Tracheal collapse rates in Kenyan working cart and pack donkeys

Abstract

Background

Donkeys (Equus asinus) in East Africa are recognised as working with poorly designed harnesses and carts. The donkey cart used in regions of Africa, specifically Meru County, Kenya, appears to place the cart's weight solely on the donkey's mid‐cervical region. The ventral area of the neck is vulnerable to external pressure on the trachea, which is superficially located.

Objectives

To compare the presence and severity of endoscopic abnormalities in the upper airway and trachea of Meru County working cart donkeys compared to pack donkeys. To determine associations between endoscopic and physical evidence of harness injury.

Study Design

Cross‐sectional.

Methods

At voluntary health clinics, physical examination and endoscopy of the upper airway and trachea were evaluated and recorded at rest on 66 donkeys (38 cart donkeys and 28 pack donkeys). The presence and grade of tracheal collapse, tracheal ring abnormalities, tracheal mucus, and pharyngitis were subsequently determined on the endoscopy recordings by blinded evaluators.

Results

Significant associations were identified between work type and tracheal collapse severity, ring abnormalities, pharyngitis, hairless skin, and thickened skin, with these findings more common in cart than pack donkeys (all p < 0.03). Moderate tracheal collapse (grade 3) was present in 40% (95% CI: 24%–57%) of cart and 0% (0%–12%) of pack donkeys. Intra‐tracheal ring abnormalities were noted in 45% (29%–62%) of cart and 0% (0%–12%) of pack donkeys. Physical examination findings were significantly and positively associated with endoscopy findings (all p < 0.002).

Main Limitations

Standing endoscopy was performed at rest and dynamic endoscopy during the pulling of carts was not performed.

Conclusion

Moderate tracheal collapse was seen in cart donkeys when compared to pack donkeys. Physical examination findings at rest were associated with and moderately predictive of tracheal collapse.

1. INTRODUCTION

Recognised welfare issues for donkeys in developing countries include excessive workloads, long working hours, improper husbandry, and poor handling. ¹ An estimated 1.9 million donkeys in Kenya are used to transport water for household use and pull carts loaded with goods. ² The welfare of these and other working equids is often overlooked, despite the vast impact their work has on an owner's quality of life and financial situation. ² , ³

Harnessing methods in East Africa have been identified as a welfare concern for draught donkeys due to improper fit and function and the resulting impact this has on their musculoskeletal system. ¹ , ⁴ The cart used in Meru County, Kenya has a single central shaft with a crossbeam attached solely to the donkey's proximal to middle neck. ⁴ The current cart design does not consider recommendations to have a clear gap in breast collars designed for donkeys to prevent pressure on the trachea. ⁵ Pressure from carts and harnessing methods may result in trauma to the structures located under poorly fitting equipment. ⁶ The weight of loaded carts varies but generally ranges between 250 and 700 kg and is often pulled in steeply sloping terrain. ⁴ The force from pulling the cart is placed entirely on the donkey's ventral neck where the trachea is superficially located. ⁷ The reported workload for donkeys in this region is 1–4 h a day, 2–5 days/week. ⁴ This potential impairment of the respiratory system due to poor harnessing may have a negative impact not only on the donkeys' ability to work, but also on their health and welfare. ⁸

Tracheal collapse is a known condition of donkeys caused by factors such as external trauma, primary degeneration of tracheal cartilage, or secondary to respiratory diseases that increase respiratory effort. ⁹ An ongoing cough with inspiratory and expiratory dyspnoea is seen in chronic cases of tracheal collapse. Airway turbulence and impedance can be significantly increased with tracheal collapse, leading to decreased athletic performance and increased work necessary to breathe. ⁵ Donkeys, when compared to horses, are very stoic animals with a less sensitive cough reflex, making early detection of respiratory disease difficult. ¹⁰

Standing lateral radiographs and/or endoscopy provide a definitive diagnosis of the condition, with endoscopy able to identify subtle or chronic disease not always visible on radiography. ¹¹ Flattening of the cartilaginous tracheal rings, as seen with tracheal collapse, may also be observed as an ageing change in donkeys with varying impacts on their respiratory health. ¹² While uncommon in horses, tracheal collapse has been reported to occur in donkeys, ponies, and American Miniature Horses. ¹³ Based on post‐mortem examinations of donkeys with a substantially different workload than the donkeys in this study, aged, non‐working donkeys in the United Kingdom reported a prevalence of tracheal collapse of 8.7%. ¹⁴

This study aimed to investigate tracheal injury in working donkeys and identify physical examination findings that would help guide harness adjustments to improve their overall health and welfare. To achieve this, the study had two main objectives. The first objective was to investigate the presence of upper airway injury and tracheal collapse attributable to current methods of harnessing using a sample of pack donkeys without harnesses as a local comparison. We hypothesised that there would be an increased proportion of upper respiratory tract disease, visualised endoscopically and detected through physical examination findings in working cart donkeys in Meru County, Kenya when compared to working pack donkeys. Our second objective was to investigate associations between physical examination findings and endoscopic findings in cart donkeys, and we hypothesised that one or more physical examination findings would be indicative of tracheal pathology.

2. MATERIALS AND METHODS

A cross‐sectional, partially blinded study was performed on a convenience sample of cart and pack donkeys enrolled during voluntary health clinics in Meru County, Kenya in January 2024. The pack donkeys served as a comparison for the cart donkeys (Figure 1A), as the pack donkeys did not have a harness across the neck (Figure 1B). Sample size was based on the most conservative estimates of preliminary data gathered by the principal investigator. A minimal sample size of 32 donkeys per group was needed to detect a statistical difference in the presence of clinically relevant tracheal compression (estimated prevalences Cart Donkeys = 40%, Pack Donkeys = 10%, Alpha = 0.05, Power = 0.8, likelihood‐ratio test for two‐sample proportions).

FIGURE 1.

Examples of harness and pack equipment in Meru County, Kenya. Commonly used harnessing to pull carts (A) and working pack donkey used for hauling water (B). Note the lack of equipment on the neck of pack donkeys.

Inclusion/exclusion criteria are presented in Figure 2. As an incentive to participate in the study, vaccination for tetanus, rabies, and anthelminthic treatment was administered to donkeys, as well as correction of dental and hoof disorders when indicated. Each community was notified verbally or via text message of the date and location of the donkey health clinics where the endoscopies and physical examinations were performed. Fractious donkeys that would not tolerate twitching or venipuncture were excluded. Donkeys less than an estimated 1 year of age, and donkeys that both pulled carts and packed water were excluded from the study. Endoscopic examinations that were not diagnostic due to obstruction of the lens with mucous or lacking visualisation of the trachea were discarded.

FIGURE 2.

Flow chart of inclusion/exclusion criteria of working cart and pack donkeys in Meru County, Kenya.

At the health clinics, all donkeys were processed on a first‐come basis until the clinic closed. Thoracic auscultation, heart rate, respiratory rate, rectal temperature, age according to teeth, and physical evidence of harnessing injury such as thickened skin (Figure 3A,B) and hairless skin on the ventral neck were evaluated and recorded by the primary investigator (Martha Mellish). No difference in the hair growth on the ventral neck when compared to surrounding skin was graded as no skin lesions. Lack of hair was classified as areas of hairless skin. Thickened skin on the ventral neck was graded using a scale of none (no palpable or visual thickening), mild (palpable increase, minimal visual thickening), moderate (excessive skin visual and palpable with skin hanging ventrally up to 3 cm), and severe (excessive skin visual and palpable with skin hanging ventrally greater than 3 cm).

FIGURE 3.

Moderate skin thickening (A) and severe skin thickening (B) on ventral neck of cart donkeys in Meru County, Kenya.

After physical examination findings were recorded, physical restraint including a nose twitch, halter and chemical restraint with 0.013–0.026 mg/kg detomidine HCl, once, IV (Dormosedan, Zoetis, Canada) was used to perform the endoscopy. The upper airway and trachea to a depth of 40–60 cm from the nares (Figure 4) were evaluated with video endoscopy and recorded by a single individual (Martha Mellish). A 9 mm outer diameter, 1 m in length, modified boroscope (Articulating Borescope, Teslong, TD500) was used for the endoscopic evaluation.

FIGURE 4.

Demonstration of the position of the endoscope in the trachea at a depth of 60 cm from nares. The dark grey patch of skin on the donkey's neck indicates where the harness contacts the neck. The endoscopic examination would extend to the location where the harness is positioned.

Two blinded observers (blinded to donkey work type and physical examination findings) experienced in equine upper airway endoscopy (Yvonne Elce and Jennifer Burns; Diplomates of ACVS and ABVP) evaluated the videos and graded tracheal collapse, tracheal ring abnormalities, and tracheal mucous using a grading rubric based on published scoring methods. ¹¹ , ¹⁵ , ¹⁶ Tracheal collapse was graded from 0 (normal) to 4 (severe). See Figure 5A–D for examples of tracheal collapse grades 0 to 3 found in this study. Definitions of tracheal collapse were as follows: Grade 0 (normal): round trachea with no protrusion of trachealis muscle. Grade 1 (very mild): <25% reduction of dorsoventral tracheal lumen, slightly pendulous trachealis muscle. Grade 2 (mild): 26%–50% reduction of dorsoventral tracheal lumen, widened and pendulous trachealis muscle. Grade 3 (moderate): 51%–75% reduction of dorsoventral tracheal lumen, trachealis muscle almost in contact with dorsal surface of tracheal cartilage. Grade 4 (severe): >76% reduction of dorsoventral tracheal lumen, trachealis muscle almost in contact with dorsal surface of tracheal cartilage. Tracheal mucous was graded as none (clean, singular drop), little mucous (multiple small drops), moderate mucous (larger drops) or marked mucous (confluent or stream‐forming drops). Tracheal ring abnormalities were determined by the presence or absence of mucosa‐covered protrusions of tissue into the lumen of the trachea. ¹⁶ Pharyngitis was determined by the presence or absence of lymphoid follicles covering the visible surface of the pharyngeal mucosa. All donkeys were graded by both observers.

FIGURE 5.

Grade 0 (normal) Tracheal collapse score: Round trachea with no protrusion of trachealis muscle (A). Grade 1 (very mild) Tracheal Collapse: < 25% reduction of dorsoventral tracheal lumen, slightly pendulous trachealis muscle (B). Grade 2 (mild) Tracheal Collapse: 26–50% reduction of dorsoventral tracheal lumen, widened and pendulous trachealis muscle (C). Grade 3 (moderate) Tracheal Collapse Score: 51–75% reduction of dorsoventral tracheal lumen, trachealis muscle almost in contact with the dorsal surface of tracheal cartilage. Tracheal ring abnormalities are also present in this image (D).

2.1. Data analysis

Data was entered into an Excel database (Microsoft Corp) and imported into statistical software for analysis (Stata version SE.17.0, Stata Corp). Descriptive statistics were computed for all variables and stratified by work type (cart, pack). The Shapiro–Wilk test was used to determine the normality of continuous variables; as none were determined to be normally distributed, median and range were reported for continuous variables. Interobserver agreement for each of the endoscopic findings was calculated (percent agreement, kappa [pharyngitis, tracheal ring anomaly], weighted kappa [using software‐prerecorded weights (w); tracheal mucous, tracheal collapse]) and assessed using the following interpretations: below 0.0 (poor), 0.00–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial), 0.81–1.00 (almost perfect). ¹⁷ Associations between the type of work and categorical variables and between physical examination findings (skin thickening, hairless skin) and endoscopy findings (tracheal collapse grade, ring abnormality) were assessed using the Pearson χ ² test or, if any expected cell value was ≤1 or 20% or more expected cell values were ≤5, Fisher's exact test. Associations between the type of work and continuous variables were assessed with the two‐sample Mann–Whitney test. The Cochran–Armitage test for trend was used to assess the trend in proportions across ordered tracheal collapse grades for physical examination findings (skin thickening, hairless skin). Due to small cell frequencies, skin thickening scores were combined (none/mild, moderate/severe) for analysing associations between skin thickening and tracheal collapse grade and tracheal ring abnormalities. Binomial exact 95% confidence intervals were calculated for select variables. Statistical significance was set at p < 0.05.

3. RESULTS

The voluntary health clinics were attended by 270 donkeys, and 80 were examined endoscopically, and videos of 66 donkeys (38 cart donkeys and 28 pack donkeys) could be sufficiently evaluated for inclusion. Of the 14 donkeys for which videos could not be evaluated, five were excluded due to mucous obscuring the video, and nine were excluded as they were not scoped to a sufficient depth (40–60 cm from the nares) due to difficulty entering the trachea. Interobserver agreement for the parameters was fair to substantial (pharyngitis: 66.2% agreement, κ = 0.36, p < 0.001; tracheal ring anomaly: 74.3% agreement, κ = 0.47, p < 0.001; tracheal mucous: 90.7% agreement, κ = 0.58, p < 0.001; tracheal collapse: 88.7% agreement, κ = 0.61, p < 0.001). For all disagreements, re‐evaluation of the video by the two observers resulted in an agreed consensus score.

Signalment and findings on physical and endoscopic examination are presented in Table 1. Cart donkeys were exclusively male and had significantly more thickened skin and hairless skin (both p < 0.001) on the ventral neck, in the area where the harness sits. No significant difference was noted between the cart and pack donkeys in mucous scores, age, or health parameters at rest (respiratory rate, heart rate, rectal temperature, thoracic auscultation).

TABLE 1.

Comparison of cart and pack donkeys' endoscopic and physical examination findings, Meru County, Kenya.

		Count (%) unless otherwise noted			p a
Parameter		All donkeys	Cart donkeys	Pack donkeys
Tracheal collapse (grade)		N = 66	N = 38	N = 28	<0.001 b
Normal	0	1 (2)	0	1 (4)
Very mild	1	22 (33)	8 (21)	14 (50)
Mild	2	28 (42)	15 (40)	13 (46)
Moderate	3	15 (23)	15 (40)	0
Severe	4	0	0	0
Ring abnormalities		N = 66	N = 38	N = 28	<0.001 c
	Yes	17 (26)	17 (45)	0
	No	49 (74)	21 (55)	28 (100)
Tracheal mucous (score)		N = 64	N = 37	N = 27	0.5 b
None	0	18 (28)	12 (32)	6 (22)
Little mucous	1	34 (53)	18 (49)	16 (59)
Moderate mucous	2	11 (17)	7 (19)	4 (15)
Marked mucous	3	1 (2)	0	1 (4)
Pharyngitis		N = 61	N = 34	N = 27	0.03 c
	Yes	12 (20)	10 (29)	2 (7)
	No	49 (80)	24 (71)	25 (93)
Age (years)		N = 66	N = 38	N = 28	0.3 c
	<5	13 (20)	6 (16)	7 (25)
	5–15	38 (58)	21 (55)	17 (61)
	>15	15 (23)	11 (29)	4 (14)
Sex		N = 66	N = 38	N = 28	<0.001 c
	Jack	51 (77)	38 (100)	13 (46)
	Jenny	15 (23)	0	15 (54)
Hairless skin		N = 66	N = 38	N = 28	<0.001 c
	No	41 (62)	14 (37)	27 (96)
	Yes	25 (38)	24 (63)	1 (4)
Skin thickening		N = 66	N = 38	N = 28	<0.001 b
	None	47 (71)	20 (53)	27 (96)
	Mild	10 (15)	9 (24)	1 (4)
	Moderate	6 (9)	6 (16)	0
	Severe	3 (5)	3 (8)	0
Auscultation		N = 66	N = 38	N = 28	0.2 b
	WNL	60 (91)	33 (87)	27 (96)
	Abnormal	6 (9)	5 (13)	1 (4)
RR (breaths per minute)		N = 66	N = 38	N = 28	0.3 d
	Median (range)	30 (18–84)	30 (18–84)	30 (18–42)
Temp (°Celsius)		N = 54	N = 37	N = 17	1.0 d
	Median (range)	37.9 (35.5–39.1)	37.9 (35.5–39.1)	37.9 (37.2–38.4)
HR (beats per minute)		N = 66	N = 38	N = 28	1.0 d
	Median (range)	54 (42–96)	54 (42–96)	54 (42–85)

^a p‐value comparing cart to work donkeys for given parameter.

^b Fisher's exact test.

^c Pearson chi‐square test.

^d Two‐sample Wilcoxon rank‐sum (Mann–Whitney) test.

A significant increase in both tracheal collapse grade and presence of tracheal ring abnormalities was observed in cart donkeys when compared to pack donkeys (both p < 0.001; Table 1). Moderate (grade 3) tracheal collapse was noted in 40% (95% CI: 24%–57%) of cart donkeys and 0% (0%–12%) of pack donkeys, while intra‐tracheal ring abnormalities were noted in 45% (29%–62%) of cart donkeys and 0% (0%–12%) of pack donkeys. The presence of pharyngitis was also seen more frequently in cart than pack donkeys (29% vs. 7%; p = 0.03).

Physical examination findings were associated with endoscopic findings in cart donkeys (Table 2). The presence of moderate or severe skin thickening was positively associated with tracheal collapse grade (Cochran–Armitage test for trend, p = 0.03). The presence of moderate/severe skin thickening was moderately predictive (7/9; 78%) of the most severe level of tracheal collapse (grade 3), while no/mild skin thickening was moderately predictive (21/29; 72%) of more mild levels of tracheal collapse (grades 1 and 2). The presence/absence of hair on the ventral neck (hairless skin) was associated with tracheal ring abnormalities in cart pulling donkeys (p = 0.03). The absence of hairless skin was moderately predictive of the absence of ring abnormalities (11/14; 79%), while the presence of hairless skin was poorly predictive of the presence of ring abnormalities (14/24; 58%).

TABLE 2.

Associations between endoscopic findings and ventral neck skin in cart pulling donkeys, Meru County, Kenya (N = 38).

Endoscopic findings a	Physical examination findings, N (%) a
	Skin thickening		p	Hairless skin		p
	None/mild	Moderate/severe		No	Yes
Tracheal collapse (grade)			0.03 b			0.3 b
0	0	0		0	0
1	7 (24)	1 (11)		4 (29)	4 (17)
2	14 (48)	1 (11)		6 (43)	9 (38)
	8 (28)	7 (78)		4 (29)	11 (46)
4	0	0		0	0
Ring Abnormalities			0.05 c			0.03 d
No	19 (66)	2 (22)		11 (79)	10 (42)
Yes	10 (35)	7 (78)		3 (21)	14 (58)

^a Physical examination and endoscopic findings categorised as: Skin thickening none or mild = No change in skin or only palpable change; moderate or severe = Visual and palpable change in skin with skin hanging ventrally. Hairless skin none = No visible or palpable skin lesions; yes = Hairless skin on ventral neck. Tracheal collapse (grade): 1 = <25% reduction in tracheal lumen; 2 = 26%–50% reduction in tracheal lumen; 3 = 51%–75% reduction in tracheal lumen (grades 0 and 4 not observed). Ring abnormalities: No: normal contour of tracheal ring; yes: irregular luminal contour of tracheal ring.

^b Cochran–Armitage test for trend.

^c Fisher's exact test.

^d Pearson chi‐square test.

4. DISCUSSION

Our results supported the hypothesis that work type was strongly associated with upper respiratory tract abnormalities such as tracheal collapse and cartilage ring abnormalities in the donkeys examined in this study. Over a third of cart donkeys (40%) had grade 3 tracheal collapse, and cartilage ring abnormalities were found in 45% of endoscopic examinations. Pack donkeys did not have tracheal collapse at this grade (0%) nor any ring abnormalities (0%). This finding suggests that the repeated trauma to the ventral neck from the current method of harnessing in the cart donkeys is associated with tracheal injury and remodelling. However, dynamic endoscopy could not practically be performed to confirm this in these working donkeys. External trauma to the neck has been recognised as a cause of tracheal collapse in horses and donkeys. ⁹ , ¹⁸ Ageing changes can occur in donkeys as the cartilaginous rings that give the trachea its structure can remodel with age. This finding on its own is not known to cause respiratory disease in sedentary donkeys. ¹² The pack donkeys were used as a control group to account for this ageing change and to lend importance to the increased collapse grades seen in cart donkeys. As there was no difference in age between the two groups studied here, age does not appear to be a factor in the difference seen in the airways between the cart and pack donkeys. Concurrent lower respiratory disease such as asthma or pulmonary fibrosis is another cause of tracheal collapse. In this study, no significant difference in respiratory rate or thoracic auscultation was found. A donkey suffering from severe respiratory disorders would be expected to have abnormalities detected on physical examination. ¹⁰

Despite the increased proportion of tracheal collapse in cart donkeys, no difference in tracheal mucous grade, thoracic auscultation, respiratory rate, rectal temperature, or heart rate was found when examined at rest. Varying degrees of abnormal physical examination findings have been described with diagnosed tracheal collapse. The clinical signs can vary with the severity of the disease and concurrent lower respiratory tract disease. A donkey suffering from acute tracheal collapse often shows signs of respiratory distress such as dyspnoea, attempts at mouth breathing and may have stertorous or ‘honking’ noises audible when breathing. Instability in the cartilage rings can be evident by digital palpation of the trachea inducing collapse and resulting dyspnoea. ¹⁰ More chronic cases can suffer from a chronic cough and dyspnoea. Tracheal collapse can go undiagnosed for years in donkeys, especially ones with sedentary lifestyles. ¹⁹ One study of American Miniature horses found 14.7% 5/34 (14.7%) had grade 1–2 tracheal collapse with no abnormal physical examination findings. It is also suggested that some cases of tracheal collapse can be more static in some species (American Miniature Horses) and more dynamic in others (canines). ¹³

The donkeys surveyed here were actively working and were able to do so with the changes seen endoscopically. Donkeys that are not tolerant of the current methods of harnessing may have been culled from the population. The rate of culling in this population is not known but would be useful in assessing the effect of the observed tracheal collapse on the health and ability of the donkeys to work. Endoscopic evaluation of the trachea of donkeys while working or shortly after working would be valuable to determine the degree of dynamic tracheal collapse and health parameters during work. It is also possible that the airway lumen narrowing seen in this study is more static in nature, and a result of enlargement of the tracheal rings rather than a decrease in the stability of the rings. A radiographic and/or fluoroscopic study to evaluate the diameter of the trachea and width of the cartilage rings would be a way to quantify this.

Most pack donkeys had normal skin on the ventral neck and no hairless skin (96% for both categories) compared to cart donkeys that frequently had thickened or hairless skin (37% and 53%, respectively). The skin thickening and hairless skin seen on the cart donkeys was always observed where the cart's cross bar was in contact with the ventral neck. This could indicate the skin findings are from harnessing, which is a common welfare issue in draught animals. ⁷ In this study, there were significant associations between moderate or severe skin thickening as well as hairless skin and severity (grade) of tracheal collapse in the cart donkeys. However, these physical examination findings were only limitedly predictive of the presence or absence of the endoscopic findings (e.g., only 58% of the cart donkeys with hairless skin had tracheal ring abnormalities, while 79% without hairless skin did not have tracheal ring abnormalities). While these findings can assist with future projects involving welfare assessment of draught donkeys when endoscopy is not available, further work in other donkey cart pulling populations with larger sample sizes is needed to determine if these or other physical examination findings are useful as predictors of tracheal collapse.

We were able to recruit the calculated sample size in cart donkeys, but recruitment was slightly lower than planned in the pack donkeys due to adherence to exclusion criteria as defined above (dual purpose cart and pack donkeys, fractious behaviour, less than 1 year of age). Pack donkeys tended to be more fractious and less tolerant of handling necessary to perform the endoscopic examination. Sample size calculations were based on highly conservative estimates from preliminary data, and as such, the reduced study power was unlikely to be impactful on the study findings. Despite this reduction in study power, we identified several significant associations between work type and physical and endoscopic findings. The conditions in rural Kenya under which the study was performed did not allow dynamic endoscopy to be performed, which would have led to stronger evidence that the harness was directly responsible for the abnormalities seen in the cart donkeys. However, the observed associations for the cart donkeys between external skin findings and tracheal pathology, as well as the location within the trachea of the abnormalities, allowed an assumption of the harness being responsible. Due to a low‐cost endoscopic system, some videos were unable to be assessed due to mucus accumulation on the end of the scope. A final limitation was that while assessment of the videos was blinded to the donkey working conditions and physical examination findings, the assessment of physical examination parameters, including skin thickening and hairless skin, was not blinded, and as such, has resulted in some bias, although careful development and use of categorisation and grading scores likely reduced this bias. Other limitations include the use of health clinics for recruitment and thus the potential for selection/enrolment bias.

It is well recognised that prevention of pressure over the trachea during draught work is essential for efficient, humane draught donkey, mule, and horse work. ⁶ The increased prevalence of tracheal collapse seen in cart donkeys with poorly designed harnessing and carts in this study indicates one physical repercussion of lack of attention to cart and harness design. Future community outreach efforts and research directed towards the education of donkey owners on the benefits of improved harnessing could improve the productivity and welfare of draught donkeys. Determining if an intervention is effective in improving animal welfare is difficult and could be improved by communication of past efforts for future planning. ²⁰ This study attempts to establish one component of a baseline of current welfare concerns related to harnessing seen in working donkeys in Meru County, Kenya. A participatory action project designed to facilitate donkey owners in the design and implementation of an improved cart is planned, with the information gathered in this study as a steppingstone to motivate change.

5. CONCLUSION

There is an increased occurrence of tracheal collapse, tracheal ring abnormality, skin thickening, and hairless skin in cart donkeys compared to pack donkeys in Meru County, Kenya. Future efforts for this research program will be directed towards behaviour change of owners to use a harness and cart that relieves the external trauma on the neck and efforts to improve the welfare of the donkeys.

FUNDING INFORMATION

Farmers Helping Farmers, The Atlantic Veterinary College Internal Research Fund, Merck Animal Health Canada, and Zoetis Canada supported this project through personnel, travel funding, vaccines, and de‐wormers.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

AUTHOR CONTRIBUTIONS

Martha Mellish: Conceptualization; investigation; funding acquisition; writing – original draft; writing – review and editing; project administration; data curation; supervision; resources. Jennifer Burns: Investigation; writing – review and editing. Yvonne Elce: Investigation; writing – review and editing. Jason W. Stull: Methodology; writing – review and editing; formal analysis.

DATA INTEGRITY STATEMENT

Martha Mellish is responsible for the integrity of the data and the accuracy of the data analysis.

ETHICAL ANIMAL RESEARCH

The protocol was approved by the Animal Use and Care Committee of the University of Prince Edward Island, File No. 6012157.

INFORMED CONSENT

Explicit informed consent was obtained from each owner through verbal communication from PI and a translator.

ANTIMICROBIAL STEWARDSHIP POLICY

Not applicable.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are openly available in Beta Data at https://doi.org/10.11571/upei-roblib-data/researchdata:812 and https://doi.org/10.11571/upei-roblib-data/researchdata:813.