Journal of Ethology

, Volume 23, Issue 2, pp 77–83 | Cite as

Mechanisms of cache retrieval in long-term hoarding birds

Review

Abstract

Food hoarding and memory have primarily been studied in two bird families, the Corvidae (crows, jays, nutcrackers, etc.) and the Paridae (tits, titmice and chickadees). In both families there are species that hoard large quantities of seeds and nuts in the autumn and depend on these stores during the winter. Caches are concealed or highly inconspicuous and the most efficient way to retrieve them is to remember the exact locations. However, a long-term memory for a large number of caches may be physiologically expensive, and especially after long retention intervals, an alternative strategy could be to retrieve caches by cheaper but less efficient methods. Very few studies have been designed to investigate the decay of the memory in birds, but both field observations and experiments point in the same direction: although long-term hoarding corvids seem to possess an accurate long-term memory, long-term hoarding parids do not appear to. I discuss possible reasons for this and suggest that differences between the families in their degree of dependence on stored food or/and size-related limitations of brain capacity may be important.

Keywords

Food hoarding Paridae Corvidae Brain size Cache retrieval 

Introduction

Food storing is a way for animals to save food from periods when it is abundant for periods when it is scarce (e.g. Vander Wall 1990; Brodin 1994c). It has been shown in several field studies that hoarding birds frequently retrieve caches within hours up to a day after storing (Cowie et al. 1981; James and Verbeek 1984; Stevens and Krebs 1986; Brodin 1992). Henceforth I refer to storing with cache recovery the same day or within a few days as short-term hoarding.

However, many bird species store food over long intervals, for example when seeds and nuts are stored in the autumn to be used as winter food (for references older than 1990 see Vander Wall 1990, pp 15–20, 297–305; Marzluff and Balda 1992; Balda and Kamil 1992; Brodin 1994b). The interval between storing and retrieval can then range from weeks up to half a year or even longer. Henceforth I refer to hoarding with such long retention intervals as long-term hoarding.

For most long-term-hoarding bird species, stored food constitutes a large proportion of the winter sustenance (see references in Vander Wall 1990, pp. 15–20, 297–305; Brodin 1994b). The amount of food that can be stored by an individual is amazing (Table 1). The time aspect and the large number of caches demand highly developed retrieval capabilities of long-term hoarders since they must be able both to find large numbers of caches and to do this after long retention intervals.
Table 1

The amount of food stored per individual in one autumn by some long-term hoarding bird species

Species

Type of food

Number of nuts/seeds

Number of caches

Source

Clark’s nutcracker

Pine seeds

33,900–98,000

7,700

Tomback (1982), Hutchins and Lanner (1982)

Eurasian nutcracker

Hazelnuts

172,000

9,500

Swanberg (1951)

Marsh tit

Spruce seeds

89,000

89,000

Haftorn (1959)

Coal tit

Spruce seeds

77,000

77,000

Haftorn (1959)

Willow tit

Spruce seeds

62,000

62,000

Haftorn (1959)

Willow tit

Conifer seeds, juniper seeds, hemp nettle

46,000

46,000

Brodin (1994b)

Willow tit

Spruce seeds

150,000

150,000

Pravosudov (1985)

Siberian tit

Spruce seeds

170,000

170,000

Pravosudov (1985)

Most food-hoarding bird species store in cryptic, widely scattered locations. This makes caches difficult for pilferers to find but also difficult for the hoarder to find. Since the number of potential caching locations that are available in nature is almost infinite (e.g. Sherry 1984), the hoarder must use some mechanism to facilitate relocation of caching sites. A very effective way to find caches is to remember their exact locations. Other retrieval mechanisms, for example visual or olfactory cues, will probably not pinpoint locations as exactly as memory (Vander Wall 1982). Besides ensuring accurate relocation of caches, retrieval by memory incurs a smaller risk of losing stored food to pilferers compared to search rules of more general types that can be used by many individuals. The reason is that the knowledge is exclusive to the hoarding individual (e.g. Vander Wall 1982; Sherry 1984; Emery et al. 2004).

However, a long-term memory may not only have benefits. Storing a large number of such memories may be a costly process (Dukas 1999). For example, memories may need to be transferred from temporary to long-term storage systems, and neurons and synapses may need to be repaired and maintanied. This may depend on processes that are physiologically costly and that may require special adaptations in the brain (e.g. Krebs et al. 1989; Sherry et al. 1989; Barnea and Nottebohm 1994; Hoshooley and Sherry 2004). Also back-up neurons may be required in order to create redundancy in the form of alternative memory pathways (Dukas 1999). Finally there may be “lost opportunity costs” since brain capacity that otherwise could be used for other purposes may be occupied by memory storage.

The cheapest retrieval strategy from this perspective is a random search of the area where caches have been stored. However, considering the large number of potential caching locations, a true random search would be highly inefficient. Instead various forms of systematic search that have been labelled “restricted random search” (Haftorn 1974; Vander Wall 1990) have been reported. These strategies rely on some rule that reduces the number of potential caching locations to a smaller subset, for example: (1) preferences for certain types of hoarding sites (Moreno et al. 1981; Alatalo and Carlson 1987; Suhonen and Alatalo 1991; Brodin 1994a; Lens et al. 1994; Brotons 2000), (2) preferences for certain types of hoarding substrates (Haftorn 1956b; Brotons 2000), (3) individually exclusive hoarding areas (Andersson and Krebs 1978; Moreno et al. 1981; Brodin 1994a) or, (4) mnemonic rules (Barnea and Nottebohm 1995). Strategies of this type are less costly from a physiological perspective but incur other costs instead. For example retrieval will be less accurate and there is higher a risk of cache loss since scroungers may learn to use the strategy. Since memory decays over time the relative gain from a restricted random search compared to memory will probably increase over time.

Whereas there are numerous studies of cache retrieval after short retention intervals, only four experiments have investigated whether memory is used for retrieval after a month or more (Hitchcock and Sherry 1990; Balda and Kamil 1992; Brodin and Kunz 1997; Bednekoff et al. 1997). In addition to these controlled experiments, there are also some reports from field workers that suggest whether or not memory is used after long retention intervals (e.g. Tomback 1980; Vander Wall 1990). My aim with this paper is to summarise what we know about memory usage for cache retrieval in long-term hoarding birds. I will also discuss how evolutionary and ecological factors might act as constraints on this memory. I will only consider temporal aspects and omit questions of memory capacity since there are hardly any data on this.

Field observations

In a pioneering field study Swanberg (1951) observed how thick-billed nutcrackers (Nucifraga c. caryocatactes) retrieved hazelnuts of Corylus avellana in a way that he judged would be impossible unless the exact locations were remembered. In his study area, hazelnuts are available on the trees only in August and September. At that time the nutcrackers bury thousands of nuts in scattered locations under moss on the forest floor (Table 1). In winter they are highly dependent on these nuts, which they excavate even through deep snow cover. The habit of nutcrackers to crack the nutshells at the excavation holes made it possible for Swanberg to calculate the percentage of successful excavations. He found it to be around 86% over the whole winter. Since some nuts were likely to have been eaten by rodents, to have rotted or to have been carried away by the hoarder before opening, Swanberg suggested that the accuracy must have been almost 100%. Mattes (1982) reports almost the same proportion of successful excavations in snow in nutcrackers in Switzerland.

The close relative, Clark’s nutcracker, N. columbiana, stores large pine seeds in the autumn and depends on these supplies during the winter. Using a similar technique to Swanberg’s, Tomback (1980) found that retrieval had been successful at 72% of the excavation holes. She estimated loss of caches before retrieval to be 20–30% and hence judged that this species must also have an accuracy of almost 100%.

Although evidence of this type is not as strong as that obtained from controlled experiments, it suggests that nutcrackers possess an exact memory for caching locations that lasts long enough to retrieve caches in winter that were stored in the previous autumn. Nutcrackers belong to the family Corvidae, and retrieval behaviour that suggests a long-term memory of this type has also been reported in some other species in the family, for example the pinyon jay Gymnorhinus cyanocephalus (Marzluff and Balda 1992) and the Eurasian jay Garrulus glandarius (e.g. Goodwin 1951).

Besides corvids, food hoarding and memory have been studied mainly in the Paridae: tits, titmice and chickadees. Among parids there are also long-term hoarders that rely on food stored in the autumn for winter sustenance (Haftorn 1956a; Jansson 1982; Brodin 1994b). Many field workers who have observed parids foraging in winter have got the impression that parids, in contrast to corvids, do not know beforehand the positions of cached food that they find (Haftorn 1956b, 1974; Gibb 1960; Pravosudov 1986; Brodin and Ekman 1994; Brodin 1994a). They then show a behaviour that appears to be a good example of restricted random search, pecking at many potential caching locations as they move short distances in the trees (Haftorn 1956b, 1974). Individual birds then forage in the parts of the trees where they previously have stored food (Pravosudov 1986; Brodin 1994a), which increases the chance that the hoarders will relocate their own caches even if the specific locations are not remembered.

These observations come from studies performed under conditions with no food supplementation. In field studies where food was provided at feeders, parids show a retrieval behaviour that looks strikingly different from the one described above. Retrieving birds typically fly straight to a caching location and uncover a concealed seed (Löhrl 1950; A. Brodin, unpublished data), a behaviour suggesting that the location was remembered. In feeder experiments, stored items were retrieved within 1 or 2 days after storing (Cowie et al. 1981; Stevens and Krebs 1986; Brodin 1992). The quick recovery may be a reaction to high pilferage. Feeders attract granivores, and food stored in the vicinity may be lost if is not recovered soon after storing (Cowie et al. 1981; Stevens and Krebs 1986; Brodin 1992, 1993). The ability of parids to recover caches accurately shortly after storing suggests that the difference between the families concerns the decay of memory rather than the ability to memorise caching sites. This is supported by the accurate caching memory parids have shown after short retention intervals in aviary experiments (e.g. Sherry et al. 1981; Shettleworth and Krebs 1982, 1986).

Experiments

The impression from field observations that nutcrackers possess an accurate long-term memory has been confirmed experimentally. In an experimental room, Clark’s nutcrackers were allowed to store and then retrieve after retention intervals from 11 up to 285 days (Balda and Kamil 1992). The birds remembered the positions of caches after the longest retention intervals, although some forgetting occurred between days 183 and 285 (Fig. 1).
Fig. 1

Longest interval before retrieval (L) and longest interval before retrieval based on memory (M) according to field observations (a) and recorded in laboratory studies (b) in corvids (crossed bars) and parids (hatched bars). 1 Willow tit (calculated from Haftorn 1956a), 2 marsh tit (Löhrl 1950), 3 Eurasian nutcracker (Swanberg 1951), 4 black-capped chickadee (Hitchcock and Sherry 1990), 5 Clark’s nutcracker (Balda and Kamil 1992)

The other impression from field studies, that memory decays relatively fast in parids, is also supported by experimental data. Black-capped chickadees, Poecile atricapillus, retrieved better than random chance up to 28 days after storing but not after that (Hitchcock and Sherry 1990). In a similar experiment Brodin and Kunz (1997) found an almost identical memory decay curve in willow tits (Fig. 2). However, it should be noted that in a third experiment, Healy and Suhonen (1996) found no decay in retrieval performance between 1 and 17 days after storing in willow and marsh tits Parus palustris.
Fig. 2

Memory longevity in black-capped chickadees [Hitchcock and Sherry 1990 (H&S), filled circles] and willow tits [Brodin and Kunz 1997 (B&K), open circles]. To make the studies comparable, the proportion recovered was standardised by setting it to 1 on day 1. Data from H&S are the means of their two experiments. Standardised proportion recovered in control sessions (dotted line) was 0.37 in both studies even though procedures were different: H&S allowed the birds to search for seeds stored by the experimentors while B&K allowed the birds to search for seeds stored by conspecifics. For H&S, data are averaged from pretrials and experiments. The linear regression (solid line) is calculated for data from days 1 to 28 since H&S suggested that there was a memory effect up to 28 days after storing

Not only does the longevity of the caching memory seem to differ between the families, but so does the ability to learn observationally. Black-capped chickadees could not relocate caches 6 minutes after observing conspecifics store them (Baker et al. 1988); in contrast, pinyon jays, Mexican jays (Aphelocoma ultramarina) and Clark’s nutcrackers remembered positions of caches stored by other individuals 1 day after storing (Balda and Bednekoff 1996). A study of movement patterns during cache recovery suggests that Clark’s nutcrackers rely on site-specific memory (Kamil et al. 1999); there are no such observations in parids.

Evolution of memory longevity

Much evidence points to the fact that while long-term hoarding corvids seem to possess an accurate long-term memory, long-term hoarding parids do not seem to. There are several possible explanations for this difference, and I will discuss some of them here. These explanations are not mutually exclusive, two or more could work in concert.

Long-term hoarding may be more important for corvids than for parids

If long-term hoarding is important there will be selection for special adaptations for this. While stored food may constitute almost 100% of the winter diet for nutcrackers and pinyon jays (Giuntoli and Mewaldt 1978; Ligon 1978) it constitutes from 50% (Brodin 1994b) to 70% (Nakamura and Wako 1988) for a long-term hoarding parid such as the willow tit. Also nutcrackers and jays show their dependence on stored food in their well-known eruptive migrations from areas where storable food is scant (Vander Wall and Balda 1981; Vander Wall 1990, pp. 65–66). Finally nutcrackers are able to breed early, even in late winter, since they can feed nestlings stored food (Swanberg 1951; Vander Wall and Balda 1981). Provisioning of stored food to nestlings has been reported in one parid species, the varied tit P. varius, but it is not clear if there is any association with early breeding in this species (Higuchi 1977). It is thus possible that some corvid species have experienced a higher selection pressure for accurate cache retrieval than parids.

The body size difference between the families might affect hoarding strategies or memory capacity for allometric reasons

While corvids are among the largest passerines, parids are among the smaller species in the order. Small birds metabolise relatively more fuel than large birds and carry relatively smaller fat reserves. Thus small birds must eat more frequently than large birds, which may have consequences for foraging and hoarding strategies. During the main hoarding period, the autumn temperatures are relatively mild and food is abundant (e.g. Swanberg 1951; Haftorn 1956b). At this time there should be little effect of body mass in how much time birds can allocate to hoarding.

In winter, on the other hand, smaller birds should be more energetically stressed than larger birds. This may make proximity and fast accessability of caches important for parids, while corvids may be able to disperse caches over a wider area. Obviously the spatial distribution of caches is important for how they are retrieved. While nutcrackers, jays and rooks (Corvus frugilegus) may harvest and store food over quite large areas (Purchas 1980; Vander Wall and Balda 1981), parids will store in their year-round territories (e.g. Ekman 1989; Brodin 1993). Nutcrackers regularly transport seeds several kilometres before storing them (e.g. Vander Wall and Balda 1977). Unless caches are distributed in clumps a wider scattering of them may increase the need to remember their locations. Restricted random search will probably be more effective in a smaller area where caches have been concentrated, for example a territory.

One effect of body size that has hardly been discussed is that parids have much smaller brains and hippocampi than corvids meaning that they will have fewer neurons. Food hoarding occurs in families with larger hippocampi than nonhoarding families (Krebs et al. 1989; Sherry et al. 1989) suggesting that the number of neurons in some way may be important for memory storing capacity. To remember the individual positions of 7,700–9,500 caches is a formidable task for a nutcracker (Table 1), and if a willow tit could remember 60,000–150,000 caching positions this would be even more impressive. If brain volume ever can be predicted to have a limiting effect on storing capacity, this should be the occasion to detect it. The reason that parids need to remember more individual positions than corvids is that nutcrackers and jays may fill caching “holes” with 1–20 seeds or nuts (Mattes 1982) whereas parids store all items singly (e.g. Brodin 1994b). Below I discuss possible reasons for size-related differences in memory capacity in more detail by comparing one long-term hoarding species from each family.

Ecological differences between the families that may select for retrieval by memory

Most corvids seem to store in excavations in the ground whereas parids frequently store under tufts of lichens, in bark crevices, etc. To some degree this could also be considered as an allometric difference since the decisive factor here could be the large and powerful bills that make it possible for corvids to excavate underground caching holes. Food that is hidden below the ground over very large areas may be almost impossible to find using restricted random searches. Even if parid caches are inconspicuous and hard to find, they are probably easier to find than the underground caches of corvids. I have many years experience of observing storing parids in nature (Brodin et al. 1994, 1996; Brodin 2005) and hoarders of different species restrict their caching to certain types of species-specific sites (Haftorn 1956b). When I was searching for willow tit caches during earlier field work (Brodin 1993) I would frequently find sunflower seeds that I had provided to the birds under tufts of lichens even when I did not observe the actual caching act. Also, some species of parids will leave caches partially exposed, for example in tree needles (e.g. Brotons 2000). Caches that are visible could be relatively easy to locate with restricted random search compared to underground caches.

Comparison of brain and hippocampus size in a long-term hoarder from each family

In birds, the hippocampal complex or hippocampal formation (i.e. the hippocampus and the parahippocampus) is a part of the brain that is involved in the storage of spatial memories. I compare this structure, as an example, in the Eurasian jay and the willow tit—two highly specialised long-term hoarding species that have the largest hippocampi in their respective families (Brodin and Lundborg 2003). The jay has a body mass of around 195 g and the tit weighs 11–12 g (Table 2). Both species are forest-dwelling, largely arboreal foragers that reside in year-round territories.
Table 2

A comparison of features that might be important for memory longevity between a parid (the willow tit) and a corvid (the Eurasian jay). hc Hippocampus

Feature

Species

Ratio

Jay

Tit

Jay:tit

Body mass (g)

195a

12

16:1

hc volume (mm3)

116a

27b

4.3:1

hc/g body mass (mm3)

0.6

2.25

1:3.7

Neurons/mm3 in the hc (×103)

35c

40d

1:1.1

Neurons in hc (×106)

4.1

1.1

3.7:1

Neurons/g body mass (×104)

20.5

90

1:4.4

Potential connections (×1011)

41e

2.9e

14:1

aHealy and Krebs (1992)

bHealy and Krebs (1996)

cFrom magpie (Healy and Krebs 1993)

dFrom marsh tit (Healy et al. 1994)

eCalculated for P=0.5

It is not known exactly how memory is stored at the cellular level, but the most likely method is long-term potentiation of synapses that connect neurons (e.g. Stevens 1998). Then memory storage capacity will depend on the number of possible network connections (e.g. Horn 2004). Little is known about the degree of connectivity between neurons in the avian hippocampus. Instead I have taken data from the mouse Mus musculus cortex (Braitenberg and Schüz 1998). In the mouse cortex the probability of more than one connection between two neurons is negligible. The number of connections between neurons, C, increases with the number of neurons, n, as C=n(n−1)p2 where p is the proportion of other neurons to which a neuron connects. The network potential will increase faster with size than the number of neurons (Fig. 3), meaning the advantage of a large network compared to a smaller one will be larger than suggested by the volumetric difference.
Fig. 3

The number of potential connections as a function of the number of cells, given that the cells only connect once

A possible confounding factor in calculations of size-dependent brain capacity is that neuron densities may not be the same in brains of different size. I have not found any data on neuron densities in jays or willow tits and instead use those of hoarding relatives of similar size. Neuron density in the hippocampus of magpies, Pica pica, is around 35,000/mm3 (Healy and Krebs 1993) and in marsh tits around 40,000/mm3 (Healy et al. 1994). Though these estimates are rough, I have still compensated for the difference between them in my calculations. In reality the effect of such a difference will probably be counteracted by an increase with brain size in the number of synapses per neuron (Braitenberg and Schüz 1998).

Parids possess relatively larger hippocampi than corvids (Table 2). Although the body mass of the jay is around 16 times larger than the tit it has only about 3.6 times more neurons in the hippocampus. However, since the number of connections can increase faster than the number of neurons, the maximum number of potential connections can be calculated to be 14 times larger in the jay (Table 2).

Earlier discussions of hippocampus size in the context of food hoarding and memory in birds have focussed on relative size (Krebs et al. 1989; Sherry et al. 1989; Healy and Krebs 1992, 1996; Healy et al. 1994; Clayton and Krebs 1995; Hampton et al. 1995; Healy and Suhonen 1996; Basil et al. 1996; Clayton 1998; Brodin and Lundborg 2003), but it is also possible that absolute hippocampus size may be important. I am not suggesting that large size automatically gives high memory storage capacity, just that it may be a prerequisite for developing such a capacity.

How conclusive are the data?

Since there are so few studies on memory longevity in these two families some caution is necessary. For example, retrieval behaviour in the two families could appear to be more different than it actually is. It is possible that the behaviour that field workers have labelled restricted random search in parids in fact is memory-based retrieval. It may be good foraging economy to search nearby locations that may contain cached food at the same time as remembered caches are retrieved, and this may result in behaviour that looks like systematic search. Also, in laboratory experiments, the environment and food regime are highly artificial, and it is possible that the reported decay in retrieval success with time in parids (Hitchcock and Sherry 1990; Brodin and Kunz 1997) is due to decreasing motivation to retrieve rather than memory decay.

However, good foraging economy should also be important for corvids that need stored food to survive the winter. Also is seems strange that the laboratory environment should decrease the motivation to retrieve more in parids than in corvids. Healy and Suhonen (1996) did not find any memory decay over time in parids, but their study spanned less than half the time of the two experiments that detected a decay with time. This suggests that the latter studies were more sensitive to temporal trends. Also, various field workers have independently reached the same conclusion—that parids normally do not remember the locations of the caches they retrieve in winter (Haftorn 1956b, 1974; Gibb 1960; Brodin 1994a; Brodin and Ekman 1994).

In conclusion most evidence suggests that long-term hoarding corvids possess a memory for caching locations of long duration whereas parids do not. The most likely reasons for this are that the selection pressure for long-term memory is weaker in parids than in corvids and the memory storing capacity is limited in parids due to their small size. The reason that such size-related limitations on memory have not been discussed previously may be that possible physiological costs of memory have not been considered in this respect.

Notes

Acknowledgements

I thank Diana Tomback and Sue Healy for valuable comments and suggestions. This study was supported by a grant from the Swedish Research Council, Vetenskapsrådet.

References

  1. Alatalo RV, Carlson A (1987) Hoarding-site selection of the willow tit Parus montanus in the presence of the Siberian tit Parus cinctus. Ornis Fenn 64:1–9Google Scholar
  2. Andersson M, Krebs J (1978) The evolution of hoarding behaviour. Anim Behav 26:707–711Google Scholar
  3. Baker CM, Stone E, Baker AEM, Shelden RJ, Skillicorn P, Mantych MD (1988) Evidence against observational learning in storage and recovery of seeds by black-capped chickadees. Auk 105:492–497Google Scholar
  4. Balda RP, Bednekoff PA (1996) Observational spatial memory in Clark’s nutcrackers and Mexican jays. Anim Behav 52:833–839Google Scholar
  5. Balda RP, Kamil AC (1992) Long-term spatial memory in Clark’s nutcracker, Nucifraga columbiana. Anim Behav 44:761–769Google Scholar
  6. Barnea A, Nottebohm F (1994) Seasonal recruitment of hippocampal neurons in adult free-ranging black-capped chickadees. Proc Natl Acad Sci USA 91:11217–11221Google Scholar
  7. Barnea A, Nottebohm F (1995) Patterns of food storing by black-capped chickadees suggest a mnemonic hypothesis. Anim Behav 49:1161–1176Google Scholar
  8. Basil JA, Kamil AC, Balda RB, Fite KV (1996) Difference in hippocampal volume among food storing corvids. Brain Behav Evol 47:156–164Google Scholar
  9. Bednekoff PA, Balda RP, Kamil AC, Hile AG (1997) Long-term spatial memory in four seed-caching corvid species. Anim Behav 53:335–341Google Scholar
  10. Braitenberg V, Schüz A (1998) Cortex: statistics and geometry of neuronal connectivity. Springer, Berlin Heidelberg New YorkGoogle Scholar
  11. Brodin A (1992) Cache dispersion affects retrieval time in hoarding Willow Tits. Ornis Scand 23:7–12Google Scholar
  12. Brodin A (1993) Low rate of loss of willow tit caches may increase the adaptiveness of long-term hoarding. Auk 110:642–646Google Scholar
  13. Brodin A (1994a) Separation of caches between individual willow tits hoarding under natural conditions. Anim Behav 47:1031–1035Google Scholar
  14. Brodin A (1994b) The role of naturally stored food supplies in the winter diet of the boreal willow tit Parus montanus. Ornis Svecica 4:31–40Google Scholar
  15. Brodin A (1994c) Time aspects of food hoarding in the willow tit—an evolutionary perspective. PhD Thesis, Department of Zoology, Stockholm University, SwedenGoogle Scholar
  16. Brodin A (2005) Hippocampus volume does not correlate to food hoarding rates in the field in two closely related bird species, the black-capped chickadee Poecile atricapilla and the willow tit Parus montanus. Auk (in press)Google Scholar
  17. Brodin A, Ekman J (1994) Benefits of food hoarding. Nature 372:510Google Scholar
  18. Brodin A, Kunz C (1997) An experimental study of cache recovery by hoarding willow tits after different rentention intervals. Behaviour 134:881–890Google Scholar
  19. Brodin A, Lundborg K (2003) Is hippocampus volume affected by specialisation for food hoarding in birds? Proc R Soc Lond B Biol Sci 270:1555–1563Google Scholar
  20. Brodin A, Lens L, Suhonen J (1994) Do crested tits, Parus cristatus, store more food at northern latitudes? Anim Behav 48:990–993Google Scholar
  21. Brodin A, Lahti K, Lens L, Suhonen J (1996) A northern population of willow tits Parus montanus did not store more than southern ones. Ornis Fenn 73:114–118Google Scholar
  22. Brotons L (2000) Individual food-hoarding decisions in a nonterritorial coal tit population: the role of social context. Anim Behav 60:395–402Google Scholar
  23. Clayton NS (1998) Memory and the hippocampus in food-storing birds: a comparative approach. Neuropharmacology 37:441–452Google Scholar
  24. Clayton NS, Krebs JR (1995) Memory in food-storing birds: from behaviour to brain. Curr Opin Neurobiol 5:149–154Google Scholar
  25. Cowie RJ, Krebs JR, Sherry DF (1981) Food storing by marsh tits. Anim Behav 29:1252–1259Google Scholar
  26. Dukas R (1999) Costs of memory: ideas and predictions. J Theor Biol 197:41–50Google Scholar
  27. Ekman J (1989) Ecology of non-breeding social systems of Parus. Wilson Bull 101:263–288Google Scholar
  28. Emery NJ, Dally JM, Clayton NS (2004) Western scrub-jays (Aphelocoma californicus) use cognitive strategies to protect their caches from thieving conspecifics. Anim Cog 7:37:43Google Scholar
  29. Gibb JA (1960) Populations of tits and goldcrests and their food supply in pine plantations. Ibis 102:163–208Google Scholar
  30. Giuntoli M, Mewaldt LR (1978) Stomach contents of Clark’s nutcrackers collected in Western Montana. Auk 95:595–598Google Scholar
  31. Goodwin D (1951) Some aspects of the behaviour of the Jay (Garrulus glandarius). Ibis 93:602–625Google Scholar
  32. Haftorn S (1956a) Contribution to the food biology of tits especially about the storing of surplus food. Part III. The willow tit (Parus atricapillus L.). Det Kgl Norske Videnskabers Selskabs Skrifter 3:1–79Google Scholar
  33. Haftorn S (1956b) Contribution to the food biology of tits especially about storing of surplus food. Part IV. A comparative analyses of Parus atricapillus L., P. cristatus L. and P. ater L. Det Kgl Norske Videnskabers Selskabs Skrifter 4:1–54Google Scholar
  34. Haftorn S (1959) The proportion of spruce seeds removed by the tits in a Norwegian spruce forest in 1954–55. Kgl Norske Videnskabers Selskabs Forhandlinger 32:121–125Google Scholar
  35. Haftorn S (1974) Storage of surplus food by the Boreal chickadee (Parus hudsonicus) in Alaska, with some records on the Mountain chickadee (P. gambeli) in Colorado. Ornis Scand 5:145–161Google Scholar
  36. Hampton RR, Sherry DF, Shettleworth SJ, Khurgel M, Ivy G (1995) Hippocampal volume and food-storing behavior are related in parids. Brain Behav Evol 45:54–61Google Scholar
  37. Healy SD, Krebs JR (1992) Food storing and the hippocampus in corvids: amount and volume are correlated. Proc R Soc Lond B Biol Sci 248:241–245Google Scholar
  38. Healy SD, Krebs JR (1993) Development of hippocampal specialisation in a food-storing bird. Behav Brain Res 53:127–131Google Scholar
  39. Healy SD, Krebs JR (1996) Food storing and the hippocampus in Paridae. Brain Behav Evol 47:195–199Google Scholar
  40. Healy SD, Suhonen J (1996) Memory for locations of stored food in willow tits and marsh tits. Behaviour 133:71–80Google Scholar
  41. Healy SD, Clayton NS, Krebs JR (1994) Development of hippocampal specialisation in two species of tit (Parus spp.). Behav Brain Res 61:23–28Google Scholar
  42. Higuchi H (1977) Stored nuts (Castanopsis cuspidata) as a food resource of nestling varied tits (Parus varius). Tori 26:9–12Google Scholar
  43. Hitchcock CL, Sherry DF (1990) Long-term memory for cache sites in the black-capped chickadee. Anim Behav 40:701–712Google Scholar
  44. Horn G (2004) Pathways of the past: the imprint of memory. Nat Rev Neurosci 5:108–121Google Scholar
  45. Hoshooley JS, Sherry DF (2004) Neuron production, neuron number, and structure size are seasonally stable in the hippocampus of the food-storing black-capped chickadee (Poecile atricapillus). Behav Neurosci 118:345–355Google Scholar
  46. Hutchins HE, Lanner RM (1982) The central role of Clark’s nutcracker in the dispersal and establishment of whitebark pine. Oecologia 55:192–201Google Scholar
  47. James PC, Verbeek NAM (1984) Temporal and energetic aspects of food storage in northwestern Crows. Ardea 72:207–215Google Scholar
  48. Jansson C (1982) The year round diets of the willow tit Parus montanus Conrad and the crested tit P. cristatus L. Ph D Thesis, Faculty of Natural Sciences. University of Göteborg, GöteborgGoogle Scholar
  49. Kamil AC, Balda RP, Good S (1999) Patterns of movement and orientation during caching and recovery by Clark’s nutcrackers, Nucifraga columbiana. Anim Behav 57:1327–1335Google Scholar
  50. Krebs JR, Sherry DF, Healy SD, Perry H, Vaccarino AL (1989) Hippocampal specialization of food-storing birds. Proc Natl Acad Sci USA 86:1388–1392Google Scholar
  51. Lens L, Adriaensen F, Dhondt AA (1994) Age-related hoarding strategies in the crested tit Parus cristatus: should the cost of subordination be reassessed? J Anim Ecol 63:749–755Google Scholar
  52. Ligon JD (1978) Interdependence of pinon jays and pinon pines. Ecol Monogr 48:111–126Google Scholar
  53. Löhrl H (1950) Beobachtungen zur Soziologie und Verhaltensweise von Sumpfmeisen (Parus palustris communis) im Winter. Z Tierpsychol 7:417–424Google Scholar
  54. Marzluff JM, Balda RP (1992) The pinyon jay—behavioral ecology of a colonial and cooperative corvid. T & A D Poyser, LondonGoogle Scholar
  55. Mattes H (1982) Die Lebensgemeinshaft von Tannenhäher und Arve. Swiss Fed Inst For Res Rep 241:1–74Google Scholar
  56. Moreno J, Lundberg A, Carlson A (1981) Hoarding of individual nuthatches Sitta europaea and marsh tits Parus palustris. Holarctic Ecol 4:263–269Google Scholar
  57. Nakamura H, Wako Y (1988) Food storing behaviour of willow tit Parus montanus. J Yamashina Inst Ornithol 20:21–36Google Scholar
  58. Pravosudov VV (1985) Search for and storage of food by Parus cinctus lapponicus and P. montanus borealis (Paridae). Zoologicheskii Zhurnal 64:1036–1043Google Scholar
  59. Pravosudov VV (1986) Individual differences in the behavior of the Siberian tit (Parus cinctus Bodd.) and the willow tit (Parus montanus Bald.) in foraging and storing food. Soviet J Ecol 4:60–64Google Scholar
  60. Purchas TPG (1980) Feeding ecology of rooks (Corvus frugilegus) on the Heretaunga plains, Hawke’s Bay, New Zealand. Neth J Zool 7:557–578Google Scholar
  61. Sherry DF (1984) What food-storing birds remember. Can J Psychol 38:304–321Google Scholar
  62. Sherry DF, Krebs JR, Cowie RJ (1981) Memory for the location of stored food in marsh tits. Anim Behav 29:1260–1266Google Scholar
  63. Sherry DF, Vaccarino AL, Buckenham K, Herz RS (1989) The hippocampal complex of food-storing birds. Brain Behav Evol 34:308–317Google Scholar
  64. Shettleworth SJ, Krebs JR (1982) How marsh tits find their hoards: the roles of site preference and spatial memory. J Exp Psychol Anim Behav Proc 8:354–375Google Scholar
  65. Shettleworth SJ, Krebs JR (1986) Stored and encountered seeds: a comparison of two spatial memory tasks in marsh tits and chickadees. J Exp Psychol Anim Behav Proc 12:248–257Google Scholar
  66. Stevens CF (1998) A million dollar question: does LTP = memory? Neuron 20:1–2Google Scholar
  67. Stevens TA, Krebs JR (1986) Retrieval of stored seeds by marsh tits (Parus palustris) in the field. Ibis 128:513–525Google Scholar
  68. Suhonen J, Alatalo RV (1991) Hoarding sites in mixed flocks of willow and crested tits. Ornis Scand 22:88–93Google Scholar
  69. Swanberg PO (1951) Food storage, territory and song in the thick-billed nutcracker. In: Hörstadius S (ed) Proceedings of the 10th international ornithology congress 1950, Uppsala. Almqvist & Wiksell, Uppsala, pp 545–554Google Scholar
  70. Tomback DF (1980) How nutcrackers find their stores. Condor 82:10–19Google Scholar
  71. Tomback DF (1982) Dispersal of whitebark pine seeds by Clark’s nutcracker: a mutualism hypothesis. J Anim Ecol 51:451–467Google Scholar
  72. Vander Wall SB (1982) An experimental analysis of cache recovery in Clarks nutcracker. Anim Behav 30:84–94Google Scholar
  73. Vander Wall SB (1990) Food hoarding in animals. University of Chicago Press, ChicagoGoogle Scholar
  74. Vander Wall SB, Balda RP (1977) Coadaptions of the Clarks nutcracker and the pinon pine for efficient seed harvest and dispersal. Ecol Monogr 47:89–111Google Scholar
  75. Vander Wall SB, Balda RP (1981) Ecology and evolution of food-storage behavior in conifer-seed-caching corvids. Z Tierpsychol 56:217–242Google Scholar

Copyright information

© Japan Ethological Society and Springer-Verlag 2005

Authors and Affiliations

  1. 1.Department of Theoretical Ecology, Ecology BuildingLund UniversityLundSweden

Personalised recommendations