Pelvic bone structures in free-ranging Eurasian lynx (Lynx lynx) from Switzerland: a radiological pelvimetry study

The observation of severe pelvic malformations in Eurasian lynx (Lynx lynx) from a population reintroduced to Switzerland raised the question as to whether inbreeding may contribute to the development of congenital pelvic malformations. We aimed at providing baseline data on the pelvic morphology of Eurasian lynx from the reintroduced populations in Switzerland, at assessing potential differences in pelvic conformation between the two main Swiss populations, among age classes and between sexes, and at detecting pelvic anomalies. We performed measurements of 10 pelvic parameters on the radiographs of 57 lynx of both sexes and different ages taken from 1997-2015. We calculated two ratios (vertical diameter/acetabula; sagittal diameter/transversal diameter) and two areas (pelvic outlet and inlet) to describe the shape of the pelvis. Our results showed that the Eurasian lynx is a mesatipelvic species, with a pelvis length corresponding to approximatively 20% of the body length. We found no statistically significant differences between the two examined populations but observed growth-related pelvis size differences among age groups. Sexual dimorphism was obvious in the adult age group only: two parameters reflecting pelvic width were larger in females, likely to meet the physiological requirements of parturition. By contrast, pelvis length, conjugata vera, diagonal conjugata, vertical diameter and sagittal diameter were larger in males, in agreement with their larger body size. Accordingly, the ratio between the sagittal and transversal diameters was significantly larger in males, i.e. adult males have a different pelvic shape than adult females. Furthermore, pelvimetry highlighted one adult individual with values outside the calculated reference range, suggesting a possible congenital or developmental pathological morphology of the internal pelvis. Our work generated baseline data of the pelvic morphology including growth and sexual dimorphism of the Eurasian lynx. These data could also be useful for estimating age and sex in skeletal remains.


41
The Eurasian lynx (Lynx lynx) vanished from Switzerland and other European countries at the the boundary between the pelvic and the abdominal cavity. The pelvic outlet area (POA) 162 represents the lower circumference of the lesser pelvis.  177 We stored the data in a MS Excel © spreadsheet and imported them into the statistical software 178 NCSS (NCSS 10 Statistical Software, 2015;NCSS, LLC;Kaysville, Utah, USA, 179 ncss.com/software/ncss). We excluded five <4 month-old lynx from the data analysis because 180 their pelvic values were much lower (pelvis length <10 cm) than those of other juveniles and 181 these animals were too few to create an additional age class. The 57 remaining lynx included 182 32 animals from the Alps, 23 from the Jura, and two from a recently reintroduced population 183 nucleus in north-eastern Switzerland. There were 26 males, 30 females and one animal of 184 unknown sex (adult, Jura population); and 24 juveniles (4-12 months old), 11 subadults and We used the Shapiro-Wilk normality test to assess the normal distribution of each 187 variable. First, we tested for differences between the two main populations (Alps and Jura), 188 among age classes and between sexes for each variable (i.e. pelvic measurements) in a 189 univariable approach, using either the parametric unpaired Student t test (if variables were 190 normally distributed) or the Mann-Whitney U test (if variables were not normally distributed).

191
Subsequently, population, age class and sex were combined in a multivariable approach. We 192 applied a MANOVA for each pelvic measurement (outcome variable), including the 193 population, age class and sex as explanatory variables. The level of significance was set to 194 0.05. In addition to the five juveniles <4 months, we excluded the adult lynx of unknown sex 195 and the two lynx from the population in north-eastern Switzerland from the multivariable 196 models, i.e., these models included 54 individuals: 32 from the Alps (16 juveniles, 4 subadults 197 and 12 adults) and 22 from the Jura (7 juveniles, 6 subadults and 9 adults). Thus, the results 198 apply to specimens >4 months from the two populations Alps and Jura.

199
Male Eurasian lynx are expected to be larger in size than females, at least in the 200 subadult and adult age classes (17). Therefore, we assessed differences in body length and 201 pelvis length between sexes for each of the three age classes using unpaired Student t tests.

202
Additionally, we calculated the ratio between the pelvis length and the body length for each 203 specimen, and we tested for differences between the ratio of males and females for each of the 204 three age classes using unpaired Student t tests or Mann-Whitney tests as appropriate. 205 We did not include BETHLI in the statistical analyses but we discuss how this 206 animal's data compares to the measurements of six other specimens (three females and three 207 males) of the same age class and with similar pelvic length (controls).

231
All linear radiographic pelvic measurements increased with age (Table 1 and 2, Fig 3). There 232 was a statistically significant difference (MANOVA) between the three age groups in both 233 males and females for all parameters except VD ( Table B in S1 Table). This parameter also 234 increased from the juvenile to the adult class and there was an outlier among adults (

249
Regarding the pelvic height/width relationships, the VD/AC ratio significantly 250 decreased with increasing age in both sexes (Table 1 and 2, Fig 3). The SGD/TD ratio did not 251 significantly differ between age groups (Table 1 and Table 2).