Bone fractures in roadkill Northern Tamandua Tamandua mexicana (Mammalia: Pilosa: Myrmecophagidae) in Costa Rica

Northern Tamandua Tamandua mexicana is one of the most common roadkill species encountered on Costa Rican highways. Ten roadkill Northern Tamanduas were collected along different roads in Costa Rica and moved to a veterinary facility where appendicular radiologic studies were undertaken. The number of fractures present in each individual varied from zero to five (mean=2.6), with only one animal sustaining no fractures at all. Most fractures were present in the humerus (31%), followed by the ulna and ilium (both 19%), whilst the cranial portion of the body represented the highest number of fractures (61%). These data can contribute, not only to establishing causes of animal-road-mortalities, but also to the future understanding and decision-making of clinical actions for animals injured on the roads.


INTRODUCTION
Highways are significant factors of wildlife mortality; they interfere with natural migration routes and are responsible for habitat fragmentation, which is one of the main causes of bio diversity loss (De la Ossa-V & Galván-Guevara 2015). A further negative impact of roads is the resulting collision with a vehicle, namely, roadkill.
Studies have shown that roadkill seriously decreases animal populations and can even result in local extinctions creating a risk factor for xenarthran persistence, even to non-threatened species (Ribeiro et al. 2017). Among Costa Rican wildlife species, the Common Opossum Didelphis marsupialis and a species of anteater Tamandua mexicana (Image 1) dominate roadkill numbers (Monge-Nágera 2018b).
Monge-Nágera (2018a) found that T. mexicana was the most common roadkill species (n=73), followed by the Common Oppossum (n=66) from data gathered over a four year period (2014)(2015)(2016)(2017)(2018) T. mexicana had higher numbers of roadkill during the dry season (December to April) than the wet season (May to November) in Costa Rica and Colombia (Nadjar & De la Ossa 2013; Monge-Nágera 2018a) possibly because ants are scarcer in dryer habitats than when it is wet, causing anteater species to travel further in search of food. This may result in them crossing roads more frequently and becoming roadkill although this is just speculation since the seasonal behavior of T. mexicana is poorly known (Nadjar & De la Ossa 2013; Monge-Nágera 2018a) and further study of their behaviour is needed

14804
bone radiographic appearance as well as the location of the fracture is vital in assisting the treatment of injured animals. Furthermore, increased public awareness of the number of wildlife injured or killed on Costa Rican roads may have contributed to the mitigation methods being applied.
T. mexicana is a species of a medium-sized anteater ranging from southern Mexico to northwestern Andes in South America and lives in diverse forest ecosystems (Navarrete & Ortega 2011). It is classified as "Least Concern" by the IUCN due to its wide distribution, and presumably, large population, however, the current population trend is unknown ( The purpose of this communication is to use roadkill T. mexicana as tools to demonstrate the location and frequency of bone fractures resulting from an impact with a vehicle. This information can be a valuable reference in future medical and surgical procedures at rescue centers.

METHODS
Ten roadkill of T. mexicana were collected between April to July 2016, on different roads in Costa Rica. For each individual, the date and time of collection, geographic coordinates and the person responsible for collecting them were recorded. We believe that two of the animals were at least one year old. The remaining eight individuals consisted of mature animals based on closed epiphyseal growth plates. Based on the presence of the reproductive organs, we were able to sex three of the T. Mexicana confirming there were two females and one male; three of the carcasses had no organs, whilst the remaining four individuals were unclear due to damage to the abdominal cavitiy.
The 10 T. Mexicana (all deceased) were transferred to a veterinary establishments where appendicular radiology studies were carried out using an Ultra 12040HF (Diagnostic Imaging Systems®) X-ray unit with a CR2000 Plus Scanner and CR200 Navigator Software (Diagnostic Imaging Systems®). Lateral (L) and anteroposterior (AP) orthogonal views of the four extremities were taken. Each animal received six radiographs, with AP projections of the forelimbs and hind limbs performed on both limbs at the same time. Each affected bone, in all animals, had only one fracture, which made further classification easier. For two of the animals, x-rays of the hind legs could not be done due to extensive damage to the legs. For the remaining eight, x-rays of the forelimbs (including the scapula, humerus, radius/ulna, carpus, metacarpals, and phalanges) and x-rays of the hind limbs (including the ilium, ischium, pubis, femur, tibia and fibula, tarsals, metatarsals and phalanges) were all taken.

RESULTS
Of the 10 roadkill, a total of 54 radiographs were taken (Figures 1 & 2). Two individuals had open epiphyseal growth lines indicating that they were juveniles and still growing. The number of fractures present in each individual varied from zero to five (mean=2.6); only one animal did not present any fractures, while three of them had a total of five fractures in different bones (Table 1). Sixty-one percent of the fractures occurred in the cranial portion of the body with the majority observed in the humerus (31%) followed by the ulna (19%) and then the radius (11%). The remaining 39% were present in the caudal portion of the body with the majority observed in the ilial (19%), followed by the femur (12%) and then tibia/fibula (8%); no fractures were observed in the scapula, ischium, pubis, carpal/tarsal, metacarpal/tarsal and phalanges.

DISCUSSION
There are few studies that utilize road-killed animals as a beneficial tool for learning, specifically for evaluating fractures on animal-roadkill species. Most studies that evaluate roadkill-animal-fractures mainly focus on domestic animals (for example, dogs and cats) (Minar et al. 2013;Martínez-Hernández et al. 2017 3  5  1  2  2  3  0  3  0  2  3  2  25 Key: L-left | R-right | H-humerus | R-radius | U-ulna | F-femur | T-tibia | I-ilium. Individuals T5 and T8 have no hind limb x-rays (x) | --means no fractures | X-means the individuals did not have the limb. percentage of fractures were found in the cranial portion of the T. mexicana whilst the most commonly affected bones in cats were found in the caudal portion (such as, the femur (28.2%) and pelvis (24.8%). The same contrast was also noticed with dogs with more fractures observed in the pelvis (15.8%), the femur (14.8%) and the tibia (14.8%) (Phillips 1979). Reasons for a higher percentage of fractures found in the cranial portion of the T. mexicana is likely due to a combination of dense roadside habitat and the anatomy of xenarthrans whose plantigrade locomotion, coupled with short limbs, provides low agility and relatively slow locomotion (Ribeiro et al. 2017). They emerge from the forest onto the road, providing a driver with limited visibility ad reaction time, and are often immediately hit by a vehicle in the frontal lateral position (Ribeiro et al. 2017).

Total overall
Other reasons for xenarthrans being vulnerable to roadkill are their poor vision, which may mean they simply do not see approaching vehicles (Ribeiro et al. 2017). Furthermore, speeding vehicles and poor driver visibility (particularly in areas with dense roadside vegetation) may also influence roadkill occurrence (Collinson et al. 2019).
Similar findings for fractures found on roadkill and comparable with T. mexicana is a study undertaken in North America, where an average of four fractures per individual in Virginia opossums Didelphis virginuanus were observed (Mead & Patterson 2009). This species is a medium-sized mammal, with head and body length ranging from 325 to 500 mm, weighing between 2-5.5 kg (Novakl 1991) and thus similar to the T. mexicana. Mead & Patterson (2009) reported several findings for opossums collected on roads, where the majority of skeletal injuries occurred in the cranial portion of the skeleton (for example, 54% rib fractures and 23% scapular fractures). Similar to the findings of Mead & Patterson (2009) on opossums, we also observed more than one fracture per individual in T. mexicana; this is likely due to them being of medium-sized resulting in multiple fractures when colliding with a vehicle (Cross 2012). Taxa

CONCLUSION
We found little literature evaluating long bone radiology in T. mexicana, therefore the data from our study which provides examples (and images) of normal appendicular x-rays plus the anatomic bone fracture location of trauma, can be used as a reference for further medical or biological studies, specifically at rescue and rehabilitation centers. Veterinarians can use epidemiological information to better understand surgical treatment of T. mexicana (and other species of similar size and behaviour), particularly for injured individuals that can then be rehabilitated and released back into the wild.