Abstract

When a forager encounters an unfamiliar type of food, it must decide whether to eat it and risk being poisoned or avoid eating it and risk forfeiting a potentially valuable resource. Birds typically respond to such situations with “dietary wariness”; they show a transient aversion to approaching new food (neophobia), and many individuals also show a much longer lasting reluctance to consume the new food (dietary conservatism), even once neophobia has waned. Very little is known about how these processes, together termed “wariness,” are controlled. We therefore present a series of experiments investigating how wariness of novel foods in domestic chicks, Gallus gallus domesticus, can be deactivated and reactivated by different experiences of colored foods, varying in their degree of novelty and palatability. We found that prior experience of a single novel color of palatable chick crumbs was sufficient to deactivate both neophobia and dietary conservatism of any other novel color of crumbs tested. Relatively little prior experience of a novel training food was needed to deactivate neophobia, after which the birds would peck at any other novel food. In contrast, much more extensive experience of eating a novel training food was needed before the birds would incorporate other novel foods into their diet. Chicks needed direct physical contact with the training food before they overcame their wariness to eat another novel food. However, observational learning was sufficient to encourage them to peck at the food (overcoming their neophobia). Reinstating wariness was much more easily achieved than its deactivation. We discuss these surprising results in relation to the foraging behavior of wild and domestic birds.

How do foragers extend their diets from known safe foods to include new foods of unknown value? When animals first encounter unfamiliar food, they are often hesitant to approach it. This apparent fear of novelty, termed neophobia, was first described in rats encountering novel objects (Barnett 1958) but has since been described in many animal groups in response to novel foods (reviews in Brigham and Sibly 1999; Kelly 2001; Mappes et al. 2005; Marples et al. 2005). Neophobia toward food is a short-lived response, lasting only a few minutes in most animals (Marples and Kelly 1999). In birds, however, a second response to novel foods, termed dietary conservatism, has been described, in which some (but not all) members of the population continue to refuse to eat even fully palatable novel food, for very extended periods—often lasting months or years (Marples et al. 1998; Kelly 2001). This trait has been shown to have a genetic basis (Marples and Brakefield 1995) and the population to split into those with the trait, detectable after a few minutes of avoidance, and those without it, who eat novel food readily. Thus, it is possible to detect birds with long-lasting dietary conservatism using a test lasting only a few minutes (Marples and Kelly 1999). This protracted reluctance to eat novel food continues even though the birds may touch and investigate it thoroughly and watch other individuals consuming it. Unlike food neophobia, dietary conservatism, therefore, does not imply fear of the food but rather a refusal to extend the diet to include the new food type. Therefore, in order for an animal to extend its diet to include the novel food, it must overcome both its neophobia and its dietarily conservative avoidance of that food.

For brevity, the 2 processes of food neophobia and dietary conservatism together can be referred to as dietary wariness (cf. Mappes et al. 2005; Marples et al. 2005). Neophobia can be measured separately as the time before a bird first touches a new food item, that is, when fear of approach has subsided; and dietary wariness is the total time before the food type is eaten whenever encountered, that is, when the item has been accepted into the diet (Kelly and Marples 2004).

Expansion of the diet is of considerable interest in a number of ecological and commercial contexts. For instance, it is important for understanding the evolutionary origins of prey signaling systems because highly conservative foragers have been shown to allow even very conspicuous new prey morphs to survive and spread through a prey population simply as a result of their novelty (Thomas et al. 2003, 2004). This provides the opportunity for predators to learn about any associations between novel colors and unprofitability of prey and may, therefore, facilitate the evolution of aposematism (Mappes et al. 2005; Marples et al. 2005).

Understanding dietary change is also increasingly important for conservation because changes in climate and human activities such as farming practices are altering the foods available to many species (Fuhrer 2003; Logan et al. 2003) and changing their geographical distribution (Sanchez-Lafuente et al. 2001; Simmons et al. 2004; Suarez-Seoane et al. 2004; Chen and Kang 2005; Martin and Fitzgerald 2005).

In a commercial context, broiler chickens are given several changes in diet over the course of their development, and a considerable reduction in growth rate is evident with each dietary change (Cooper 1971; Murphy and Duncan 1977; Poole 1999). This happens despite selection over many generations for weight gain, so one would expect poultry to show a lower level of food wariness than many wild species. Control of dietary wariness could be used to alleviate the problems associated with dietary change and so improve the welfare and productivity of captive poultry.

Despite its relevance, very little is known about how dietary wariness is mediated or controlled. Previous studies show that experience of a number of colors of food may reduce the neophobic response to a different novel color, provided that the experience did not include distasteful prey. For example, chicks which had been familiarized with food of a range of colors (red, yellow, and green) were far less hesitant to approach and peck at food of an unfamiliar color (blue) than were chicks without such experience (Jones 1986). Similarly, Schlenoff (Schlenoff 1984) familiarized blue jays, Cyanocitta cristata, with either red or blue palatable seeds, as well as yellow seeds that were either palatable or unpalatable. The experience of unpalatable yellow novel food inhibited these birds from expanding their diet to novel red or novel blue foods, at least in the short term.

In our present study, we explore the deactivation and reactivation of both neophobia on its own, and overall dietary wariness, in domestic chicks, greatly extending the work described above. We present 4 experiments: The first demonstrates that wariness of a novel color of food may be deactivated by prior exposure to different novel colors. The second explores how long an exposure to the novel food is needed in order to deactivate wariness, and the third investigates what type of exposure is needed. The final experiment demonstrates the effect of experience with distasteful novel food, showing that wariness can be easily and rapidly reactivated. We discuss the implications of these findings for understanding dietary expansion in birds.

GENERAL METHODS

Housing and maintenance of chicks

In each experiment, we obtained day-old male chicks of the “Cobb 500” strain from a commercial hatchery and housed them together in a 2-m-long × 60-cm-wide wooden holding pen, with the floor covered in wood shavings. The chicks were kept at an ambient temperature of approximately 20 °C, with infrared heat lamps provided at one end of the pen for extra warmth. The birds were fed undyed (i.e., light brown in color) chick starter crumbs (William Connolly & Sons, Red Mills, Goresbridge, Co. Kilkenny, Ireland) and water from gravity feeders. These were available ad libitum throughout the study period, with the exception of 1 h before experimental trials, when the chicks were deprived of food but still had access to water. Each chick was marked with a unique color combination on its head, using permanent marker pens. This had no adverse effect on the chicks, which did not appear to respond to these marks on themselves or on the other chicks. Most of the chicks were designated as experimental chicks, but a few (number defined in each experiment below) were randomly chosen to be “buddy” chicks. These were companion chicks placed in a separate section of the test arena behind chicken wire, so they were in visual and auditory contact with the test chick. The buddy chick could not influence the foraging preferences of the experimental chick because it could not approach the test food, but it prevented the experimental chick from becoming distressed at being alone (Marples and Roper 1996).

Pretraining

When the chicks were 2 days old, they were familiarized with the testing arena, which consisted of a circular cardboard arena, 40 cm in diameter and 35cm high. A small section of the arena was separated with chicken wire to contain the buddy chick during training and testing. Each chick was familiarized with this experimental setup over 4 “pretraining” sessions, during which it was placed in the arena and allowed to eat its familiar food, undyed chick crumbs. After these pretraining sessions, the chicks were assigned to treatment groups, balanced across these groups for how quickly they had eaten in the final pretraining session.

Measuring neophobia and dietary wariness

In our experiments, the duration of neophobia was measured as the latency to begin pecking at a novel food. This, therefore, represents the time during which a bird is afraid to approach and make contact with a novel food, so the food is safe even from investigatory attacks (Marples and Kelly 1999). In contrast, the duration of overall dietary wariness was measured as the latency from the start of the trial until the bird ate the novel food consistently (i.e., ingested 10 crumbs in experiment 1, or 3 crumbs in all subsequent experiments). The duration of overall wariness therefore incorporates both the duration of neophobic avoidance plus the duration of avoidance due to dietary conservatism. Overall wariness is therefore a biologically meaningful measurement of the time taken to incorporate a novel food into the diet and therefore of the time for which the prey is protected to some extent by its novelty.

EXPERIMENT 1: THE EFFECT OF PRIOR EXPERIENCE OF NOVEL COLORS ON DIETARY WARINESS

In experiment 1, we used 120 experimental chicks and 4 buddy chicks. Thirty-six chicks were assigned to the “control” group (group 1) and received only undyed chick crumbs during their training sessions. Eighteen chicks were assigned to group 2 (blue, single-color treatment), and 18 were assigned to group 3 (red, single-color treatment). The blue food was made by dying chick crumbs with blue food coloring (O'Brien's liquid blue FCF) and the red with red food coloring (O'Brien's Christmas red). Forty-eight chicks were assigned to the “multicolored” group (group 4) and were each familiarized with 4 different colors during training sessions: red, green (O'Brien's liquid green 90), orange (20 parts of O'Brien's lemon yellow to 1 part of O'Brien's Christmas red), and blue. We used 48 chicks in this group so that all possible orders of presentation for the 4 colors during training sessions could be used twice (see below).

Each chick was given 4 training sessions; one in the morning and one in the afternoon of the 2 days after their pretraining. In each session, the chick was placed into the main part of the arena, and one of the buddy chicks (previously fed to satiation) was placed in the small “buddy area.” Each experimental chick was offered a small pile of food of the color dictated by its treatment group (see Table 1). Birds in the multicolored treatment group were offered one of the 4 colors at each training session, so that each chick had encountered all 4 colors by the end of the 4 training sessions. This treatment was balanced for the order in which the 4 colors were presented to the chicks such that all 24 possible orders were used twice. Thus, birds in group 4 experienced 4 different novel colors of food, whereas those in groups 2 and 3 experienced one novel color of food on 4 occasions.

Table 1

The design of experiment 2(i), testing the effect of short-term exposures to novel food

Treatment group Training session 1 (1 min) Training session 2 (1 min) Training session 3 (1 min) Training session 4 (1 min) Test session (3 min) 
Group 1 (N = 12) Uncolored Uncolored Uncolored Uncolored Blue 
Group 2 (N = 12) Red Uncolored Uncolored Uncolored Blue 
Group 3 (N = 12) Red Red Uncolored Uncolored Blue 
Group 4 (N = 12) Red Red Red Uncolored Blue 
Group 5 (N = 12) Red Red Red Red Blue 
Treatment group Training session 1 (1 min) Training session 2 (1 min) Training session 3 (1 min) Training session 4 (1 min) Test session (3 min) 
Group 1 (N = 12) Uncolored Uncolored Uncolored Uncolored Blue 
Group 2 (N = 12) Red Uncolored Uncolored Uncolored Blue 
Group 3 (N = 12) Red Red Uncolored Uncolored Blue 
Group 4 (N = 12) Red Red Red Uncolored Blue 
Group 5 (N = 12) Red Red Red Red Blue 

Each training session lasted for 3 min or until the chick had eaten 10 crumbs, whichever was shorter. Trials were kept to a maximum of 3 min because a lone chick tended to go to sleep if left to forage for longer than that. If the chick had not pecked at the food at all after 2 min, it was picked up and placed next to the food to ensure that it had seen it, whereon most of these chicks began to peck. If they did not peck before the end of the trial (18 chicks in total, 17 of which were in the control group), they were scored as taking 180s to peck, that is, the full trial duration.

On the day after training, all chicks were tested for their response to a novel color of food. Black was chosen to be the novel color as it was clearly distinct from the 4 colors used during training. Chicks were offered a small pile of black food, which was novel for all the treatment/control groups. This “test session” continued for 3 min or until the chick had eaten 10 of the crumbs, whichever was shorter. The latency to peck at the food and the latency to eat 10 crumbs were recorded. The chicks were observed directly from a distance of approximately 1 m, and pecking and eating were readily distinguished from this distance.

Results

Effects on neophobia

As the single-color treatments (groups 2 and 3) were not significantly different from each other, they were pooled for the analysis. The chicks that had been familiarized with either a single color (groups 2 and 3) or multiple colors (group 4) were faster to peck at the novel black food than were the control chicks (group 1) that had never experienced dyed chick crumbs (Figure 1a, Kruskal–Wallis test among treatment groups: H2 = 30.44, P < 0.001). Thus, we found no evidence that experience of a single color, be it blue (group 2) or red (group 3), was any less effective at deactivating neophobia than experience of 4 different colors (group 4), despite the latter treatment involving considerably more experience of novelty (Dunn's post hoc test groups 2 vs. 4, rank difference = 4.09, critical value = 18.65, P = not significant [NS]; and groups 3 vs. 4, rank difference = 7.7, critical value = 18.26, P = NS).

Figure 1

The effect of prior experience with novel colored food on (a) neophobia and (b) overall wariness. Bars show mean latency (s) to (a) start pecking at or (b) eat the novel food after 4 training sessions with different colors of food. Group 1 (control), N = 36; group 2 (single color, blue), N = 18; group 3 (single color, red), N = 18; and group 4 (4 colors), N = 48. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

Figure 1

The effect of prior experience with novel colored food on (a) neophobia and (b) overall wariness. Bars show mean latency (s) to (a) start pecking at or (b) eat the novel food after 4 training sessions with different colors of food. Group 1 (control), N = 36; group 2 (single color, blue), N = 18; group 3 (single color, red), N = 18; and group 4 (4 colors), N = 48. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

Effects on overall dietary wariness

The experiment revealed that the chicks were also much faster to eat the black food consistently (i.e., ingest 10 crumbs, as opposed to simply pecking at the food) if they had previously been familiarized with novel colored food (Figure 1b, Kruskal–Wallis test among treatment groups: H2 = 40.08, P < 0.001). This shows that experience of eating novel colored food deactivates dietary conservatism as well as neophobia. As with neophobia, the number of colors with which the chicks had been familiarized made little difference to the degree of deactivation of wariness because experience with one color on 4 occasions was not significantly less effective than experience with 4 different colors (Dunn's post hoc test groups 2 vs. 4, rank difference = 7.38, critical value = 18.65, P = NS; and groups 3 vs. 4, rank difference = 15.96, critical value = 18.26, P = NS).

Thus, our first experiment showed that deactivation of both aspects of wariness does not appear to depend on the breadth of experience of novelty, but it is still possible that it may depend on its extent, that is, for how long the birds had experienced the training exposure to novel food. Furthermore, experiment 1 only tested the effect using one color of test food (black). The following experiments investigated the extent and type of experience needed to cause this deactivation of wariness with a wider range of colors of test food.

EXPERIMENT 2: THE DURATION OF EXPOSURE TO NOVEL FOOD NEEDED TO DEACTIVATE DIETARY WARINESS

Method

This experiment comprised 2 sets of trials: (i) one set investigating short exposures to the training food and (ii) the second set using much longer exposures to the food during training.

Short-term exposures

Sixty chicks were pretrained as described in the general methods and then divided into 5 treatment groups as set out in Table 1.

The chicks were given a series of 4 training sessions, 2 per day, each lasting 1 min, in which they were offered either uncolored or red chick crumbs in a small pile in the testing arena. The treatment groups differed only in the number of training sessions in which red food was offered. The day after the final training session, each chick was offered a second novel color of food (blue), during a 3-min test session in the test arena. We recorded the latency to the first time each chick pecked at this food and the latency to it eating 3 of the crumbs. Observations from experiment 1 revealed that chicks that ate 3 crumbs would invariably continue until they had eaten 10, so the criterion for incorporation of the novel food into the diet was reduced to 3 crumbs for this and subsequent experiments.

Long-term exposures

After the usual pretraining, 60 chicks were divided into 4 treatment groups that differed in the duration of their training. Group 1 acted as a control group, having no exposure to novel food during the training sessions. Group 2 were exposed to the training food for 12 min in one continuous exposure and with the rest of the chicks from that group (social training) in a large arena (80 cm in diameter). Groups 4 and 5 experienced longer durations of social training with the novel food (25 min and 40 min, respectively). As chicks tend to fall asleep after about 3 min of foraging alone, these birds were trained socially. This ensured that the exposure time was meaningful.

The day after training was complete, each bird underwent a 3-minute test session in which it was offered the test food, blue dyed chick crumbs, in the normal test arena. We recorded the latency to the first peck at the blue chick crumbs and the latency to eat 3 crumbs.

Results

Effects of short-term exposures on neophobia

Short-term exposures to red food were effective in reducing neophobia (Figure 2a, Kruskal–Wallis test across all groups: H4 = 16.67, P = 0.002; significant Dunn's post hoc tests groups 1 vs. 2, P < 0.05; group 1 vs. groups 4 and 5, P < 0.001; groups 3 vs. 5, P < 0.05). This reduction was even significant after only 1 min of exposure (Mann–Whitney U test between groups 1 and 2: U = 38, N1 = N2 = 12, P = 0.036). However, neophobia was not entirely deactivated even by 2 min of exposure because the latency to peck continued to fall with greater exposure to red food during training (Figure 2a, Spearman rank correlation between latency to peck and exposure time for groups 3–5, rS = −0.293, N = 48, P = 0.044).

Figure 2

The effect of the duration of training with novel colored food on neophobia (a, b) and overall wariness (c, d). Bars show the mean latency (s) to peck at novel blue food after short-term training (a, c) and longer term training (b, d) with different durations of exposure to novel red food. Short-term training (a, c): group 1 (control), N = 12; group 2 (1 min training), N = 12; group 3 (2 min training), N = 12 min; group 4 (3 min training), N = 12; and group 5 (4 min training), N = 13. For full details of treatments, see Table 1. Longer term training (b, d): group 1 (control), N = 11; group 2 (12 min training), N = 12; group 3 (25 min training), N = 12; and group 5 (40 min training), N = 13. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

Figure 2

The effect of the duration of training with novel colored food on neophobia (a, b) and overall wariness (c, d). Bars show the mean latency (s) to peck at novel blue food after short-term training (a, c) and longer term training (b, d) with different durations of exposure to novel red food. Short-term training (a, c): group 1 (control), N = 12; group 2 (1 min training), N = 12; group 3 (2 min training), N = 12 min; group 4 (3 min training), N = 12; and group 5 (4 min training), N = 13. For full details of treatments, see Table 1. Longer term training (b, d): group 1 (control), N = 11; group 2 (12 min training), N = 12; group 3 (25 min training), N = 12; and group 5 (40 min training), N = 13. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

Effects of longer term exposures on neophobia

Compared with no exposure, social exposure for 12 min to red food significantly reduced the aversion to blue food (Figure 2b; Mann–Whitney U test between groups 1 and 2, U = 0.00, Z = −4.31, N = 11, 12, P < 0.001). Longer term exposure to the red training food did not significantly reduce the neophobic response further (Figure 2b, Spearman rank correlation across groups 2, 3, and 4, rs = −0.216, N = 37, P = 0.199; Dunn's post hoc test shows no differences between groups 2, 3, and 4). Thus, neophobia was effectively deactivated by 12 min of exposure to a single color of food.

Effects of short-term exposure on overall dietary wariness

Wariness was not deactivated by short-term training (Figure 2c) as the latency to eat 3 blue crumbs was not significantly altered by any of the short-term training regimes. (Kruskal–Wallis test across all groups: H4 = 3.05, P = 0.549.)

Effects of longer term exposure on overall dietary wariness

Longer term exposure to the training food (Figure 2d) did reduce wariness (Kruskal–Wallis test among all groups: H3 = 10.41, P = 0.015), but at least 40 min of exposure was needed for effective deactivation. The groups which had experienced red food for up to 25 min were not significantly quicker to eat 3 crumbs of the test food than the control birds who had never experienced novel food at all (Kruskal-Wallis test between groups 1, 2, and 3: H2 = 2.84, P = 0.242, and Dunn's post hoc test across all groups shows no differences between groups 1, 2, and 3). However chicks in group 4 (which had been exposed to the training food for 40 min) began to eat significantly more quickly than the control chicks (Mann–Whitney U test between groups 1 and 4: U = 33, N1 = 13, N2 = 11, P = 0.005, Dunn's post hoc test across all groups shows group 1 vs. group 4, P < 0.01).

EXPERIMENT 3: THE TYPE OF EXPERIENCE NEEDED TO DEACTIVATE DIETARY WARINESS

Method

In 2 sets of trials, 56 chicks in one set and 60 chicks in the other were pretrained as above, then divided into 5 treatment groups. The only difference between the 2 sets of trials was that in one set the training food was red and the test food blue and in the other set the training food was blue and the test food red. As these gave very similar results, the 2 sets of trials were pooled and will be described together.

All chicks were given 4 training sessions, each lasting 3 min. The control group (group 1) was given no experience of novel foods during training. The chicks in groups 2 and 3 were kept behind a barrier made of chicken wire during training, so they could see the food but not approach it. Group 2 chicks just saw a pile of the colored food, whereas group 3 chicks observed a buddy chick eating the novel colored food (the buddy chick having previously been familiarized with this food). The chicks in group 4 were allowed to eat the food during their training sessions. The aim of the group 5 training was as a positive control, in which the chicks were fully accustomed to the training food, so that all novel food avoidance was deactivated. To do this, they were given the usual 4 exposures to feeding on the training food in the test arena and then an additional 48 min of access to the food in a social situation. The results of the previous experiment would suggest that their dietary wariness should have been effectively deactivated by such training.

The day after training, the birds were each given a 3-min test session in which they were offered a second new color of food (blue or red), depending on which color had not been used in their training sessions. Their latencies to peck at the food and eat 3 of the chick crumbs were recorded.

Results

Certain types of exposure to the food reduced both neophobia (Kruskal–Wallis test among all groups: H4 = 48.15, P < 0.001) and overall wariness (Kruskal–Wallis test among all groups: H4 = 65.47, P < 0.001) to the next novel food they encountered. Dunn's post hoc tests revealed that chicks which did not have direct experience of eating the novel food during training (groups 2 and 3) were not significantly quicker than the control birds (group 1) either to peck at the test food (Figure 3a) or to eat it consistently (Figure 3b). So, observation of the food or seeing a conspecific eating the food was not sufficient to deactivate neophobia or overall wariness in most of the birds. In contrast, chicks which did have direct experience of eating the novel food (groups 4 and 5) were significantly faster than control birds to peck at the novel food (Dunn's post hoc test P < 0.01) but only those which had extensive experience (group 5) were faster than control to eat the food (P < 0.01). Thus, this experiment demonstrates that direct experience of eating novel food was much more effective at deactivating both neophobia and overall wariness than simply seeing the food or watching it being eaten by other chicks.

Figure 3

The effect of different types of training with novel food on (a) neophobia and (b) overall wariness. Bars show the mean latency (s) to (a) peck at or (b) to eat 3 crumbs (s) of novel food after training with different durations of exposure to another color of novel food. Group 1 (control), N = 23; group 2 (observe novel food 4 × 3 min), N = 25; group 3 (observe conspecific eat novel food 4 × 3 min), N = 22; group 4 (4 × 3 min eating novel food, solo foraging), N = 24; and group 5 (4 × 3 min eating novel food, solo foraging followed by 48 min social foraging) N = 24. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

Figure 3

The effect of different types of training with novel food on (a) neophobia and (b) overall wariness. Bars show the mean latency (s) to (a) peck at or (b) to eat 3 crumbs (s) of novel food after training with different durations of exposure to another color of novel food. Group 1 (control), N = 23; group 2 (observe novel food 4 × 3 min), N = 25; group 3 (observe conspecific eat novel food 4 × 3 min), N = 22; group 4 (4 × 3 min eating novel food, solo foraging), N = 24; and group 5 (4 × 3 min eating novel food, solo foraging followed by 48 min social foraging) N = 24. Error bars show ±1 standard error. Bars bearing different letters are significantly different with P < 0.05.

As in experiment 2, overall wariness was considerably more resistant than neophobia to deactivation. The birds which were allowed to eat the training food for 12 min showed reduced neophobia (pecking latency Dunn's post hoc test groups 1 vs. 4, P < 0.001) but no significant reduction in overall wariness (eating latency Dunn's post hoc test groups 1 vs. 4, P = NS). Only after an hour of training, much of it in a social situation did the birds readily eat the novel test food (eating latency Dunn's post hoc test groups 1 vs. 5, P < 0.001).

Our final experiment investigated whether wariness can be reactivated by encounters with unpalatable food.

EXPERIMENT 4: REACTIVATION OF WARINESS BY EXPERIENCE OF UNPALATABLE FOOD

Method

Thirty chicks were trained in a social context on palatable novel red food for 2 h, so their wariness of novel food was expected to be effectively deactivated (as shown in the experiments above). The chicks were then divided into 2 equal groups. Group 1 experienced one 3-minute training session, again in a social group, with red food that had been soaked with denatonium benzoate (“Bitrex,” a solution of 2.5 w/v in water). This is an extremely bitter substance that has been shown to cause single-trial avoidance learning in chicks (Marples and Roper 1997; Skelhorn and Rowe 2005). The chicks were allowed to peck at the distasteful food until they showed a “headshake response,” a behavior characteristic of disgust (i.e., rejection of distasteful food) in chickens (Lowndes and Stewart 1994; Rickard et al. 1994). Only chicks in group 1 that showed this response were included in the experiment to ensure that they had definitely experienced the unpalatability of the red food. Group 2 (the “wariness deactivated” group) were offered an equivalent training trial, but in their case, the red food was fully palatable.

The following day, all chicks had a 3-min test session with fully palatable novel blue food, in order to determine whether their wariness responses had been reactivated by experience of the distasteful red food. Their latencies to peck at the food and eat 3 of the chick crumbs were recorded.

Results

The birds which had experienced distasteful red food took significantly longer to begin pecking at the novel test food than the control group (Mann–Whitney U test between groups 1 and 2: U = 44.5, N1 = 15, N2 = 14, P = 0.008). Thus, a single experience of unpalatable food, even after extensive experience during which that color of food was palatable, was enough to cause the birds to avoid pecking at the next novel food encountered (Figure 4a).

Figure 4

The effect of a single encounter with distasteful novel food on (a) neophobia and (b) overall wariness. Bars show the mean latency (s) to (a) peck at and (b) to eat 3 crumbs of novel blue food after a single experience of distasteful red food (group 1) or after experience of only palatable red food (group 2). Error bars show ±1 standard error.

Figure 4

The effect of a single encounter with distasteful novel food on (a) neophobia and (b) overall wariness. Bars show the mean latency (s) to (a) peck at and (b) to eat 3 crumbs of novel blue food after a single experience of distasteful red food (group 1) or after experience of only palatable red food (group 2). Error bars show ±1 standard error.

The latency of the chicks in group 2 to eat the food was significantly shorter than in group 1 (Figure 4b, Mann–Whitney U test between groups 1 and 2: U = 45.0, N1 = 15, N2 = 14, P = 0.001), indicating that overall wariness was reactivated by just a single experience of unpalatable food. Furthermore, there was a striking and highly significant difference in the number of chicks that ate any of the food during the 3-min test session. None of the 15 chicks in group 1 ate the blue crumbs after experiencing distasteful red crumbs, whereas 8 of the 14 chicks which had only experienced palatable food ate the test food (Fisher's Exact test P < 0.001).

It is also interesting to note that even after 2 h of exposure to the palatable novel red food in a social environment during training, 6/14 of the chicks in group 2 did not eat the blue test food, showing that dietary conservatism was not deactivated in these individuals. This suggests that a substantial proportion of the population may be very resistant indeed to the deactivation of the dietary conservatism component of their overall wariness.

DISCUSSION

Avoidance of novel food is likely to be an adaptive trait in a world where many unfamiliar potential foods are toxic and the effects of such toxicity may be delayed. However, it must sometimes be advantageous for foragers to try new foods and extend their diet (for example when seasonal foods become available or unavailable). To do this, they would need to overcome both their neophobia and dietary conservatism. Our experiments comprise the first systematic investigation into how wariness can be deactivated, and the results have a number of important implications.

First, our experiments demonstrate that neophobia is only one component of the decision to begin eating a new food. For example, in experiment 2(ii), neophobia was effectively deactivated by training for 12 min and was significantly reduced after a single minute of exposure to a novel food, whereas it took between 25 and 40 min to deactivate overall wariness in the majority of chicks (and longer in some individuals). The discrepancy between these 2 times can only be due to the second component of overall wariness, namely dietary conservatism, delaying dietary incorporation even once neophobia has been deactivated.

Second, our results reveal that deactivation of both components of wariness required the individual to have direct contact with the training food. Observational learning (by watching other individuals foraging) was not sufficient. This is surprising because chicks are highly social animals, and there is good evidence (Wauters et al. 2002) that chicks learn to identify foods from their mother through social facilitation (the chick copying their mother's behavior) and possibly local enhancement (the mother directing the chick's attention to the food).

A third striking aspect of our results was that the chicks did not need experience of a particular color of food for their wariness to that color of food to be deactivated. Experience of a single novel food was enough to speed up acceptance into the diet of the next novel food encountered, even though the 2 novel foods were completely different in color. This finding is reminiscent of generalization of one cue to another (Weiss and Weissman 1992; Baddeley and Jones 2001; Jones et al. 2001; Jansson and Enquist 2003) but is far more extreme than classic generalization. Birds have previously been found to generalize from the training color only as far as different shades of that same color, and responses fall off quickly when the hues become too different (Weiss and Weissman 1992; Jones et al. 2001). In contrast, in our study, the birds were less averse to a completely different color after experience of the training food.

One explanation of this is that the birds were using novelty itself as a cue and were learning during the first experience that novel objects were safe and could be incorporated into the diet. This implies that these birds did not learn that any particular color is safe but learned that novel foods in general were safe. Consistent with this view, we found in experiment 4 that a single brief experience of unpalatability (not accompanied by any toxicity) was enough to reverse the effect of 2 h of familiarization with the palatable form of that novel food. Thus, experience of novelty only deactivated wariness when all experiences of novel food were positive, suggesting that the birds were indeed learning the association between novelty and the palatability of novel food.

An alternative interpretation is that the birds may have formed a mental image of this food type (chick crumbs) when they first encountered it and learned to associate this image with edibility (Dukas and Kamil 2001; Blough 2002). After experience of a novel color of this food, the chicks then generalized their mental image to accept similar looking food of any color, that is, color ceased to be a salient part of the mental image of that food.

Our experiments do not allow us to distinguish between these 2 explanations, but both differ from the orthodox view of how birds respond to novel food and generalize about familiar foods. Whichever cognitive process is responsible, chicks abandon their wariness of novel colors of foods more easily than might be expected given the potential loss of fitness incurred by eating toxic food (Marples et al. 1989). Indeed, it is hard to imagine that dietary wariness would ever have evolved if novel colors of food were always safe to eat.

If our findings using domestic chicks are representative of wild foragers, then foragers which encounter a large number of different consistently palatable food types would have their wariness deactivated most of the time. Our results show that there is, however, a striking asymmetry in the effect of palatable and unpalatable experiences because a relatively long period of exposure with palatable novel food was required to turn off wariness, whereas these responses can be reactivated by only a very brief experience of unpalatable novel food. Therefore, a wild forager's willingness to eat novel food is likely to be reduced very quickly whenever it encounters unpalatable food. Whether a forager shows wariness on encountering a new food item will therefore depend on its recent history of encounters with palatable and unpalatable food. Thus, a patchwork of foraging strategies might be expected across spatially heterogeneous environments. These findings may therefore help to explain the very variable responses of wild birds to novel prey that have been observed in field studies (Lindström et al. 2001; Ruxton et al. 2004; Mappes et al. 2005). It is clearly important that we understand both the activation and deactivation of wariness if we are fully to understand the foraging decisions of foragers in the wild.

Although patterns of forager wariness in natural environments appear to be complex, our study shows that controlling wariness in captive birds may be quite simple, and this has some very useful applications. For example, dietary wariness could be deactivated in commercially reared poultry simply by feeding them from hatching on more than one food color, thereby making them less averse to the changes in diet which are needed during normal husbandry.

To conclude, the results presented here reveal for the first time the ease with which wariness toward food types can be manipulated by even brief and limited exposure to novelty and distastefulness. We have outlined some of the implications that these results may have for understanding foraging behavior of animals in the wild, the evolutionary pressures that they impose on their prey, and control of dietary wariness in captive birds.

Funding

Cardiff University to R. J. T.; Science Foundation Ireland.

We thank Mark Brown, John Skelhorn, and the anonymous referees for their very helpful comments on previous drafts of this article and Mike Speed for assistance in designing and carrying out one of the experiments reported here.

References

Baddeley
R
Osorio
D
Jones
CD
Colour generalisation by domestic chicks
Behav Brain Sci
 , 
2001
, vol. 
24
 pg. 
654
 
Barnett
SA
Experiments on neophobia in wild and laboratory rats
Br J Psychol
 , 
1958
, vol. 
49
 (pg. 
195
-
201
)
Blough
DS
Measuring the search image: expectation, detection, and recognition in pigeon visual search
J Exp Psychol Anim Behav Process
 , 
2002
, vol. 
28
 (pg. 
397
-
405
)
Brigham
AJ
Sibly
RM
Cowan
PD
Feare
CJ
A review of the phenomenon of neophobia
Advances in vertebrate pest management
 , 
1999
Furth (Germany)
Filander Verlag
(pg. 
67
-
84
)
Chen
B
Kang
L
Insect population differentiation in response to environmental thermal stress
Prog Nat Sci
 , 
2005
, vol. 
15
 (pg. 
289
-
296
)
Cooper
JB
Colored feed for turkey poults
Poult Sci
 , 
1971
, vol. 
50
 (pg. 
1892
-
1893
)
Dukas
R
Kamil
AC
Limited attention: the constraint underlying search image
Behav Ecol
 , 
2001
, vol. 
12
 (pg. 
192
-
199
)
Fuhrer
J
Agroecosystern responses to combinations of elevated CO2, ozone, and global climate change
Agric Ecosyst Environ
 , 
2003
, vol. 
97
 (pg. 
1
-
20
)
Jansson
L
Enquist
M
Receiver bias for colourful signals
Anim Behav
 , 
2003
, vol. 
66
 (pg. 
965
-
971
)
Jones
CD
Osorio
D
Baddeley
RJ
Colour categorization by domestic chicks
Proc R Soc Lond B Biol Sci
 , 
2001
, vol. 
268
 (pg. 
2077
-
2084
)
Jones
RB
Responses of domestic chicks to novel food as a function of sex, strain and previous experience
Behav Process
 , 
1986
, vol. 
12
 (pg. 
261
-
271
)
Kelly
DJ
Dietary conservatism in passerines: the influences of novel odour and novel colour [PhD dissertation]
 , 
2001
[Dublin (Ireland)]
University of Dublin
Kelly
DJ
Marples
NM
The effects of novel odour and colour cues on food acceptance by the zebra finch, Taeniopygia guttata
Anim Behav
 , 
2004
, vol. 
68
 (pg. 
1049
-
1054
)
Lindström
L
Alatalo
RV
Lyytinen
A
Mappes
J
Predator experience on cryptic prey affects the survival of conspicuous aposematic prey
Proc R Soc Lond B Biol Sci
 , 
2001
, vol. 
268
 (pg. 
357
-
361
)
Logan
JA
Regniere
J
Powell
JA
Assessing the impacts of global warming on forest pest dynamics
Front Ecol Environ
 , 
2003
, vol. 
1
 (pg. 
130
-
137
)
Lowndes
M
Stewart
MG
Dendritic spine density in the lobus parolfactorius of the domestic chick is increased 24-h after one-trial passive avoidance training
Brain Res
 , 
1994
, vol. 
654
 (pg. 
129
-
136
)
Mappes
J
Marples
N
Endler
JA
The complex business of survival by aposematism
Trends Ecol Evol
 , 
2005
, vol. 
20
 (pg. 
598
-
603
)
Marples
NM
Brakefield
PM
Genetic variation for the rate of recruitment of novel insect prey into the diet of a bird
Biol J Linn Soc
 , 
1995
, vol. 
55
 (pg. 
17
-
27
)
Marples
NM
Brakefield
PM
Cowie
RJ
Differences between the 7-spot and 2-spot ladybird beetles (Coccinellidae) in their toxic effects on a bird predator
Ecol Entomol
 , 
1989
, vol. 
14
 (pg. 
79
-
84
)
Marples
NM
Kelly
DJ
Neophobia and dietary conservatism: two distinct processes?
Evol Ecol
 , 
1999
, vol. 
13
 (pg. 
641
-
653
)
Marples
NM
Kelly
DJ
Thomas
RJ
Perspective: the evolution of warning coloration is not paradoxical
Evolution
 , 
2005
, vol. 
59
 (pg. 
933
-
940
)
Marples
NM
Roper
TJ
Effects of novel colour and smell on the response of naive chicks towards food and water
Anim Behav
 , 
1996
, vol. 
51
 (pg. 
1417
-
1424
)
Marples
NM
Roper
TJ
Response of domestic chicks to methyl anthranilate odour
Anim Behav
 , 
1997
, vol. 
53
 (pg. 
1263
-
1270
)
Marples
NM
Roper
TJ
Harper
DGC
Responses of wild birds to novel prey: evidence of dietary conservatism
Oikos
 , 
1998
, vol. 
83
 (pg. 
161
-
165
)
Martin
LB
Fitzgerald
L
A taste for novelty in invading house sparrows, Passer domesticus
Behav Ecol
 , 
2005
, vol. 
16
 (pg. 
702
-
707
)
Murphy
LB
Duncan
IJH
Attempts to modify responses of domestic fowl towards human beings. 1. Association of human contact with a food reward
Appl Anim Ethol
 , 
1977
, vol. 
3
 (pg. 
321
-
334
)
Poole
TB
The UFAW handbook on the care and management of laboratory animals
 , 
1999
Oxford
Blackwell Science
Rickard
NS
Poot
AC
Gibbs
ME
Ng
KT
Both non-NMDA and NMDA glutamate receptors are necessary for memory consolidation in the day-old chick
Behav Neural Biol
 , 
1994
, vol. 
62
 (pg. 
33
-
40
)
Ruxton
GD
Sherratt
TN
Speed
MP
Avoiding attack: the evolutionary ecology of crypsis and mimicry
 , 
2004
Oxford
Oxford University Press
Sanchez-Lafuente
AM
Valera
F
Godino
A
Muela
F
Natural and human-mediated factors in the recovery and subsequent expansion of the purple swamphen Porphyrio porphyrio L. (Rallidae) in the Iberian Peninsula
Biodivers Conserv
 , 
2001
, vol. 
10
 (pg. 
851
-
867
)
Schlenoff
DH
Novelty—a basis for generalization in prey selection
Anim Behav
 , 
1984
, vol. 
32
 (pg. 
919
-
921
)
Simmons
RE
Barnard
P
Dean
WRJ
Midgley
GF
Thuiller
W
Hughes
G
Climate change and birds: perspectives and prospects from southern Africa
Ostrich
 , 
2004
, vol. 
75
 (pg. 
295
-
308
)
Skelhorn
J
Rowe
C
Tasting the difference: do multiple defence chemicals interact in Müllerian mimicry?
Proc R Soc Lond B Biol Sci
 , 
2005
, vol. 
272
 (pg. 
339
-
345
)
Suarez-Seoane
S
Osborne
PE
Rosema
A
Can climate data from METEOSAT improve wildlife distribution models?
Ecography
 , 
2004
, vol. 
27
 (pg. 
629
-
636
)
Thomas
RJ
Bartlett
L
Marples
NM
Kelly
DJ
Cuthill
IC
Prey selection by wild birds can allow novel and conspicuous colour morphs to spread in prey populations
Oikos
 , 
2004
, vol. 
106
 (pg. 
285
-
294
)
Thomas
RJ
Marples
NM
Cuthill
IC
Takahashi
M
Gibson
EA
Dietary conservatism may facilitate the initial evolution of aposematism
Oikos
 , 
2003
, vol. 
101
 (pg. 
458
-
466
)
Wauters
AM
Richard-Yris
MA
Talec
N
Maternal influences on feeding and general activity in domestic chicks
Ethology
 , 
2002
, vol. 
108
 (pg. 
529
-
540
)
Weiss
SJ
Weissman
RD
Generalization peak shift for autoshaped and operant key pecks
J Exp Anal Behav
 , 
1992
, vol. 
57
 (pg. 
127
-
143
)