Ateleute, Santos & Alvarado & Sääksjärvi & Noort & Villemant & Brady, 2018

PARAPHYLY OF ATELEUTE View in CoL View at ENA

As per its current taxonomic definition ( Townes, 1970; Kasparyan & Hernandez, 2001; Bordera & Sääksjärvi, 2012), the genus Ateleute is rendered paraphyletic by Tamaulipeca , a group that is morphologically well-characterized and readily diagnosable (see Taxonomy section below). While the clade including the Old World taxa is stable and wellsupported, the relationships among the Neotropical forms currently assigned to Ateleute have low support, and clade composition does not correspond to any putative morphological synapomorphies. While these taxa still fit the ‘broad’ definition of Ateleute , the morphological variation observed across the examined species is much higher than previously recognized. Several described New World species, including most of the ones described by Bordera & Sääksjärvi (2012) and the only described Nearctic species of the genus, A. carolina Townes , could not be obtained for sequencing, and their affinities remain uncertain. Hence, further work with better sampling of the New World taxa is needed in order to provide a more thorough morphological characterization of the group and establish sound generic limits.

One obvious solution would be to synonymize Tamaulipeca in order to render Ateleute monophyletic. However, lumping all of the considerable phenotypic diversity of the group under a single genus is more likely to complicate classification and taxonomic identification. At the same time, considering the limitations of the results observed herein, attempting to provide a full generic classification based on the data presently available is clearly premature. Hence, it seems more appropriate to progressively delimit monophyletic and morphologically diagnosable groups as more data accumulates. The proposal of Duwalia gen. nov. represents a step towards that direction, while the description of the aberrant A. boitata sp. nov. intends to highlight the unexplored phenotypic diversity in the genus.

BIOGEOGRAPHY

Drawing biogeographic inferences for Ateleutinae from the recovered tree topology is not straightforward. The fact that the earliest diverging lineages are all from the Neotropical or Australian regions may suggest that the subfamily as a whole has a Gondwanan history. Within Ateleute , all examined species that currently occur in areas geologically belonging to Laurasia ( A. linearis , A. densistriata Uchida and three undescribed species from South-East Asia) were recovered in a single clade nested within a Gondwanan background. Species from Madagascar, the Afrotropical region at large and the Australasian region appear scattered across several groups within the clade.

It is tempting to hypothesize that these patterns may have been driven by past vicariance events, such as the break-up between Africa and South America separating the Neotropical lineages from the Old World clade. However, the taxonomic sampling of the current analyses is far from complete; many of the known species are not represented in the phylogeny, and the Ateleutinae as a whole clearly include many undescribed species. In addition to that, the lack of information regarding divergence times casts doubt on whether the observed diversification patterns may be chronologically consistent with putative vicariant events. A vicariancedriven scenario would suggest the Ateleutinae as an ancient lineage within Ichneumonidae , as the oldest fossils known for the family date from the Lower Cretaceous ( Zhang & Rasnitsyn, 2003; Kopylov, 2010; Kopylov & Zhang, 2015).

Currently there are no known fossils for Ateleutinae , and the currently precarious understanding of the ichneumonid fossil record ( Spasojevic et al., 2018) requires rigorous study and re-evaluation of fossil taxa before any reliable molecular-clock-type analysis to infer divergence timing for the subfamily. It is noteworthy that there seems to be little ‘geographic conservatism’ in the evolution of Ateleutinae ; species occuring in each region seem to be derived from multiple lineages. This is consistent with a scenario of very old divergence times, predating vicariant events such as continental splits, but it may also suggest a considerable amount of dispersal among closely related lineages. For example, multiple lineages seem to be present in Madagascar, which separated from the Indian peninsula 88 Mya, and from Africa 135 Mya ( Briggs, 2003; Ali & Aitchison, 2008). This implies either that these lineages were already present in Madagascar by then, or that dispersal between island and continent has subsequently occurred multiple times.