Effects of predation danger on migration strategies of sandpipers

We examine the potential selective importance of predation danger on the evolution of migration strategies of arctic‐breeding calidrid sandpipers. Adult calidrids truncate parental care for reasons not obviously related to levels of food abundance on the breeding areas or at migratory stopover sites, suggesting that a different trade‐off occurs between providing additional care and adult survivorship. The southward migrations of adult western sandpipers precede those of migratory peregrine falcons by almost a month. By moving early and quickly, adults remain ahead of migrant falcons all the way to their non‐breeding areas, where they rapidly moult flight feathers. They complete the moult just as falcons arrive in late September–October. By migrating early, they avoid exposure to falcons when they are unusually vulnerable, due to the requirements for fuelling migratory flight and of wing feather moult. Juvenile western sandpipers migrate south just as falcon numbers start to increase, but do not moult flight feathers in their first winter. Pacific dunlin use an alternative strategy of remaining and moulting in Alaska after falcons depart, and migrating to their overwintering sites after migrants have passed. East of the Rocky Mountains, the southbound migration of falcons begins 4–6 weeks later. Southbound semipalmated sandpipers make extended migratory stopovers, but their lengths of stay shorten prior to falcon migration to the sites in September. Predation danger also may affect the evolution of migration routes. Southbound western sandpipers fly directly from Alaska to southern British Columbia, in contrast to the multi‐stage journey northward along the Alaska panhandle. We estimate that a direct flight would be more economical on northward migration, but may be avoided because it would expose sandpipers to higher mass‐dependent predation danger from migratory falcons, which travel north with sandpipers. By contrast, few raptors are present in Alaska during preparation for the southward flight. A temporal and spatial window of safety may also permit semipalmated sandpipers to become extremely vulnerable while preparing for trans‐Atlantic southward flights. Danger management may account for the these previously enigmatic features of calidrid migration strategies, and aspects of those of other birds.     

Experimental evidence that migrants adjust usage at a stopover site to trade off food and danger

Rich habitats, intensive feeding, and large fuel deposits areassumed to improve the capability for long-distance migrationby birds but may also heighten their vulnerability or exposureto predators. Studies of habitat use by migrants have emphasizedthe importance of feeding, and relatively few studies have consideredhow migrants manage the dangers inherent in acquiring and storingfuel. Migrant western sandpipers (Calidris mauri) stop overon coastal mudflats characterized by a strong feeding–dangergradient, with both food and danger decreasing with distancefrom the shoreline. We experimentally manipulated danger byadding obstructive cover and measured sandpiper usage alongthis gradient. We compared sandpiper usage along a transectextending 100 m on either side of the obstruction with thaton matched control transects without obstructions. The droppingdensity accumulated during a low-tide period provided a sensitivemeasure of sandpiper usage. Mean usage on control transectswas 2.3 droppings/m² and was lower by 1.5 droppings/m² (65%)on treatment transects. Usage did not differ between controland treatment transects at the furthest distances from the obstruction,the difference increased with proximity to the obstruction,and was greater by on average 0.9 droppings/m² on the oceanwardside (low food abundance) than on the shoreward side (high foodabundance). All these findings were predicted by danger managementtheory. Our study provides experimental evidence that migrantbirds are sensitive to danger on stopover and has implicationsfor understanding migration strategies.     

Migration of two calidrid sandpiper species on the predator landscape: how stopover time and hence migration speed vary with geographical proximity to danger

The effects of relative fuel load on migration speed and on vulnerability have been investigated, but the effects of seasonal variation in predation danger on the amount of fuel and duration of stopover have not been considered. We analyzed seasonal patterns of stopover residence times for western and semipalmated sandpipers Calidris mauri and C. pusilla on southward migration in relation to the passage of migratory peregrine falcons Falco peregrinus. We predicted that individuals on stopover far in advance of the seasonal arrival of falcons would adjust stopover length and hence relative fuel load to migrate slowly and cautiously. We predicted that individuals on stopover later in the season would increase migratory speed as the arrival of migratory falcons came closer, while individuals on stopover under or behind the passage of falcons would migrate slowly. Adult and juvenile semipalmated and adult western sandpipers migrated prior to seasonal increases in peregrine abundance, and as predicted, the seasonal patterns of their stopover durations are consistent with an increase in the speed of migration as the date of peregrine arrival approached. Juvenile western sandpipers, in contrast, migrating concurrently with falcons, slowed their speed of migration as predator abundance increased. Stopover patterns differ between species due to different relative fuel loads. The results fit predictions made based on a ‘mortality‐minimizing’ migration strategy.     

Foraging speed in staging flocks of semipalmated sandpipers: evidence for scramble competition

Foraging speed is a key determinant of fitness affecting both foraging success and predator attack survival. In a scramble for food, for instance, evolutionary stable strategy models predict that speed should increase with competitor density and decrease when the risk of attack by predators increases. Foraging speed should also decrease in richer food patches where the level of competition is reduced. I tested these predictions in fall staging flocks of semipalmated sandpipers (Calidris pusilla) foraging for an evasive prey. Capture rate of these prey decreased with sandpiper density as the presence of competitors reduced the availability of resources for those behind. Foraging speed was evaluated indirectly by measuring the time needed to cross fixed boundaries on mudflats over 6 years. As predicted, foraging speed increased with sandpiper density and decreased with food density, but, unexpectedly, increased closer to obstructive cover where predation risk was deemed higher. When foraging closer to cover, from where predators launch surprise attacks, the increase in foraging speed may compensate for an increase in false alarms that interrupted foraging. While foraging in denser flocks decreases foraging success, joining such flocks may also increase safety against predators. In semipalmated sandpipers that occupy an intermediate position in the food chain, foraging behavior is influenced simultaneously by the evasive responses of their prey and by the risk of attack from their own predators.     

Indirect effects of lemming cycles on sandpiper dynamics: 50 years of counts from southern Sweden

The bird-lemming hypothesis postulates that breeding success of tundra-nesting geese and waders in Siberia follows the cyclic pattern of lemming populations, as a result of predators switching from lemmings to birds when the lemming population crashes. We present 50 years of data on constant-effort catches of red knot Calidris canutus and curlew sandpiper C. ferruginea at an autumn migratory stopover site (Ottenby) at the Baltic Sea, supplemented with literature data on winter censuses of dark-bellied brent goose Branta b. bernicla and white-fronted goose Anser albifrons in northwestern Europe, and waders in Germany and Southern Africa. Number and proportion of juveniles in these bird populations (both our own and literature data) were compared with an index of predation pressure (calculated from the abundance of lemmings on the Taimyr peninsula), and climate indices for the North Eurasia and the North Atlantic regions. The index of predation pressure correlated significantly with the number of juveniles of red knot and curlew sandpiper, but not with number of adults. Also, this index correlated with the reproductive performance of geese and waders reported in the literature. Fourier analysis revealed a significant deviation from random noise with the maximum spectral density at the period length of 3 years for number of juvenile red knots and curlew sandpipers captured at Ottenby, abundance of lemmings, reproduction in arctic fox Alopex lagopus, and reproductive performance in geese on the Siberian tundra. Also, the date of passage at Ottenby for adult red knot and curlew sandpiper showed a spectral density peak at a period length of 3 years, the latter species also showing a peak at a period length of 5-6 years. Passage dates for adult red knot and curlew sandpiper were earlier in years of high predation pressure compared with years of low predation pressure. The fluctuations in reproductive success of the studied Siberian goose and wader species appear to be primarily influenced by biotic factors in the breeding area, rather than by abiotic factors, such as climate oscillations. Annual variations in migratory arctic bird populations may have far reaching effects in habitats along their migration routes and in their wintering areas. We suggest a link between lemming cyclicity in the Northern Hemisphere and predation pressure on Southern Hemisphere benthos, in which the signal is carried between continents by long distance migrating waders.     

Diel variation in predator abundance, predation risk and prey distribution in shallow-water estuarine habitats

Predation by visual predators is often affected by light conditions and may therefore exhibit strong diel variation. The dominant predators on grass shrimp, Palaemonetes pugio, are finfish predators that are thought to locate their prey by visual cues. We examined the response of grass shrimp to diel variation in predation risk in the nearshore shallow waters of the Chesapeake Bay. We used diel shoreline seines to assess the relative abundance of predators. We assessed the relative risk of predation with shrimp tethered at refuge (30 cm) and nonrefuge (60 cm) depths. To measure grass shrimp response to predation risk, we used dipnets to monitor habitat use. Four predominantly visual predators dominated the shoreline seine catches, Fundulus heteroclitus, Micropogonias undulatus, Morone americana and Morone saxatilis. Total predator abundance had a diel component, with dramatic nighttime decreases in total abundance, whereas guild composition and relative abundance remained unchanged. Relative predation risk for tethered shrimp exhibited significant time by habitat interaction. During the day, depth negatively affected survivorship of tethered shrimp while at night overall survivorship increased and there was no effect of depth. Shrimp habitats use reflected diel predation risks. Abundances in the near shore were highest during the day with decreased abundances at night. Together, the seine and tethering data highlight the importance for a refuge (e.g., shallow water) from predation during the daytime and a relaxation of predation pressure at night.     

Non‐breeding areas and timing of migration in relation to weather of Scottish‐breeding common sandpipers Actitis hypoleucos

The number of breeding common sandpipers has declined in Britain due to poorer return rates from non‐breeding areas. To investigate little known aspects of their annual cycle, breeding common sandpipers were fitted with geolocators to track their migrations and determine their non‐breeding areas. Ten tagged birds left Scotland on 9 July (median dates and durations are given throughout the abstract). Short‐term staging was carried out by some birds in England and Ireland, then for longer by most birds in Iberia before continuing to West Africa, arriving on 28 July. Six birds spent most of the non‐breeding season (October–February) on the coast of Guinea‐Bissau, suggesting that this is a key area. Single birds occurred in Sierra Leone, Guinea, the Canary Islands and western Sahara. The southward migration from Scotland took 17.5 d (range 1.5–24 d), excluding the initial fuelling period. The first northward movement from Africa was on 12 April. Staging occurred in either Morocco, Iberia or France. Arrival in Scotland was on 2 May. The northward migration took 16 d (range 13.5–20.5 d). The main migration strategy involved short‐ and medium‐range flights, using tail‐winds in most cases. Variation in strategy was associated with departure date; birds that left later staged for shorter durations. Coastal West Africa provides two major habitats for common sandpipers: mudflats associated with mangroves and rice fields. Although the area of mangrove has been depleted, the scale of loss has probably been insufficient to account for the decline in sandpiper numbers. Rice fields are expanding, providing feeding areas for water‐birds. Meteorological data during the migrations suggest that the weather during the southward migration is unlikely to contribute to a population decline but strong cross‐winds or head‐winds during the northward migration to the breeding grounds may do so.     

Western sandpipers have altered migration tactics as peregrine falcon populations have recovered

The presence of top predators can affect prey behaviour, morphology and life history, and thereby can produce indirect population consequences greater and further reaching than direct depredation would have alone. Raptor species in the Americas are recovering since restrictions on the use of dichlorodiphenyltrichloroethane (DDT) and the implementation of conservation measures, in effect constituting a hemisphere-wide predator-reintroduction experiment, and profound effects on populations of their prey are to be expected. Here, we document changes in the behaviour of western sandpipers (Calidris mauri) at migratory stopover sites over two decades. Since 1985, migratory body mass and stopover durations of western sandpipers have fallen steadily at some stopovers in the Strait of Georgia, British Columbia. Comparisons between years, sites and seasons strongly implicate increasing danger from the recovery of peregrine falcons (Falco peregrinus) as a causal factor. A decade-long ongoing steep decline in sandpiper numbers censused on our study site is explained entirely by the shortening stopover duration, rather than fewer individuals using the site. Such behavioural changes are probably general among migratory shorebird species, and may be contributing to the widespread census declines reported in North America.     

Life history varies with migratory distance in western sandpipers Calidris mauri

The propensity of migratory waders to remain on the non-breeding grounds during the arctic breeding season ("oversummer") in their first biological year of life ("juveniles") may be latitude, and thus migratory distance dependent. We compared the extent of preparation for northward migration of western sandpipers Calidris mauri spending the non-breeding season in México and Panamá during 1995–1998. During winter residency and premigratory periods, we measured body mass and scored the extent of dull basic versus bright alternate breeding plumage of captured juveniles and adults (second biological year or older), and obtained additional plumage scores from observations of uniquely colour banded birds. Nearly all western sandpipers in México prepared for northward migration by increasing body mass and moulting into breeding plumage. In Panamá, most adults prepared for migration, but few, if any, juveniles did so. Patterns of body mass and breeding plumage development do not generally support the hypothesis that oversummering by juveniles results directly from less efficient foraging or from resource competition with adults. We suggest instead that costs directly associated with migratory distance per se influence the life history strategies of sandpipers spending the non-breeding seasons at different latitudes. This latitudinal difference should interact with the well documented sex-ratio cline in non-breeding distribution (male western sandpipers predominating in northern parts of the range and females in southern parts). This suggests that females have more conservative life histories, prioritizing first year survivorship, relative to males that instead weight first-year breeding opportunities.     

Danger management and the seasonal adjustment of migratory speed by sandpipers

The behavior of migrating birds is governed by time‐, energy‐ and danger‐minimizing strategies. The adjustment of migration speed (i.e. the rate at which distance is covered during a migration) is a behavioral tactic that might contribute to these strategic goals. Shorter stopovers and greater fuel loads increase migration speed, but both require more intensive foraging at stopovers, making migrants more vulnerable to predators. A simple numerical model shows how seasonal alterations in migration speed can lower the exposure of western sandpipers to peregrine falcons, their most important predator. The ‘caution–speed–caution’ pattern of higher migration speed in the mid‐passage period, observed in earlier work, requires that the intensive foraging necessary heightens vulnerability, and that migrants are exposed to both migrant predators as well as predators resident at migratory stopovers.     

Shorebird predation may cause discrete generations in an amphipod prey

We examine the possible impact of intense, periodic predation by the semipalmated sandpiper Calidris pusilla, on the life history patterns of its amphipod prey, Co-rophium volutator We compare populations from two mudflats, one of which is visited by shorebirds on their annual migration south, and one which physically appears very similar but is not visited by the birds The Corophium population exposed to intense predation had two distinct peaks in density within the season, corresponding to two generations, and the two cohorts had constrained size distributions, and relatively synchronized timing of reproduction On the mudflat not visited by sandpipers, densities increased m spring and then remained constant through summer Reproduction was continuous The mid-summer decline in amphipod density on the mudflat used by sandpipers could not be attributed directly to sandpiper predation as had been previously argued We interpret the decline as the result of a synchronized natural die-off after reproduction Selective predation on large amphipods may contribute to the second peak in density by increasing juvenile survivorship due to the removal of competing adults Size-selective predation by sandpipers causes the overwintering cohort to have a restricted size range m the autumn, and this synchrony persists through to the reproductive period of the following spring We do not exclude the possibility that these differences in life history have been influenced in an evolutionary sense by the long history of intense periodic predation     

Competition in foraging flocks of migrating semipalmated sandpipers

I examined the effect of competitor density on foraging success in staging semipalmated sandpipers (Calidris pusilla) foraging on a burrowing amphipod (Corophium volutator) in each of two study years. Little is known about the effect of competitor density when predation attempts disturb prey, causing a temporary decrease in food availability. Controlling for Corophium density and other potentially confounding factors such as temperature, pecking rate and capture rate increased linearly with sandpiper density. Success rate, the ratio of captures to pecks, was not influenced by sandpiper density. The effect of sandpiper density was similar in each of the two study years and was documented early and late in the low tide period. The increase in foraging rate is argued to be a response to increased competition for rapidly depleting prey at the temporal scale of exploitation by a flock. Potential fitness costs associated with higher foraging costs may include decreased ability to distinguish between the profitability of different prey and reduced vigilance against predators.     

Control mechanisms of diel vertical migration: theoretical assumptions

We explore control mechanisms underlying the vertical migration of zooplankton in the water column under the predator-avoidance hypothesis. Two groups of assumptions in which the organisms are assumed to migrate vertically in order to minimize realized or effective predation pressure (type-I) and to minimize changes in realized or effective predation pressure (type-II), respectively, are investigated. Realized predation pressure is defined as the product of light intensity and relative predation abundance and the part of realized predation pressure that really affects organisms is termed as effective predation pressure. Although both types of assumptions can lead to the migration of zooplankton to avoid the mortality from predators, only the mechanisms based on type-II assumptions permit zooplankton to undergo a normal diel vertical migration (morning descent and evening ascent). The assumption of minimizing changes in realized predation pressure is based on consideration of DVM induction only by light intensity and predators. The assumption of minimizing changes in effective predation pressure takes into account, apart from light and predators also the effects of food and temperature. The latter assumption results in the same expression of migration velocity as the former one when both food and temperature are constant over water depth. A significant characteristic of the two type-II assumptions is that the relative change in light intensity plays a primary role in determining the migration velocity. The photoresponse is modified by other environmental variables: predation pressure, food and temperature. Both light and predation pressure are necessary for organisms to undertake DVM. We analyse the effect of each single variable. The modification of the phototaxis of migratory organisms depends on the vertical distribution of these variables.     

Body size shapes inter‐specific migratory behaviour: evidence from individual tracks of long‐distance migratory shorebirds

Migration is a common phenomenon across many animal taxa. Understanding how migration scales with body size across species is fundamental in the development of migration theory and in making size‐related predictions. Although aerodynamic theory and ecophysiological scaling laws have assisted greatly in generating such predictions, their verifications have been limited by a lack of empirical data across a range of body sizes. The recent development of ultra‐light tracking devices and its rapid application to migration now allows us to put theory to the test. We used tracking data of seven closely related migratory sandpiper species (family Scolopacidae) along the East Asian‐Australasian Flyway to compare their migratory behaviour when migrating towards the breeding grounds as a function of size (50–750 g). We found a marked decline in migration speed (migration distance divided by total migration duration, including time at stopover sites and in flight) with size. Departure date from the non‐breeding (i.e. wintering) ground and arrival date at the breeding ground also scaled negatively with size. Total migration duration, migration distance, total staging duration (the number of days staying at stopover sites plus days preparing, i.e. fuelling, prior to initial migration) and step length (distance covered within one migratory leg) were not significantly related with size. Correction for phylogeny showed consistent results for all variables. Besides improving our fundamental understanding of inter‐specific variation in migration behaviours, the finding of a clear scaling with size in migration speed and migration timing highlight differential size related capabilities and constraints of migrants. Migratory birds, including sandpipers, are declining on a global scale and particularly along the East Asian‐Australasian Flyway. This notion of size‐dependency in migratory traits may have a bearing on their vulnerability to specific environmental disturbances along their flyways.     

Relationship between Distance to Cover,Vigilance and Group Size in Staging Flocks of Semipalmated Sandpipers

Cover can be either a source of protection or a source of danger for foragers. Distance to cover creates a gradient in predation risk that allows examining adjustments in anti‐predator behaviour such as group size and vigilance. As distance to obstructive cover increases, both group size and vigilance are expected to decrease given that individuals have more time to react to a more distant source of danger. I provide an empirical test of these predictions in staging semipalmated sandpipers (Calidris pusilla) in a system controlling for many confounding factors that have marred earlier research. Controlling for food density, forager density and phenotypic attributes often correlated with distance to cover, I found that as distance to obstructive cover increased, sandpipers foraged in sparser groups, were less flighty and to some extent less vigilant. Such controlled studies are needed to re‐assess the relationship between distance to cover and anti‐predator behaviour.     

Evidence that the parasitic nematode Skrjabinoclava manipulates host Corophium behavior to increase transmission to the sandpiper, Calidris pusilla

We found evidence that a nematode (Skrjabinoclava morrisoni) adaptivelymanipulates the behavior of its intermediate host (the amphipod Corophiumvolutator) to increase its likelihood of transmission to itsfinal host (the semipalmated sandpiper, Calidris pusilla). We foundthat male and female amphipods parasitized by nematodes increasedtheir surface activity in the field during daytime, but notduring nighttime hours. Increased surface activity is knownto increase susceptibility of amphipods to predation by sandpipersduring the day, but not at night, when sandpipers do not feedvisually. Also, as predicted by the manipulation hypothesis,only late-stage (infective) larvae of nematodes were associatedwith behavioral changes of amphipods. We found no evidence thatparasites were associated with other amphipod behaviors in thelaboratory, such as trail complexity, distance traveled, orburrow-probing activity of crawling males as would be expectedif parasitized hosts altered their own behavior. Survivorshipof amphipods was also unaffected by parasitism, which may favorparasite transmission. Thus, behavioral changes of parasitizedhosts were simple, and their expression was context-dependentand related to likelihood of predation. We argue that maturationtimes of nematodes in relation to migration schedules of sandpipers providea narrow window of opportunity and may explain why nematodes manipulateamphipod behavior.     

Predation Danger Can Explain Changes in Timing of Migration: The Case of the Barnacle Goose

Understanding stopover decisions of long-distance migratory birds is crucial for conservation and management of these species along their migratory flyway. Recently, an increasing number of Barnacle geese breeding in the Russian Arctic have delayed their departure from their wintering site in the Netherlands by approximately one month and have reduced their staging duration at stopover sites in the Baltic accordingly. Consequently, this extended stay increases agricultural damage in the Netherlands. Using a dynamic state variable approach we explored three hypotheses about the underlying causes of these changes in migratory behavior, possibly related to changes in (i) onset of spring, (ii) potential intake rates and (iii) predation danger at wintering and stopover sites. Our simulations showed that the observed advance in onset of spring contradicts the observed delay of departure, whereas both increased predation danger and decreased intake rates in the Baltic can explain the delay. Decreased intake rates are expected as a result of increased competition for food in the growing Barnacle goose population. However, the effect of predation danger in the model was particularly strong, and we hypothesize that Barnacle geese avoid Baltic stopover sites as a response to the rapidly increasing number of avian predators in the area. Therefore, danger should be considered as an important factor influencing Barnacle goose migratory behavior, and receive more attention in empirical studies.     

Trade-offs, condition dependence and stopover site selection by migrating sandpipers

Western sandpipers Calidris mauri on southward migration fly over the Gulf of Alaska to the Strait of Georgia, British Columbia, where they stop for a few days to replenish reserves before continuing. In the Strait, individuals captured on the extensive tidal mudflats of the Fraser estuary (∼25000 ha) are significantly heavier (2.71 g, or >10% of lean body mass) than those captured on the small (<100 ha) mudflat of nearby Sidney Island. Previous work has shown that the difference cannot be attributed to seasonal timing, size, age or gender effects, and here we compare predictions made by six hypotheses about a diverse set of data to explain why, partway through a migratory journey of ∼10000 km, birds have such different body masses at two stopover sites within 40 km of each other. The 'trade-off' hypothesis – that the large Fraser estuary offers safety from predators, but a lower fattening rate, while the small Sidney Island site is more dangerous, but offers a higher fattening rate – made six successful predictions, all of which were upheld by the data. All other hypotheses failed at least one prediction. We infer that calidrid sandpipers arriving in the Strait of Georgia with little fat remaining (and therefore low body mass) choose to take advantage of the high feeding rate at small sites like Sidney Island because they are less vulnerable to avian predators than are individuals with higher fat reserves, who instead elect to feed at large open sites like the Fraser estuary mudflats.     

Adverse wind conditions during northward Sahara crossings increase the in‐flight mortality of Black‐tailed Godwits

Long‐distance migratory flights are predicted to be associated with higher mortality rates when individuals encounter adverse weather conditions. However, directly connecting environmental conditions experienced in‐flight with the survival of migrants has proven difficult. We studied how the in‐flight mortality of 53 satellite‐tagged Black‐tailed Godwits (Limosa limosa limosa) during 132 crossings of the Sahara Desert, a major geographical barrier along their migration route between The Netherlands and sub‐Saharan Africa, is correlated with the experienced wind conditions and departure date during both southward and northward migration. We show that godwits experienced higher wind assistance during southward crossings, which seems to reflect local prevailing trade winds. Critically, we found that fatal northward crossings (15 deaths during 61 crossings) were associated with adverse wind conditions. Wind conditions during migration can thus directly influence vital rates. Changing wind conditions associated with global change may thus profoundly influence the costs of long‐distance migration in the future.     

Do sex and habitat differences in antipredator behavior of Western Sandpipers Calidris mauri reflect cumulative or compensatory processes?

Individuals manage their risk of predation in different ways in different situations. We studied the use of anti-predator behavior by Western Sandpipers (Calidris mauri) at Bahía Santa María, northwestern Mexico, foraging in three habitats that differed in presumed predation danger. Brackish flats are completely open, making them theoretically less dangerous for feeding sandpipers than mangroves and cattail marshes, which have closer visual horizons. Western Sandpipers are sexually dimorphic, with females about 15% longer-billed and 10% heavier than males. We previously showed that male and female sandpipers differed in their habitat choice and relative body mass in ways consistent with differential responses to predation danger (Fernández and Lank in Condor 108:547–557, 2006). Contrary to expectations, however, females were overrepresented in more dangerous habitats. Here, we examine differential usage across habitats and between the sexes of three anti-predator tactics—flock size, density within flocks, and vigilance rate—that may be used cumulatively to reinforce safety, or as trade-offs that compensate for levels of usage of each. We hypothesized, and found, that ordered differences occur among habitats, and that controlling for other factors, females were more cautious than males. For the most part, the use of these three tactics appeared to be cumulative, rather than compensatory. However, with respect to habitat use, birds appeared to compensate for the higher probability of mortality intrinsic to the use of higher-danger habitats by increasing the use of vigilance, foraging in tighter flocks, and maintaining lighter body weights (females only). Thus, both cumulative and compensatory processes operate among anti-predator tactics to determine the net level of safety and trade-off against other factors.