool. London 199: Nelson, R An introduction to behavioral endocrinology. Sinauer, Sunderland, Massachusetts. Nomura, O., K. Nishimori, O. Nakabayashi, H. Yasue, and S. Mizuno Determination by modified RT-PCR of transcript amounts from genes involved in sex-steroid synthesis in chicken organs including brain. J. Steroid Biochem. Molec. Biol. 67: Nordby, J., S. Campbell, and M. Beecher Ecological corelates of song learning in song sparrows. Behav. Ecol. 10: Norris, D Vertebrate endocrinology. Academic Press, New York. Nowicki, S. and G. F. Ball Testosterone induction of song in photosensitive and photorefractory male sparrows. Horm. Behav. 23: Robel, P. and E. Baulieu Dehydroepiandrosterone (DHEA) is a neuroactive neurosteroid. Ann. N.Y. Acad. Sci. 774: Röhss, M. and B. Silverin Seasonal variation in the ultrastructure of the Leydig cells and plasma levels of luteinizing hormone and steroid hormones in juvenile and adult male great tits Parus major. Ornis Scand. 14: Scanes, C. G., P. Cheeseman, J. G. Phillips, and B. K. Follett Seasonal and age variation of circulating immunoreactive luteinizing hormone in captive Herring gulls, Larus argentatus. J. Zool. London 174: Schlinger, B. A. and G. V. Callard Aromatization mediates aggressive behavior in quail. Gen. Comp. Endocrinol. 79: Schlinger, B., N. Lane, W. Grisham, and L. Thompson Androgen synthesis in a songbird: A study of Cyp17 (17 -hydroxylase/c17,20-lyase) activity in the zebra finch. Gen. Comp. Endocrinol. 113: Schwabl, H. and E. Kriner Territorial aggression and song of male European robins (Erithacus rubecula) in autumn and spring: Effects of antiandrogen treatment. Horm. Behav. 25: Sharp, P. J., R. Moss, and A. Watson Seasonal variation in plasma luteinizing hormone levels in male red grouse, Lagopus lagopus scoticus. J. Endocr. 64:1 44. Silverin, B Effects of long-acting testosterone treatment on free-living pied flycatchers, Ficedula hypoleuca, during the breeding period. Anim. Behav. 28: Silverin, B. (ed.) Endocrinology of the non-breeding season in birds. Ornis Scandinavica 23. Silverin, B., M. Baillien, and J. Balthazart Territorial aggression correlates with preoptic aromatase activity but not with plasma testosterone in free-living pied flycatchers. Proc. Soc. Behav. Neuroendocrinol. Abs. p 64. Silverin, B. and P. Deviche Biochemical characterization and seasonal changes in the concentration of testosterone-metabolizing enzymes in the European great tit (Parus major) brain. Gen. Comp. Endocrinol. 81: Silverin, B., P. A. Viebke, and J. Westin Plasma levels of luteinizing hormones and steroid hormones in free-living winter groups of willow tits (Parus montanus). Horm. Behav. 18: Silverin, B., P. A. Viebke, and J. Westin Hormonal correlates of migration and territorial behavior in juvenile willow tits during autumn. Gen. Comp. Endocrinol. 75: Smith, G. T., E. A. Brenowitz, M. D. Beecher, and J. C. Wingfield Seasonal changes in testosterone, neural attributes of song control nuclei and song structure in wild songbirds. J. Neurosci. 17: Soma, K., R. Bindra, J. Gee, J. Wingfield, and B. Schlinger. 1999c. Androgen-metabolizing enzymes show region-specific changes across the breeding season in the brain of a wild songbird. J. Neurobiol. 41: Soma, K., V. Hartman, J. Wingfield, and E. Brenowitz. 1999b. Seasonal changes in androgen receptor immunoreactivity in the song nucleus HVc of a wild bird. J. Comp. Neurol. 409: Soma, K. K., K. A. Sullivan, A. D. Tramontin, C. J. Saldanha, B. A. Schlinger, and J. C. Wingfield. 2000b. Acute and chronic effects of an aromatase inhibitor on territorial aggression in breeding and non-breeding male song sparrows. J. Comp. Physiol. A 186: Soma, K. K., K. Sullivan, and J. Wingfield. 1999a. Combined aromatase inhibitor and antiandrogen treatment decreases territorial aggression in a wild songbird during the nonbreeding season. Gen. Comp. Endocrinol. 115: Soma, K. K., A. D. Tramontin, and J. C. Wingfield. 2000a. Oestrogen regulates male aggression in the non-breeding season. Proc. R. Soc. London; B 267: Soma, K. K. and J. C. Wingfield Endocrinology of aggression in the nonbreeding season. In N. J. Adams and R. H. Slotow (eds.), Proc. 22 Int. Ornithol. Congr. Durban, pp BirdLife, South Africa. Soma, K. K. and J. C. Wingfield Dehydroepiandrosterone in plasma of a song bird. (submitted). Soma, K. K., A. Wissman, E. Brenowitz, and J. C. Wingfield. 2000c. Effects of dehydroepiandrosterone (DHEA) on behavior and neuroanatomy of a song bird. Soc. Neurosci. Abstr. 26:1762. Spurr, E. and H. Milne Adaptive significance of autumn pair formation in the common eider Somateria mollissima (L.) Ornis Scand. 7: Tsutsui, K. and T. Yamakazi Avian neurosteroids. I. Pregnenolone synthesis in the quail brain. Brain Res. 678:1 9. Ukena, K., Y. Honda, Y. Inai, C. Kohchi, R. Lea, and K. Tsutsui Expression and activity of 3 -hydroxysteroid dehydrogenase/ 5-4 -isomerase in different regions of the avian brain. Brain Res. 818: Vanson, A., A. P. Arnold, and B. A. Schlinger ß-hydroxysteroid dehydrogenase/isomerase and aromatase activity in primary cultures of developing zebra finch telencephalon: Dehydroepiandrosterone as substrate for synthesis of androstendione and estrogens. Gen. Comp. Endocrinol. 102: Wikelski, M., M. Hau, W. D. Robinson, and J. C. Wingfield Seasonal endocrinology of tropical passerines a comparative approach. In N. Adams and R. Slotow (eds.), Proceedings of the 22nd International Ornithological Congress, pp BirdLife South Africa, Johannesburg. Wingfield, J. C Short-term changes in plasma levels of hormones during establishment and defense of a breeding territory in male song sparrows, Melospiza melodia. Horm. Behav. 19: Wingfield, J. C. 1994a. Control of territorial aggression in a changing environment. Psychoneuroendocrinol. 19: Wingfield, J. C. 1994b. Regulation of territorial behavior in the sedentary song sparrow, Melospiza melodia morphna. Horm. Behav. 28:1 15. Wingfield, J. C. and T. P. Hahn Testosterone and territorial behaviour in sedentary and migratory sparrows. Anim. Behav. 47: Wingfield, J. C., R. E. Hegner, A. M. Dufty, A. M. Jr., and G. F. Ball The challenge hypothesis : Theoretical implications for patterns of testosterone secretion, mating systems, and breeding strategies. Am. Nat. 136: Wingfield, J. C., J. Jacobs, and N. Hillgarth Ecological constraints and the evolution of hormone-behavior interrelationships. Ann. N.Y. Acad. Sci. 807: Wingfield, J., J. Jacobs, K. Soma, D. Maney, K. Hunt, D. Wisti-

10 20 J. C. WINGFIELD AND K. K. SOMA Peterson, S. Meddle, M. Ramenofsky, and K. Sullivan Testosterone, aggression, and communication: Ecological bases of endocrine phenomena. In M. Hauser and M. Konishi (eds.), The design of animal communication, pp MIT Press, Cambridge, Massachusetts. Wingfield, J. C., J. D. Jacobs, A. D. Tramontin, N. Perfito, S. Meddle, D. L. Maney, and K. Soma Toward and ecological basis of hormone-behavior interactions in reproduction of birds. In K. Wallen and J. Schneider (eds.), Reproduction in context, pp MIT Press, Cambridge. Wingfield, J. C. and D. Monk Control and context of yearround territorial aggression in the non-migratory song sparrow, Melospiza melodia morphna. Ornis Scand. 23: Wingfield, J. C. and D. Monk Behavioral and hormonal responses of male song sparrows to estrogenized females during the non-breeding season. Horm. Behav. 28: