themselves or for their chicks?

Transcription

1 Ecology , Food patch use by parent birds: to gather food for Blackwell Publishing, Ltd. themselves or for their chicks? SHAI MARKMAN*, BERRY PINSHOW*, JONATHAN WRIGHT and BURT P. KOTLER* *Jacob Blaustein Institute for Desert Research and Department of Life Sciences, Ben-Gurion University of the Negev, Sede Boqer Campus, 84990, Midreshet Ben-Gurion, Israel; and School of Biological Sciences, University of Wales, Bangor, Gwynedd LL57 2UW, UK Summary 1. Despite its importance in the evolution of central-place foraging strategies, few studies have distinguished between patch use by parent birds for their own consumption vs. patch use for the purpose of provisioning young in the nest. 2. Adult Palestine sunbirds (Nectarinia osea) consume nectar and arthropods, but feed only arthropods to their nestlings. This partial separation of food types allowed us to experimentally test sex-specific trade-offs in parental foraging effort between two patches. 3. A variable patch on one side of a pair s territory contained only artificial nectar for parental self-feeding (in varying sucrose concentrations between 0 25 and 0 75 mol kg 1 H 2 O). A constant patch on the other side of the territory contained artificial nectar for parental self-feeding (sucrose concentration = 0 25 mol kg 1 H 2 O), plus a dish of flightless houseflies for provisioning young. 4. Experimental increases in sucrose concentration in the variable patch led the parents to: (a) consume greater amounts of sucrose solution from this patch; (b) collect more flies from the constant patch to feed their young; and (c) persistently use the 0 25 mol kg 1 H 2 O feeder near the dish of flies. By titration (gradually changing the sucrose concentration in the variable patch), we demonstrate that at concentrations of 0 47 mol kg 1 H 2 O for males and 0 55 mol kg 1 H 2 O for females, the birds spent an equal amount of time foraging in both patches. Therefore, sucrose concentration in the variable patch (for self-feeding only), had to be higher for the females than for the males, in order to balance the additional value females placed upon the flies provided for chick-feeding in the constant patch. 5. These results reflect previously described sex-specific parental roles in this species, with females provisioning young mainly in the nest and males patrolling more and guarding against nest predators. Furthermore, we demonstrate for the first time the fine details of alternative adaptive trade-offs in foraging effort by parent birds when using patches for self-feeding or for delivery of food to the young. Key-words: behavioural titration, foraging, Nectarinia osea, parental care, sunbirds. Ecology (2004) 73, Ecological Society Introduction Central place foraging is where an animal brings food to a certain location, such as a cache or to its young in Correspondence: S. Markman, School of Biosciences, Main Building, Park Place, Cardiff University, Cardiff CF10 3TL, UK. Tel (0) ; Fax: + 44 (0) ; markmans@cardiff.ac.uk Present address: Institute of Biology, NTNU, Trondheim, N- 7491, Norway. the nest (Orians & Pearson 1979). When parent birds feed nestlings, their nutritional and energetic requirements are often different from those of their young (Murphy 1996). For example, many frugivorous and granivorous species supplement their nestlings diets with arthropod prey that contains important nutrients, such as proteins, to support their rapid growth (Gill 1990). This phenomenon may lead to different sizes or types of prey being consumed by parents as opposed to their nestlings. Foraging decisions regarding patch use and prey types should therefore be influenced both by

2 748 S. Markman et al. the demands of the young as well as the need of the parents to feed themselves. These effects can best be studied separately when parents and young eat different foods (Ydenberg 1994; Ydenberg et al. 1994). The value of various food patches can be assessed by controlled experiments in which animals are subjected to small changes in their food resources, while any responses in their foraging behaviour are recorded. This was termed behavioural titration by Kotler & Blaustein (1995). In behavioural titration experiments, two types of patches are usually presented to the forager. At the start, the patches differ from each other in either the amount of food they contain, and/or in the character of their placement (e.g. a patch safe from predators vs. a risky patch). The animal decides to spend different amounts of time in and/or to harvest different amounts of food from each type of patch. The experimenter then gradually changes the quality of one of the patches to the point where the forager invests equal time or effort foraging in the two types of patches. Built into this is the assumption that the animal can compare and equalize the costs and benefits of patch exploitation. We took advantage of the fact that parent Palestine sunbirds (Nectarinia osea, Bonaparte) eat both nectar and arthropods, whereas they feed only arthropods to their young (Markman, Pinshow & Wright 1999), and combined this clear division of food types with an element of spatial separation in a behavioural titration experiment on foraging parents. Our goal was to evaluate the value of self-feeding compared to the opportunity to provision young. We addressed a number of novel questions regarding the evolution of sex differences in central place foraging tasks by parent birds. First, does the quality of food consumed by the parents in a given patch affect their use of other patches? Second, how do parents optimize the time they spend foraging in patches used only for self-feeding compared to patches used for both selffeeding and the delivery of food to their young? The behavioural outcomes of these foraging decisions also yield information concerning how parents view their environment and how they rank the quality of patches used for more than one purpose (for non-parent foragers, see: Abramsky, Rosenzweig & Pinshow 1991). Recent provisioning and foraging models assumed that the energy content of food consumed by parent birds will affect their rate of food gathering and delivery to the young (Ydenberg 1994; Ydenberg et al. 1994; McNamara & Houston 1997). Our earlier experiments partially support these models, as we have shown that subtle increases in supplementary sucrose concentration affect levels of parental chick feeding in the sunbirds (Markman, Pinshow & Wright 2002). We further assumed that in order to maximize both their food consumption and delivery of food to their young, parent sunbirds should incorporate constantly updated information regarding the relative profitability and the relative locations of different food patches into their foraging decisions. Therefore, we predicted that an increase in supplementary sucrose concentration in a particular patch will allow parent sunbirds to increase their rate of energy intake, and thus (1) increase the amount of arthropods that they gather to feed on themselves and to deliver to their offspring from another food patch, and (2) continue to exploit a less profitable nectar source for their own use, which is close to a valuable source of arthropod prey for their nestlings. The above predictions can be extended to explore sex-specific differences in patch use by parent sunbirds as a function of the parental investment roles played by each gender. For example, males that patrol their territory may have access to a wider range of information and foraging options, whereas females mainly care for the young in the nest and so might be limited in their foraging choices to patches closer to this central place and/or to nestling food sources. We tested the above predictions by examining how parent sunbirds use two different food patches, which served either for self-feeding or for both self-feeding and delivery of food to the young. Methods EXPERIMENTAL PROCEDURE All pairs of Palestine sunbirds used in this study were feeding their first successful brood of the season. All nests contained broods of three nestlings, the maximum natural brood size in this species (Goldstein & Yom-Tov 1988). On the day that the third egg was laid in each nest, the parents were offered two 100 ml feeders containing a 1 25 mol kg 1 H 2 O sucrose solution, in order to habituate them to using the feeders (see Markman et al. 2002). These feeders were initially placed together, about 5 m in front of the nest. Every day thereafter, each feeder was moved about 10 m in opposite directions in relation to the nest site. We continued moving the two feeders until each one was about 50 m away from the nest. When the first nestling hatched, we filled one feeder (selected randomly) with a 0 25 mol kg 1 H 2 O sucrose solution, and we put a dish of commercially purchased flightless houseflies (Musca domestica, Linnaeus) next to it. On the other side of the territory, we filled the second feeder with a 0 25 mol kg 1 H 2 O sucrose solution, but did not add a fly dish. Observations of these two food patches during the two following days revealed that parent sunbirds spent much more time feeding in the patch with the flies than in the patch with only a sucrose feeder. The experiment itself was conducted as follows: on one side of a parents territory we presented a constant patch that contained a feeder with a sucrose solution of 0 25 mol kg 1 H 2 O that served the parents but could not be fed to their young, and an adjacent dish with flightless houseflies (available to parents both for themselves and for delivery to their young). This patch was named constant because we did not change the

3 749 Patch use by parent birds sucrose concentration throughout the experiment, and always provided flies a few hours before each observation. As the birds visited this patch regularly to feed and probably to sample from it, they encountered the flies before the actual observations started. To ensure that this was the case, we waited until both parents had begun to use the fly dish before starting our observations. We varied the concentration of the sucrose solution (again, available to parents for self-feeding but not available for feeding to young) in the variable patch feeder on the other side of the territory every 2 days. The sucrose concentrations used were 0 25, 0 35, 0 45, 0 55, 0 65 and 0 75 mol kg 1 H 2 O. These concentrations are within the range of concentrations that Palestine sunbirds consume in nature (Nahari 1980). We accounted for evaporation by placing an inaccessible control feeder containing a solution of the same concentration next to each usable one (see Markman et al. 2002). The volumes of both accessible and inaccessible feeders were calculated by weighing them (± 0 01 g) before and after each observation session. All the feeders used in this experiment were filled daily and contained much more food than could be consumed by the parents in a day. We recorded data only on the second day of a given sucrose concentration treatment to ensure that birds had sufficient time to become aware of the treatments involved and adjust their behaviour accordingly. Every other day, we also swapped the positions of the constant and the variable patches relative to the nest. This was to avoid any side or location bias in feeder use by the birds (see Jackson, Nicolson & Lotz 1998). We used an experimental design similar to a Latin square (Zar 1996) in which each pair of sunbirds was exposed to all six levels of sucrose concentrations in a different order. In total, we used 36 pairs of birds, and thus tested 36 unique order combinations of the six treatment concentrations. Each pair was therefore tested with a constant patch and each of the six different sucrose concentrations in the variable patch. Six pairs of parents were tested on any one day, with all the different sucrose concentrations being used across the six different territories, thereby controlling for the effects of brood age, season, or weather conditions on a particular day. DATA COLLECTION The experiment began when the oldest nestling was 3 days old (hatching date = day 0) and lasted until the oldest nestling was 14 days old. Each pair of sunbirds was observed for 1 h per day, between and h, on the second day of presenting each sucrose concentration. The order of observation of the different pairs was changed every day, thereby minimizing biases due to time of day effects. When the food patches were placed on each side of a territory, the two different observers began working together. One watched the variable patch and the other the constant patch. To remove observer bias, pairs of observers switched patches every other day of observation. During the 1-h observation sessions, we recorded the total time that the birds spent feeding in a patch. We further recorded the number of flies taken by each parent and calculated the total amount of nectar taken from each feeder per hour. Due to the lack of a third observer at the nest, it was not possible to assess how many of the flies were consumed by the parents and how many they delivered to their young. However, assuming that, on average, both male and female parents consumed nectar at the same rate (S. Markman, unpublished data), we estimated how much nectar each parent consumed by multiplying the proportion of time that each parent spent feeding at the feeder by the total volume consumed by both parents. STATISTICAL ANALYSIS Data were analysed using repeated-measures ANOVA (Sokal & Rohlf 1997), with treatment (i.e. the six different sucrose concentrations in the variable patch) as the repeated measure, following von Ende (1993). Individual territory, and therefore the order of sucrose concentration treatments, did not have any significant effects on any of the variables related to the time spent at the different feeders, nor on the amounts of food taken from them (repeated-measures ANOVA, all P > 0 05). In addition, there were no significant interaction terms between territory and sucrose treatment in the variable patch. Therefore, the effect of territory or treatment order was left out of all the following analyses of sucrose treatment. We used stepwise multiple regression analysis to test for the linearity of any effect of sucrose concentration in the variable patch (holding P = 0 05 to enter and P = 0 1 to remove). The linear effect of sucrose concentration in the variable feeder was forced into the regression model, and sucrose concentration-squared was stepped in to test for a polynomial term. Sucrose concentration-squared was thus added to the model only if it explained a significant additional portion of the variance in the mean data. If additional terms, such as sucrose concentration-cubed, were suspected to have an effect, they were also stepped into the model. The average shapes of the polynomial curves were determined for each dependent variable using the mean data for all birds together. The chosen shape for regression lines was then fitted for each bird separately in order to obtain individual equation parameters and titration points for further comparison and analysis. The average titration points for male and female parents were calculated by equalizing the value of x (x = sucrose concentration in the variable patch) in two regression equations: (a) time spent feeding in the variable patch vs. the sucrose concentration in the variable patch and (b) time spent feeding in the constant patch vs. the sucrose concentration in the variable patch. This resulted in the sucrose concentration at the point of equality where parents used the constant and the variable patches for the same amount of time. Wherever

4 750 S. Markman et al. possible, this procedure was used to obtain a titration point for each individual parent. Because of the wide variation in the data of some individuals, there were cases where it was impossible to equalize the two regression lines for a given bird. Hence, even if they contributed to the overall average results, some individual parents did not provide separate titration points because they preferred one patch to the other over all six sucrose concentrations. Results AMOUNT OF FOOD TAKEN Both male and female parent Palestine sunbirds significantly increased the amount of sucrose solution that they took from the variable patch as its concentration increased (Table 1a, Fig. 1). Generally, females consumed significantly more solution than males (Table 1a). There was a linear increase in the amount of food taken by both parents from the variable patch (Fig. 1; effect of sucrose treatment: for males: R 2 = 0 88, d.f. = 4, P = 0 006; for females: R 2 = 0 84, d.f. = 4, P = 0 009). For both sexes, the number of flies taken from the constant patch increased linearly as the sucrose concentration in the variable patch increased (Table 1b, Fig. 2, effect of sucrose treatment: for males: R 2 = 0 91, d.f. = 4, P = 0 003; for females: R 2 = 0 95, d.f. = 4, P = 0 001). Females took significantly more flies from the dish than did males (Table 1b, Fig. 2). Fig. 1. The relationship between the amount of sucrose solution (mean ± SE) consumed by Palestine sunbirds and sucrose concentration at a feeder in a variable food patch, where the birds were given a choice of either that patch, or a constant food patch (at the opposite side of their territory) with a feeder containing a constant concentration of 0 25 mol kg 1 H 2 O sucrose solution alongside a dish of flightless houseflies. Males are represented by filled circles and bold regression line (y = 2 15x 0 28), and females by open circles and fine regression line (y = 2 51x 0 39). The amount of food consumed by males and females from the sucrose feeder in the constant patch did not change significantly as sucrose concentration in the variable patch increased (Table 1c, Fig. 3; effect of sucrose treatment: for males: R 2 = 0 17, d.f. = 4, P = 0 402; for females: R 2 = 0 00, d.f. = 4, P = 0 969). Table 1. Repeated-measures ANOVAs of the amounts of food taken per hour by parent Palestine sunbirds from: (a) a feeder containing a sucrose solution of variable concentration; (b) a dish of flightless houseflies; and (c) a feeder containing a constant sucrose concentration. The effects of parental sex (Sex), and sucrose concentration in the variable patch (Treatment) are given, along with interaction terms Source d.f. MS F P (a) Amount taken from a variable sucrose solution feeder Sex Error Treatment < Sex treatment < Error (b) Number of flightless houseflies taken from a dish Sex Error Treatment < Sex teatment Error (c) Amount taken from a constant sucrose solution feeder Sex < Error Treatment Sex treatment Error

5 751 Patch use by parent birds Fig. 2. The relationship between the number of flightless houseflies taken per hour (mean ± SE) by Palestine sunbirds from a constant food patch (containing a dish of flightless houseflies alongside a feeder of constant concentration of 0 25 mol kg 1 H 2 O sucrose solution) and the sucrose concentration in a variable food patch situated at the opposite end of their territory. Males are represented by filled circles (y = 8 40x ), and females by open ones (y = 11 91x ). Fig. 3. The relationship between the amount of sucrose solution (mean ± SE) consumed by Palestine sunbirds from a constant food patch (containing a feeder of constant concentration of 0 25 mol kg 1 H 2 O sucrose solution alongside a dish of flightless houseflies) and the sucrose concentration in a variable food patch situated at the opposite end of their territory. Males are represented by filled circles and females by open ones. Generally, males consumed a significantly greater amount from the constant sucrose feeder than did females (Table 1c, Fig. 3), and there was a significant interaction between parental sex and sucrose treatment. This interaction reflects the difference between males and females in the amount of sucrose consumed at intermediate concentrations (Table 1c, Fig. 3). TIME SPENT FEEDING IN A PATCH Parent birds of both sexes significantly increased their time spent feeding in the variable patch as its sucrose Fig. 4. The relationship of the time (mean ± SE) spent by (a) male and (b) female Palestine sunbirds feeding in one of two food patches: a constant food patch containing a feeder with a constant concentration of 0 25 mol kg 1 H 2 O sucrose solution alongside a dish of flightless houseflies and a variable food patch containing only a feeder with different concentrations of sucrose solution, and the sucrose concentration at the feeder in the variable patch. Filled circles represent time spent feeding in the variable patch. (a) males: y = x ; (b) females: y = x x Open circles represent time spent feeding in the constant patch. (a) males: y = x; (b) females: y = x x ). The arrow indicates the average titration point (i.e. the sucrose concentration at which parents spent equal amounts of time at the two patches). concentration increased (Table 2a, Fig. 4a,b). There was a significant interaction term between sucrose treatment and parental sex (Table 2a, Fig. 4a,b). In males, the increase was linear with increasing sucrose concentration (Fig. 4a; R 2 = 0 88, d.f. = 4, P = 0 006), whereas in females the increase was non-linear (Fig. 4b; effect of sucrose treatment: R 2 = 0 84, d.f. = 4, P = 0 009; effect of sucrose treatment squared: t = 3 29, d.f. = 3, P = 0 046). Parent sunbirds spent significantly more and more time feeding on the flies in the constant patch as the sucrose concentration in the variable patch increased (Table 2b, Fig. 5). Females spent significantly more time feeding on flies than males, and there was a significant interaction term between parental sex and sucrose treatment

6 752 S. Markman et al. Table 2. Repeated-measures ANOVAs of the time spent feeding by parent Palestine sunbirds at: (a) a feeder containing a sucrose solution of variable concentration; (b) a dish of flightless houseflies; (c) a feeder containing a constant sucrose concentration; and (d) a feeder containing a constant sucrose concentration and an associated dish of flightless houseflies. The effects of parental sex (Sex), and sucrose concentration in the variable patch (Treatment) are given, along with interaction terms Source d.f. MS F P (a) Time spent at a variable sucrose solution feeder Sex Error Treatment < Sex treatment < Error (b) Time spent at a dish of flightless houseflies Sex < Error Treatment < Sex treatment < Error (c) Time spent at a constant sucrose solution feeder Sex Error Treatment < Sex treatment Error (d) Combined time spent at a constant sucrose solution feeder and at a dish of flightless houseflies Sex < Error Treatment < Sex treatment < Error Fig. 5. The relationship between time spent (mean ± SE) by Palestine sunbirds feeding on flightless houseflies in a constant food patch, containing a feeder with a constant concentration of 0 25 mol kg 1 H 2 O sucrose solution and a dish of flightless houseflies, and the concentration of sucrose solution in a variable food patch containing only a feeder with different concentrations of sucrose solution, situated at the opposite end of the birds territory. Males are represented by filled circles (y = 8 40x ), and females by open ones ( y = 11 91x ). (Table 2b, Fig. 5). This interaction reflects the greater increases in the time females spent feeding on flies compared to males (Fig. 5, effect of sucrose treatment; for males: R 2 = 0 93, d.f. = 4, P = 0 003; for females: R 2 = 0 84, d.f. = 4, P = 0 010). Males spent significantly more time at the constant sucrose feeder than females, and there was a significant interaction term between sucrose treatment and parental sex (Table 2c, Fig. 6). This interaction was largely because males showed a marginally significant decrease in the time they spent feeding at the constant sucrose feeder with sucrose concentration in the variable feeder (Fig. 6; linear effect of sucrose treatment: R 2 = 0 66, d.f. = 4, P = 0 049), whereas the time females spent at the constant sucrose feeder did not change significantly (Fig. 6; R 2 = 0 35, d.f. = 4, P = 0 213). However, compared to the other treatments, there were also large sex differences in responses to the 0 35 and 0 45 mol kg 1 H 2 O concentrations (Fig. 6). In general, the sucrose concentration in the variable patch had a significant effect on the time the birds spent feeding in the constant patch (Table 2d, Fig. 4a,b). Females spent significantly more time in the constant

7 753 Patch use by parent birds Fig. 6. The relationship between time spent feeding (mean ± SE) by Palestine sunbirds at a feeder containing a constant concentration of 0 25 mol kg 1 H 2 O sucrose solution alongside a dish of flightless houseflies and the concentration of sucrose solution in a variable food patch containing only a feeder with different concentrations of sucrose solution. Males are represented by filled circles ( y = x), and females by open ones. patch than males, even though they spent less of their time feeding from the constant sucrose feeder (and more of their time feeding on flies). There was a significant interaction term between sucrose treatment and parental sex. Males did not significantly change the time they spent feeding in the constant patch with increasing sucrose concentration in the variable patch (Fig. 4a; R 2 = 0 09, d.f. = 4, P = 0 565), while the time spent by females increased significantly in a non-linear fashion (Fig. 4b; effect of sucrose treatment: R 2 = 0 53, d.f. = 4, P = 0 101; effect of sucrose treatment squared: t = 7 80, d.f. = 3, P = 0 004). TITRATION POINTS The average titration point is that at which birds spent an equal amount of time feeding in the variable and in the constant patches, and it is calculated from the relevant pair of regression equations (see Methods). For the 36 males the average titration point was 0 50 mol kg 1 H 2 O (Fig. 4a). The average titration point for the 36 females was mol kg 1 H 2 O (Fig. 4b). Of the 36 males used in the experiment, we failed to determine titration points for six individuals. Five males always preferred the constant patch to the variable patch at any sucrose concentration, and one always preferred the variable patch to the constant patch. Of the 36 females we failed to find titration points for five; three preferred the constant patch to the variable patch, and two preferred the variable patch to the constant patch. Of the birds that did show individual titration points, females had significantly higher titration points than males (t-test, t = 3 70, d.f. = 59, P < 0 001; average titration point for females = 0 55 ± 0 10 mol kg 1 H 2 O; average titration point for males = 0 47 ± 0 07 mol kg 1 H 2 O). Discussion Palestine sunbirds, like rufous hummingbirds (Blem et al. 2000), were sensitive to even small changes in the quality of their nectar source and adjusted their foraging effort accordingly. They increased the time spent in, and amount of sucrose solution taken from, the variable patch as its concentration increased. This ability to differentiate between small changes in food quality and to respond accordingly appears to be extremely important for animals whose locomotion is energetically expensive, such as sunbirds and hummingbirds (Tamm 1989; Tiebout 1991). The more nectar the birds consumed from the more concentrated variable patch, the more time they could afford to feed from the constant patch. Although we could not assess how many of the flies taken from the constant patch were delivered to the young, it is likely that only a small proportion of the flies was required to satisfy the parents protein and other nutrient requirements (Roxburgh & Pinshow 2000). Therefore, parents probably delivered most of the flies that they collected to their young in the nest. This conclusion is supported by previous data from Markman et al. (2002), which shows that increased sucrose concentrations in solitary feeders (from 0 25 mol kg 1 H 2 O to 1 25 mol kg 1 H 2 O) allowed female sunbirds to increase their delivery rate of arthropods from the natural environment to their young by 38%, and males by 30%. The total increases in arthropod delivery rates by parents in our previous work (Markman et al. 2002) were reflected in an increase of about 25% in mean nestling body mass. As the sucrose concentration in the variable patch increased, parents did not decrease the amount they consumed from the constant sucrose feeder. This was probably because they visited the nearby fly dish more often as the sucrose concentration in the variable patch increased and were therefore able to profit from the use of the nearby constant sucrose feeder, thereby negating the need to pay additional travel costs to fly to the variable feeder (Mitchell 1990). In addition, given the alternating patch locations used during the experiment, and the need for parents to keep updating information regarding patch quality, it is also possible that they would profit from sampling the constant sucrose feeder regularly, even though they preferred the higher quality sucrose in the variable patch. Indeed, it has been shown that both great tits (Parus major) and blue tits (P. caeruleus) forage selectively in rich resource patches around their nests, while continually sampling alternative food patches even in distant parts of their territory (Naef-Daenzer 2000). Female sunbirds participate more in chick feeding than males, and this may have led to them to spend more time than males in the constant patch for the purpose of collecting flies for their chicks. Thus, sex differences in the pattern of use of the constant and variable patches, presumably also reflecting differences in natural patch use throughout the territory, were the result of the division of care levels between parents (Markman,

8 754 S. Markman et al. Yom-Tov & Wright 1995, 1996). This suggests that once animals shift from being simply foragers to being parent-foragers, their decision-making will differ according to the nature and intensity of their parentalcare duties. In addition, females fed more than males at the variable patch and showed a greater shift towards the variable patch feeder when it contained higher sucrose concentrations than the constant patch. This indicates that females may have been less concerned with alternative activities elsewhere in the feeding territory than males. These differences between males and females may be explained in part by the fact that males spent more time patrolling their territory and were therefore able to feed in many different patches. In contrast, females spent much more time close to their nest and, consequently, the feeders. Similar sex-specific differences in food patch use have been reported in calliope hummingbirds (Stellula calliope), with the apparently time-limited males preferring to use a feeder that provided a fast intake rate of a low sucrose concentration over a feeder with low intake rate of a high concentration. The females, being more energy-limited, had the opposite response (Tamm 1989). Our assumption that sunbirds compare the foraging rewards across patches led us to predict that there should be a titration point in terms of equal time spent foraging in the two patches. The existence of titration points in the present study supports this view. Previous behavioural titration experiments have shown that nonbreeding foragers are capable of equalizing their foraging efforts across resource patches (e.g. Abrahams & Dill 1989; Nonacs & Dill 1990). The present study demonstrates, for the first time, that this is also true of reproductively active, central-place foragers, such as parent sunbirds. Females had, on average, higher titration points than males, and this was presumably because females spent more time than males collecting flies for provisioning young in the nest. Therefore, the sucrose solution in the variable patch had to be more concentrated for the females in order to balance the additional time they preferred to spend foraging on the flies in the constant patch. A recent model by McNamara & Houston (1997) deals with the provisioning of young based upon the assumption that the forager has one set of foraging options associated with feeding itself ( self-feeding ) and another set of foraging options associated with provisioning the young. For example, in the case of a parent bird feeding its nestlings, the parent may go to one location to look for one prey type to bring back to the young (i.e. the socalled provisioning option ) and may go to another location to look for other items to eat (i.e. the so-called self-feeding option ). A key prediction of this model is that as the net energy gained while self-feeding increases, the parent will increase its rate of food delivery to the young, as shown empirically to be the case in the present study (see also Markman et al. 2002). In addition, it is apparent from our analysis that the foraging options of self-feeding and delivery to nestlings are often combined within the same foraging trip, and might also depend upon more than one alternative option and the spatial arrangement of food patches. To conclude, parent Palestine sunbirds increased their foraging effort spent in the variable patch used for self-feeding only, as its quality increased. As predicted, the increase in the sucrose concentration in the variable patch led parents: (1) to consume greater amounts of sucrose solution in this patch; (2) to increase their use of other food patches containing prey for delivery to offspring (i.e. the flies in the constant patch); and (3) to persistently use some less profitable food sources, found near those sources used for delivery to the young, for their own consumption (i.e. the constant sucrose feeder). Furthermore, the differences between male and female parents, in their use of the food patches, reflected the sex-specific differences in the parental tasks of either delivering arthropods to the nest (mostly by females) or guarding the territory (mostly by males) (e.g. Markman et al. 1995, 1996). We showed that parent sunbirds gradually equalized their foraging effort as the quality of two patches converged, incorporating the profitability of both selffeeding and food collection for young in the nest. The results of this study demonstrate the existence of critical sex-specific trade-offs in parental foraging effort within and between patches of varying profitability and spatial proximity. Acknowledgements We thank Drs Ron Ydenberg, Yoram Yom-Tov, Arnon Lotem and Yael Lubin for commenting on early drafts of this paper. We are grateful to Dr Beat Naef-Daenzer and an anonymous referee for valuable comments on the final version of this paper. We also thank Noam Naveh, Lia Elazar, Boaz Lederman and Etty Markman for help in data collection. This study was supported by a grant from the Mitrani Department of Desert Ecology (MDDE) to S.M. and Israel Science Foundation grant number 3/98 to B.P. It is paper no. 407 of the MDDE. References Abrahams, M.V. & Dill, L.M. (1989) A determination of the energetic equivalence of the risk of predation. Ecology, 70, Abramsky, Z., Rosenzweig, M.L. & Pinshow, B. (1991) The shape of a gerbil isocline measured using principles of optimal habitat selection. Ecology, 72, Blem, C.R., Blem, L.B., Felix, J. & van Gelder, J. (2000) Rufous hummingbird sucrose preference: precision of selection varies with concentration. Condor, 102, Gill, F.B. (1990) Ornithology. W.H. Freeman, New York. Goldstein, H. & Yom-Tov, Y. (1988) The breeding biology of the orange-tufted sunbird. Ardea, 76, Jackson, S., Nicolson, S.W. & Lotz, C.N. (1998) Sugar

9 755 Patch use by parent birds preferences and side bias in cape sugarbirds and lesser double-collared sunbirds. Auk, 115, Kotler, B.P. & Blaustein, L. (1995) Titrating food and safety in a heterogeneous environment: when are the risky and safe patches of equal value? Oikos, 74, Markman, S. Pinshow, B. & Wright, J. (2002) Manipulation of food resources reveals sex-specific trade-offs between parental self-feeding and offspring care. Proceedings of the Royal Society of London, Series B, 269, Markman, S., Pinshow, B. & Wright, J. (1999) Orange-tufted sunbirds do not feed nectar to their chicks. Auk, 116, Markman, S., Yom-Tov, Y. & Wright, J. (1995) Male parental care in the orange-tufted sunbird: behavioural adjustments in provisioning and nest guarding effort. Animal Behaviour, 50, Markman, S., Yom-Tov, Y. & Wright, J. (1996) The effect of male removal on female parental care in the orange-tufted sunbird. Animal Behaviour, 52, McNamara, J.M. & Houston, A.I. (1997) Currencies for foraging based on energetic gain. American Naturalist, 150, Mitchell, W.A. (1990) An optimal control theory of diet selection the effects of resource depletion and exploitative competition. Oikos, 58, Murphy, M.E. (1996) Nutrition and metabolism. Avian Energetics and Nutritional Ecology (ed. C. Carey), pp Chapman & Hall, New York. Naef-Daenzer, B. (2000) Patch time allocation and patch sampling by foraging great and blue tits. Animal Behaviour, 59, Nahari, N. (1980) The Palestine sunbird (Nectarinia osea) feeding behaviour and its relation to its geographical distribution. MSc Thesis, Tel Aviv University, Tel Aviv [in Hebrew]. Nonacs, P. & Dill, L.M. (1990) Mortality risk versus food quality trade offs in a common currency: ant patch preferences. Ecology, 71, Orians, G.H. & Pearson, N.E. (1979) On the theory of central place foraging. Analysis of Ecological Systems (eds D.J. Horn, R. Mitchell & G.R. Stairs), pp Ohio State University Press, Columbus. Roxburgh, L. & Pinshow, B. (2000) Nitrogen requirements of an old world nectarivore, the orange-tufted sunbird (Nectarinia osea). Physiological and Biochemical Zoology, 73, Sokal, R.R. & Rohlf, F.J. (1997) Biometry, 3rd edn. W.H. Freeman, New York. Tamm, S. (1989) Importance of energy costs in central place foraging by hummingbirds. Ecology, 70, Tiebout, H.M. (1991) Daytime energy management by tropical hummingbirds: responses to foraging constraint. Ecology, 72, von Ende, C.N. (1993) Repeated-measures analysis: growth and other time-dependent measures. Design and Analysis of Ecological Experiments (eds S.M. Scheiner & J. Gurevitch), pp Chapman & Hall, New York. Ydenberg, R.C. (1994) The behavioral ecology of provisioning in birds. Ecoscience, 1, Ydenberg, R.C., Welham, C.V.J., Schmid-Hempel, R., Schmid-Hempel, P. & Beauchamp, G. (1994) Time and energy constraints and the relationships between currencies in foraging theory. Behavioral Ecology, 5, Zar, J.H. (1996) Biostatistical Analysis, 3rd edn. Prentice Hall, Upper Saddle River, NJ. Received 20 September 2003; accepted 19 December 2003