Splitters versus Lumpers? Subspecies designations must rely on robust morphological and/or genetic data—a response to Nabholz et al. (2024)

an important comment on our study on the phylogeography of Oedipoda latreille, 1829, reflecting some discussions we also had internally when writing the paper.While they agree with our conclusions of raising Oedipoda mauritanica lucas, 1847, Oedipoda meridionalis ramme, 1913 and Oedipoda pentagonalis Jannone, 1937 to species level, they challenge our synonymization of a subspecies described by the senior author (Oedipoda fuscocincta morini Defaut, 2006).Nabholz et al. (2024) argue that numerous cases exist in which mitochondrial phylogenies are not in line with valid species designations.We completely agree with this and we have documented such unresolved species groups before, such as the Sphingonotus caerulans group (Husemann et al. 2014), or the various species in the tribe Bryodemini (kock et al. 2024).For the same reason, we have refrained from taxonomic conclusions on the species status of Oedipoda germanica (latreille, 1804) and Oedipoda fuscocincta lucas, 1847.However, in these cases robust morphological data (and in the case of Sphingonotus Fieber, 1852 also bioacoustic data) support the species status of the taxa in question.In the case of subspecies, the situation is way more complex as subspecies definitions lack objective criteria and frequently fail to reflect distinct evolutionary units (Braby et al. 2012).We agree that our dataset is dominated by mitochondrial data, which is not unusual due to the slower mutation rates of nuclear DNA.For the same reason, a separate analysis of the ItS2 dataset does not provide substantial insights into the phylogeny of the genus.the ItS2 phylogeny presented by Nabholz et al. (2024) does not include any branch support indicators, but based upon our own analyses we can confirm that in addition to poor resolution, bootstrap support values and posterior probabilities of the ItS2 dataset alone were not acceptable to be publishable.Adding more comprehensive nuclear genetic data would be desirable, but due to their large genomes, genomic studies have long been challenging in orthoptera, particularly in oedipodinae (Hawlitschek et al. 2023).Meanwhile such issues can be overcome by sampling genome wide data using transcriptomics or genome reduction techniques such as hybrid capture or ddrAD to create more extensive SNP datasets (see for example Schmidt et al. 2024).We agree that taxonomic conclusions from genetic data alone must be undertaken with care and all evidence must be considered.this is particularly true for mitochondrial genes, which may be affected by introgression, incomplete lineage sorting or NuMts as correctly stated by Nabholz et al. (2024).When we started sequencing mitochondrial genes for inferring phylogenies in orthoptera (Hochkirch 2001), we soon realized that CoI (the "barcoding gene") is particularly difficult to apply in orthoptera phylogenies, as this gene appears to be frequently copied into the nuclear genome, leading to unresolved phylogenies (see for example Hawlitschek et al. 2017;kock et al. 2024).For this reason, we relied on more stable gene fragments, including NDS, which is a combination of 16S rDNA, trNA-leu and NADH-Dehydrogenase subunit 1 (and therefore unlikely to be copied as a whole into the nuclear genome) as well as 12S rDNA, which is known to be quite stable in orthoptera.this, however, does not prevent any erroneous conclusions from mitochondrial introgression or incomplete lineage sorting.Nabholz et al. (2024) argue with the Chorthippus biguttulus group as an example of a species group that cannot be separated with mitochondrial data alone (note that we follow the globally accepted taxonomy here (Cigliano et al. 2024) and still refer to Chorthippus Fieber, 1852 rather than Gomphocerippus roberts, 1941).We do not doubt the validity of Chorthippus biguttulus (linnaeus, 1758), Chorthippus mollis (Charpentier, 1825) and Chorthippus brunneus (thunberg,