Evaluating the impact of age, sex, and lameness on the problem-solving ability of broiler chickens

Cognition encompasses the mental processes involved in perceiving, processing, storing, and acting on information from the environment (Shettleworth, 2001). This includes a wide range of cognitive skills such as sensory perception memory, learning, decision making, reasoning, and problem solving (Dukas, 2004).

The cognitive abilities of chicken are already highly developed in early life, with research demonstrating that the cognitive abilities of chicks include object permanence (Regolin et al., 1995; Vallortigara et al., 1998), recognizing partially occluded objects (Regolin and Vallortigara, 1995), episodic memory (Vallortigara et al., 1998), transitive inference (Daisley et al., 2010), and numerical cognition (Rugani, 2018). In older birds, particularly in laying hens, research has demonstrated the presence of many different cognition skills (reviewed in Marino, 2017; Garnham and Løvlie, 2018), including several studies that have shown the problem-solving abilities of poultry (Schmelz and Krause, 2021; Wiles et al., 2023; Hillemacher and Tiemann, 2024).

Problem-solving abilities are an important part of an individual’s life, which help in finding solutions to the challenges that might be faced during their lifetime, and allowing birds to cope and adapt to their environment and to environmental changes more effectively (Franks, 2018). Problem solving requires cognitive skills to integrate the information gained in order to solve an issue (Rowell et al., 2021). With innovative problem solving, an animal uses a new behavior or adapts an existing behavior to solve an unfamiliar problem. Cognitive abilities, including problem-solving abilities, can be impacted by several different factors such as chronic stress (Mendl, 1999; Lormant et al., 2020), social instability (Arnould et al., 2024), undernutrition (Ferreira et al., 2021), environmental complexity (Tahamtani et al., 2015), genetics (Gupta et al., 2017), and pain (Khera and Rangasamy, 2021).

Lameness and walking impairment have been major welfare concerns in broiler chickens over the past 30 years and continue to do so today (Kestin et al., 1992; Bradshaw et al., 2002). Research reports that 77%–99% of broilers display some degree of gait defect (a gait score above 0) within commercial settings (Knowles et al., 2008; Tahamtani et al., 2018; Granquist et al., 2019) and that 5.4%–27.6% are considered lame (a gait score ≥3) (Knowles et al., 2008; Bassler et al., 2013; Kittelsen et al., 2017; Louton et al., 2018; Tahamtani et al., 2018; Granquist et al., 2019). Lameness and walking impairments can result from several different disorders that can be infectious, developmental, and degenerative (Bradshaw et al., 2002). The growth rate (Kestin et al., 2001; Riber and Wurtz, 2024), the genotype (Kestin et al., 2001; Riber and Wurtz, 2024), and a high body weight (Kestin et al., 2001) are key aetiological factors for leg disorders and associated lameness. However, other factors have also been found to impact the development of lameness (reviewed in Bradshaw et al., 2002; Liu et al., 2023). Age and sex have also been found to influence lameness, with incidences of lameness increasing with age (Kestin et al., 2001) and lameness being more common in males (Sørensen et al., 2000). Lameness reduces mobility, hindering a bird’s ability to access resources such as feed and water, can restrict the performance of desired behaviors, and prevent birds from escaping antagonistic interactions (Weeks et al., 2000; Bradshaw et al., 2002; Butterworth et al., 2002). Furthermore, lameness, when the leg health issues inhibit or hinder the bird’s agility and ability to maneuver around, is often painful, as evidenced by lame birds self-administering analgesics (Danbury et al., 2000). Analgesic treatment increases mobility, as well as the time spent standing, and improves gait in lame broilers (McGeown et al., 1999; Caplen et al., 2013; Hothersall et al., 2016). Pain can impact the attention, motivation, memory, sensory processing, executive function, and decision making, all of which are important components of problem solving (Khera and Rangasamy, 2021). Therefore, the pain associated with lameness has the potential to impact problem solving. Not all forms of lameness have been proven to be painful (i.e., mechanical deformities). However, other aspects associated with lameness, such as stress or feeling unwell, have also been shown to impact cognitive abilities (Lindqvist and Jensen, 2009). To understand whether lameness impacts cognition, specifically innovative problem solving, and the role that age and sex play in this, we performed this study with the objective of evaluating how age, sex, and lameness affect the innovative problem-solving abilities of broiler chickens.

A total of 88 broiler chickens (Ross 708), 44 females (F) and 44 males (M), were sourced from a commercial hatchery and reared until 56 days of age (x¯ = 4.71 kg). The birds were randomly distributed into four sex-separated pens of 2.32 m2 (two female pens and two male pens) at a density of 22 birds/pen. Each pen contained fresh wood shavings as bedding, an automatic nipple drinker water line, and a commercial-type tube feeder. Supplemental feeders and drinkers were provided from placement until day 7. Birds had ad libitum access to feed and water throughout the trial. All birds were fed ad libitum the same standard corn/soybean-based diet consisting of a four-stage feeding program. Starter feed (crumble) was provided between days 0 and 14, grower (pellets) was provided from day 14 to day 28, and finisher 1 and finisher 2 (pellets) were given from 28 to 42 days and from 42 to 56 days, respectively. The temperature was set at 32°C at placement and gradually decreased to a final set point of 18°C by the end of the trial, with relative humidity at 60%–70%. Lighting was provided at 40 lx for 23 h of light and 1 h of dark initially, and then gradually decreased to 18 h of light and 6 h of dark, at 10 lx, for the remainder of the experiment. Litter quality was scored on a scale from 0 to 4 (previously described in Welfare Quality, 2009) on each of the experimental days (the litter quality scores were 0.5, 2, and 2.75 on days 14, 28, and 42, respectively).

To assess the problem-solving abilities of broilers, the birds were presented with a problem-solving task at three different ages: 14, 28, and 42 days. Three different tasks were used, which were rotated across birds and ages so that each bird encountered each task only once, and each task was presented equally across the age groups. This was done to prevent any bird from performing the same task more than once and to account for potential differences in the task difficulty by distributing them evenly across ages. The three problem-solving tasks were a stringpull test, a cup test, and a wire removal test (Figure 1). These were modified from tasks used to test the cognitive abilities of other avian species described by Griffin and Diquelou (2015) and Barrett et al. (2022). Task A was the string-pull test, which consisted of a clear cylinder with a string attached to a food reward. Birds had to pull the string to reach the food. Task B was the wire removal test, which consisted of a clear box with food inside and wire bars across it. The wires had to be removed to access the food. Task C was the cup test, which consisted of three clear cups with lids, with one of the cups containing a food reward. Birds had to remove the correct lid to access the feed. The feed reward consisted of their standard feed. All tasks were designed so that birds could see the feed, but not peck directly at the feed, ensuring that they had to use different physical skills, but the same cognitive mechanisms, to solve the task. The birds were placed in the pen (1.2 m × 1.2 m) facing the task, approximately 50 cm from the task, and were given 5 min to complete the task. After completion or after 5 min elapsed, the birds were returned to their home pen. After all of the birds had been individually tested, they were put into groups and presented the problem-solving task again. A total of 12 groups were evaluated, six groups offemales and six groups of males. To standardize the conspecifics for group testing, each bird was consistently placed in the same group for all three problem-solving tests throughout the duration of the study. After completing the problem-solving tasks, the birds were given a minimum of 10 min to rest, after which the gait score (GS) was determined.

To assess neophobia and the problem-solving ability, the latency to the birds’ first movement (initial step) after placement, the latency to the birds’ first interaction with the tasks, the latency from the start of the test to the completion of the task (duration), the number of attempts performed to solve the task, and the success rate at solving the task were recorded. First interaction was defined as the first time the bird intentionally touched the task with their beak or feet. For birds that did not interact with the task, the maximum score was assigned. An attempt was defined as each individual time the bird pecked at or manipulated the task. The task was considered successfully completed if the bird accessed the food and ingested it. Success rate was measured on a binary scale, with 0 being unsuccessful and 1 being successful. In the group test, only success was measured, and the bird that solved the task was identified by the observer, and the bird ID number was recorded.

The GS was used to evaluate the birds’ walking ability. Birds were allowed to walk freely for 1.5 m and then were scored by two trained assessors. They were scored on a six-point scale, with a score of 0 indicating a normal gait and a score of 5 indicating complete lameness and inability to walk or support weight on their legs. A gait score of 0 (GS 0) describes a bird walking normally with no detectable abnormality; GS 1 birds had a slight defect, which was difficult to define precisely; GS 2 birds had a definite and identifiable defect; GS 3 birds had an obvious gait defect, which affected their ability to move; GS 4 birds had a severe gait effect, walking with difficulty and when driven or strongly motivated; and GS 5 birds were incapable of walking. The GS was measured on days 14, 28, and 42. Foot pad dermatitis (FPD) was scored on a five-point scale, previously described in Welfare Quality (2009). However, the FPD scores were low (with average FPD scores of 0.33, 0.24, and 0.25 for 14, 28, and 42 days, respectively) and were therefore excluded from further analysis.

Birds that did not undergo all three problem-solving tests were removed from the experiment; therefore, a total of 75 birds (39 F and 36 M) were included in the statistical analysis. Individual task data were analyzed using a two-way ANOVA for the fixed effects of age, gait, sex, age*sex, and gait*sex, with task as a covariate (PROC GLIMMIX; SAS 9.4). The Cochran–Mantel–Haenszel (CMH) test (PROC FREQ) was used to examine the consistency in problem solving across ages. Pathway analysis (PROC CALIS) was performed to understand the influence of the dependent variables on the success rate. Group task data were analyzed using a one-way ANOVA for the fixed effects of age, sex, and group/individual, with task as a covariate (PROC GLIMMIX; SAS 9.4). Mean separation was performed using Tukey–Kramer. McNemar’s test (PROC FREQ) was used to analyze whether birds that were successful in groups were also successful individually. Statistical significance was considered at p ≤ 0.05.

The results showed that the GS increased as birds aged. At 14 days, 76% of the birds had a GS of 0 and 24% had a GS of 1. At 28 days, 40% of the birds had a GS of 0, 40% had a GS of 1, and 20% had a GS of 2. Finally, at 42 days, 27% of the birds had a GS of 0, 24% had a GS of 1, 37% had a GS of 2, and 12% had a GS of 3(Figure 2).

No interaction effects were found for gait and sex or for gait and age for any of the variables measured (Table 1). Age and sex had an interactive effect on the latency to first movement, with males on day 14 (6.9 s) and females on day 28 (11.1 s) having a longer latency to first movement than males on day 28 (3.2 s) and both sexes on day 42 (2.4 and 2.6 s for females and males, respectively), but did not differ from females on day 14 (3.6 s, p < 0.01). Females on day 14 also took longer than birds on day 42, regardless of sex. Birds on day 28 (7.3 s) and day 14 (5.2 s) took longer to make their first movement compared with birds on day 42 (2.5 s, p < 0.01). The latency to first movement was not significantly impacted by the GS.

Age impacted the latency to first interaction with the task. It took longer for birds on day 14 (178.2 s) and day 28 (147.2 s) compared with birds on day 42 (91.4 s, p < 0.01) (Table 1). There was also a sex effect, with males (110.8 s) being quicker to interact with the tasks than females (164.3 s, p = 0.01). Gait had no effect on the latency to first interaction.

The total duration to solve the tasks showed an effect of age, with birds on day 42 (165.7 s) taking less time to complete the problem-solving task compared with birds on day 14 (213.2 s, p < 0.01) (Table 1). Sex also impacted the total duration to solve the tasks, with males (177.5 s) solving the tasks faster than females (205.2 s, p < 0.01). The GS affected the time taken to solve the tasks, with birds with a GS of 3 (216.2 s) taking longer to solve the tasks compared with birds with GS 0 (183.3 s), GS 1 (201.5 s), and GS 2 (194.9 s, p = 0.03).

There was an interactive effect of age and sex on the number of attempts the birds took to solve the tasks, with males on day 42 (8.8) attempting the task more times compared with females on day 42 (2.5) and males on day 14 (2.4, p < 0.01) (Table 1). No individual effects of sex, age, or gait on the number of attempts were found.

Success also differed with age, with birds on day 42 (0.54) being more successful at solving the tasks than birds on day 14 (0.31, p < 0.01) (Table 1). There was also an effect of sex on the success rate, with males (0.47) having a better success rate compared with females (0.37, p = 0.02). No effect of gait on the success rate was observed.

The CMH test demonstrated that birds showed consistency in problem-solving success across ages. The test for nonzero correlation was significant [c2 (1) = 8.50, p < 0.01], meaning that birds successful at one age had an increased likelihood of succeeding at another age. The overall success rate differed between ages, with more birds being successful at completing the task at older ages [c2 (2) = 8.59, p = 0.01]. The test for general association was significant, indicating that the success across age groups had a non-random pattern [c2 (3) = 8.50, p < 0.01].

The pathway analysis for gait, task duration, and number of attempts (Table 2) showed that the task duration had a strong negative effect on problem solving (b = −0.94, p < 0.01), while both the number of attempts and the GS had a small positive effect on the success rate (b = 0.05, p = 0.03, and b = 0.05, p = 0.02, respectively). This indicates that birds who took longer to solve the task were less likely to succeed, but that those that made more attempts were more likely to succeed. Moreover, birds with better GS were more likely to successfully complete the task. The number of attempts had no effect on the duration, and gait had no effect on either the number of attempts or the task duration.

The pathway analysis conducted to understand how the latency to first movement and the latency to interaction with the task predicted problem-solving success (Table 3) showed that both movement and interaction had a positive effect on success (b = 0.14, p = 0.01, and b = 1.02, p < 0.01, respectively). Birds that took longer to start moving and to interact with the task were more likely to complete the task. Latency to first movement had a negative effect on latency to interaction (b = −0.29, p < 0.01), indicating that birds that were slower to make the first move interacted with the task more quickly (Figure 3).

Differences were found between birds tested in groups and individually, with birds presented the tasks in groups (0.89 ± 0.053) being more successful than when presented the tasks individually (0.69 ± 0.078, p = 0.02). No effects of age or sex were found when birds were tested in groups. On day 14, 37.5% of the birds who solved the task in a group also solved it individually, while on days 28 and 42, respectively 62.5% and 66.7% of birds that previously solved the tasks individually also solved the tasks in groups.

The paired success outcomes of birds tested individually and in groups revealed that birds were significantly more likely to solve the task when tested in groups compared with individually (c2 = 11.00, p < 0.01), with a fair agreement (k = 0.26, 95%CI = 0.03–0.49). This indicates that there was some consistency in the problem solving (56% were successful both in a group and individually), but also a meaningful change (40% were successful in a group, but not individually). The likelihood of success on the task increased when the birds were in a group, indicating a social effect.

The problem-solving abilities were impacted by age, with older birds being faster and more successful at solving the tasks, and by sex, with males being faster and more successful, but not by the GS. The ability to solve the task was also consistent across ages, with birds who solved the task at one age also able to solve it at the later ages. The presence of conspecifics improved the problem-solving ability, as birds that were unable to solve the task individually were able to do so in groups.

Age had a clear impact on the problem-solving ability, with birds being both quicker and more likely to solve the task as they aged. Previous research on cognition and reasoning ability in broilers also found improved abilities with age (Panigrahy et al., 2017; Gupta et al., 2017). This may be due to spatial reasoning abilities improving with age (Gupta et al., 2017) or the birds using previous experiences to help solve the task. Laying hens in a double bisection task were found to use past experiences to navigate novel problems (Wiles et al., 2023). In this study, the tasks presented to the birds were different each time and required a different strategy to solve them. However, the birds may have been able to use some knowledge from previous tasks to solve the novel tasks, thereby increasing their speed and success. One limitation of the study is that, although the task was new for each test, the testing setup was the same for all three rounds of testing. This means that the birds may have become habituated to the testing situation and the handling associated with the testing, as previously described by Jackson et al. (2025), reducing some of the neophobia associated with the situation, which partially influenced the decrease in the time taken for the birds to respond. Furthermore, as the birds aged, they may have habituated to the handling. In future research, habituation to the testing situation may help reduce the possible influence of neophobia on the test outcomes.

During the test, at the younger ages, the birds vocalized and looked for conspecifics more than during the later ages. Both the vocalizations and the vigilance behavior during social isolation can be indicative of fear or a motivation for social reinstation (Forkman et al., 2007). In addition, at the older ages, the birds also spent more time sitting and preening, behaviors indicating comfort. The reduction in fear behaviors and the increase in comfort behaviors may indicate that, at the younger age, birds may have feared the novel environment of the testing situation, which may have affected their latency to first movement, their first interaction with the tasks, and their success rate. Particularly, fear has been shown to increase the latency to interact with novel things (Forkman et al., 2007) and to impact the cognitive abilities of poultry (de Haas et al., 2017). Subsequently, after the initial test, the birds may have been habituated to the situation and have been able to focus more on the task. Young animals may show an innate fear reaction to sudden disturbances in the environment, but rapidly become habituated to them (Steimer, 2002).

Sex also appears to have an impact on problem solving, as the males demonstrated a higher success rate, moved faster, and interacted with the task less while attempting to solve the task. This supports previous research on cognitive abilities with Vanaraja chickens, a dual-purpose breed, which showed males to also be faster in solving a Y-maze (Panigrahy et al., 2017). Their increased ability to solve the tasks could be attributed to the fact that males are more efficient at processing spatial information (Panigrahy et al., 2017). Previous studies have reported that males have shown faster latencies of movement than females when presented a food reward (Bokkers and Koene, 2002). Thus, motivation may have also been a factor in male speed and success. Males are more motivated to feed, while females are more socially motivated (Howlider and Rose, 1992; Vallortigara et al., 1998).

The GS impacted the duration, but not the success rate to solve the tasks. Birds with higher GS took longer to solve the tasks, but this did not impact how well they solved them. This indicated that the lameness or gait defects impacted the time needed to solve the task, but that the birds eventually managed to do so. It could be possible that the pain, or stress, associated with lameness impacted the bird’s attention and processing speed, as pain and stress have been shown to impact both attention and processing ability in other species (Low, 2013; Khera and Rangasamy, 2021). However, the motivation of the food reward was high enough that it did not impact task solving. To fully understand this, more research looking into the different aspects of cognition and how they are impacted by pain in chickens is needed. Overall, there did not appear to be an impact of the GS on the birds’ cognitive abilities. However, caution should be taken when interpreting these results as, within our trial, the highest GS was 3, which only occurred on day 42 and was only obtained by a very low number of birds (10 birds in total). With the highest GS only occurring at the oldest age, the improved cognitive ability with age may have masked the impact of GS on problem solving. Furthermore, the distance the birds had to walk to reach the task, i.e., < 1 m, may not have been sufficient to elicit a pain response in birds with only moderate lameness (GS of 3). Future research that uses a larger number of birds and the full range of GS should be performed to fully understand how lameness impacts cognitive abilities.

Success at problem solving at one age increased the chances of a bird being successful at the next age, meaning that certain birds were consistently able to solve the problem-solving task across different ages. In addition, the pathway analysis also showed that birds that took longer to move and interact with the task were more likely to solve the task. This suggests that birds that took the time to evaluate the situation before acting were more likely to solve the task. Similarly, birds who took more attempts to solve the task were more likely to be successful, indicating that persistence is also important in task success. This highlights the importance of individual strategies, cognitive abilities, interest, and motivation in bird success. The authors also noted a clear difference between individuals in the birds’ initial response to the task and that this initial response impacted how well the birds could solve the task. Some birds were more interested in conspecifics and reuniting with them, often responding immediately by vocalizing and walking around the pen to find the conspecifics and ignoring the task, while other birds took more time to respond and showed more interest in the task. Individual differences in personality are likely to have an impact on task performance. Previous research has shown that personality can impact cognitive task success in broilers, with a study that examined free-ranging broilers showing that low-ranging individuals did better in the test than high-ranging individuals (Ferreira et al., 2021). The impact of personality on problem-solving tasks and its consistency across time was also observed in other avian species, such as zebra finches (Barrett et al., 2022) and red jungle fowl (Zidar et al., 2018).

Birds had a higher success rate in groups overall, and of the birds that solved the task in the group, 40% were not able to solve the task during the individual test. The increase in the likelihood of solving the task when other birds were present points to social buffering, where the presence of conspecifics reduced the stressfulness of the situation (Edgar et al., 2015), allowing the birds to focus more on the task, thereby increasing the probability of success. Social facilitation may have played a role, as it was observed that one bird’s interaction with or interest in the task led to other birds also showing interest in or interacting with the task. Other behaviors related those involved with task solving, such as foraging and eating, have been found to be socially facilitated in chickens (Collins and Sumpter, 2007; Ogura and Matsushima, 2011). Further studies to investigate exactly how the presence of conspecifics impacts cognitive abilities are needed to fully elucidate this effect. Particularly, as within commercial settings broilers are housed in groups, conspecifics are likely to be present when facing a situation that requires problem solving. It is also important to take this effect into account in future studies on broiler cognition.

Overall, innovative problem-solving skills are already present in broilers at 14 days, and these skills increase with age. There is a difference between sexes in problem-solving skills for problems that require some spatial skills, with males responding more quickly and more often, as well as being more successful at solving the task. Individual differences in bird approach, motivation, and interest appear to have played a large role in whether birds were successful at solving the task, with birds that are more persistent and birds that take time to assess the situation being better at solving the problems. Lameness did not have a clear impact on problem-solving ability; however, this may be due to the absence of severe lameness within the birds tested. The presence of conspecifics increased task success, which may have been due to the presence of other birds providing a form of social buffering.