Lamini

Moyano, S. R., Aragon, P. N. Sardina, Alvarez, A., Ercoli, M. D., L, L. N., Geronazzo, opez, Gonz, J. L. & Fossati, alez, 2022, Comparative anatomy of the skull of South American camelids. A contribution to their taxonomical identification, Zoologischer Anzeiger 299, pp. 82-95 : 89

4.1. New tools for Lamini identification

In a broad sense, the studies of osteological identification are typically carried out using traditional morphometric techniques, being effective to distinguish the species of South American camelids mainly by their size: small camelids (vicunas ˜and alpacas) of the large camelids (guanacos and llamas) (L’ Heureux & Hern´andez 2017, 2019 and references therein). However, separating llamas and guanacos exclusively using the size is a difficult or non-possible task due the overlapping between smaller guanacos and larger llamas. Moreover, the problem is accentuated when northern guanacos (typically smaller than southern ones) are considered. In this way, considering the size to distinguish these species is useful but not sufficient, and the consideration of anatomical traits are necessary.

Unfortunately, the cranial morphologies of llamas and guanacos are very similar, which makes taxonomic identification difficult. There are few osteological studies available ( Cartajena 2009; L’ Heureux 2010; Hern´andez 2016; L’ Heureux & Hern´andez 2017; Balcarcel et al. 2021), many of them do not reach a robust set of diagnostical features, and some other analyze non-adequate samples or have methodological flaws (e.g., minimal taxonomical representation, not considering northern guanacos).

Our analysis confirms that vicunas ˜are distinguished from guanacos and llamas by size, but also by other features, including both discrete ones and proportions. This fact was reflected in the results of the PLS-DA in which, when using both discrete and continuous matrices together or separately, specimens of vicunas ˜were identified with a 100% of accuracy in all cases (see Table S2). From our opinion, there is no enough available information in literature that allow the distinction of guanacos and llamas, a topic in where our study attempts to contribute. Moreover, our study tests statistically a novel set of features to distinguish these species, assessing the performance and confirming the discriminatory power of the proposed variables (see Table S2). These analyses also allowed testing previous assertions. Some osteological features that have been previously proposed (e.g., Pacheco Torres et al. 1986; Adaro & Benavente 1992) are not supported by the discriminant analysis (see below) and appear to be intraspecific variations wrongly identified as interspecific variations, probably in relation to too small samples.

Moreover, there exist additional difficulties related to hybridization of llamas and guanacos ( Renieri et al. 2009 and references therein). Although we confirmed a significant morphological segregation of these large camelids, and we were able to determine a set of characters that allow their separation with high efficiency (see below), the higher difficulties found by the discriminant functions constructed by different subsets of data (analyzing discrete and continuous variables alone), could be eventually related to this fact. The specimens with relatively intermediate morphologies, wrongly classified by discriminant functions, could or could not be related to hybridization events. Nevertheless, genetic data should be considered together with morphological ones to confirm this hypothesis.