Clark’s nutcrackers (Nucifraga columbiana) are 130-140g corvids that inhabit alpine environments in montane regions of the western United States and Canada. These corvids are extremely active during autumn, when plants are producing seeds in preparation for winter, and during this period; they can catch an estimated 22,000-33,000 pine seeds in over 5,000 different locations.
These corvids cache in a variety of habitats, though it is most noticeable in windswept slopes where there is little or no floral cover, and these areas also have the least amount of snow accumulation during the winter. Because of this, the nutcrackers probably rely upon these areas to survive harsh winter storms, but they also cache near their breeding territories to ensure an adequate supply for their nestlings.
Clark’s nutcrackers exhibit remarkable spatial memory, and as a result, they can accurately remember the location of seed caches up to 9 months after storage. This sophisticated memory is very important to their survival, because it is estimated that they rely upon these caches for over 80% of their energy intake during winter and breeding periods, which directly affects fitness. They also cache above the treeline, which is a much more secure environment, due to a reduction in rodent foraging potential, and alpine conditions that prevent sprouting. These extraordinary birds rely entirely upon spatial memory in order to recover these seeds, without any fixed rules for cache placement, retracing routes, or even sensory cues from the seeds they cached.
In studies done by Peter A. Bednekoff and Russell P. Balba from North Arizona University and Eastern Michigan University respectively, they first began by collecting wild caught adults of unknown age and sex. There were eight subject birds in experiment 1 and ten birds in experiment 2, and all of the nutcrackers in the study had caches and recovered Pinus edulis seeds in previous studies.
The birds were contained individually in cages that were 0.51 m wide x 0.51 m deep x 0.72 m high, with a consistent 10:14 light:dark cycle, and were fed a diet of parrot pellets, sunflower seeds, turkey starter, and Tenebrio larvae, with powdered vitamins added at least twice per week. The nutcrackers were deprived of food 24 hours prior to experimentation, though they always had access to fresh water and ground oyster shells (calcium), and their weight was monitored.
There were two experimental rooms in which these experiments were conducted; the large experimental room, and the relatively small experimental room. The large room was 15.3 m long x 9.1 m wide x 2.8 m high, with a raised floor containing 330 holes arranged in 15 rows (A-O), which were spaced 55.9 cm apart, and 22 columns (1-22) spaced 60.5 cm apart, with 11 and 12 spaced 128.3 cm apart. There were also six 2 m high perches placed around the edges of the large experimental room, with 77 various landmarks. The relatively small experimental room was 3.4 m long x 3.7 m wide x 2.6 m high, and it’s floor had 237 holes in a grid with 17 columns (A-N) and 14 rows. The holes in the floor were 5.5 cm in diameter and spaced 21 cm apart, with each hole containing a wooden plug or a 4 cm deep, sand-filled paper cup. The small room also contained within it 10 potential landmarks of various shapes and sizes.
Both rooms had a 1 m high feeding platform, and each room had rocks, bricks, and boards of various shapes and sizes, with posters on each wall. The nutcrackers would be allowed into the room through a porthole on the wall, and were observed through one-way glass during caching sessions. In order to allow an adequate adjustment period, the nutcrackers were habituated to each room configuration in one-hour sessions on 5 days during the 15-day period before the experiment took place.
121 seeds would be deposited onto the feeding platform, with 22 dehulled to facilitate rapid feeding, and subsequently, the nutcrackers would cache all of the seeds in various locations throughout the respective experimental room. Caching sessions were followed by two recovery sessions 7-9 days later, and in each recovery session they were only allowed to recover half of their caches before the lights were turned off, which signified the end of the day. During second recovery sessions, revisits to previously recovered caches were not recorded as error, because Clark’s nutcrackers retain some memory of cache locations after recovery.
The goal of experiment 1 was to test whether increasing the number of recognizable landmarks in the large experimental room would affect recovery accuracy. These two landmark treatments showed no statistical significance, and across the two treatments the nutcrackers made a similar number of caches, with a similar number of seeds, at a similar distance from the feeding platform. Though there was some observation on increased recovery accuracy due to landmarks, it varied considerably among individual birds, which showed that adding objects to the large experimental room did not consistently increase recovery accuracy. Due to these results, it was predicted that the recovery accuracy would be considerably higher in the smaller experimental room than it had been in the larger one.
As one might expect, nutcrackers recovered their caches more accurately in the smaller experimental room than in the larger one, though this difference was much greater than the differences expected due to more holes being available in the large experimental room. Clark’s nutcrackers made an average of 0.597 revisits per cache site in the small room, and 0.853 revisits per cache in the large room, though this difference is not statistically significant (p = 0.137). In conclusion, recovery accuracy was greater in a smaller area, but were not affected by the number, shape or size of objects present in the room. However, the space in which spatial memory tasks take place can affect performance, as exhibited by other studies that the authors cited.