Wielstra Lab

In 2014 the New Atlas of Amphibians and Reptiles of Europe (the first update since 1997’s original atlas) was published in the journal Amphibia-Reptilia. In the new atlas, the crested newts and marbled newts were mapped as such, even though both of these two groups encompass more than one species. However, as these species are morphologically similar and, on top of that, hybridize at their poorly documented contact zones, the authors of the new atlas did not feel they could confidently map their distributions. We are working extensively on these newts and, particularly based on our recent field trips and our new Triturus Ion Torrent protocol, have accumulated enough data to allow us to map the individual species. Hence, in a follow-up paper in Amphibia-Reptilia, we published distribution maps of all the species that comprise the marbled and the crested newt groups.

This is an overview of all the European grid cells that have Triturus localities (with cells having more than one species coloured blue rather than red). Maps for the individual species are published with the paper.

Reference: Wielstra, B., Sillero, N., Vörös, J., Arntzen, J.W. (2014). The distribution of the crested and marbled newt species (Amphibia: Salamandridae: Triturus) – an addition to the New Atlas of Amphibians and Reptiles of Europe. Amphibia-Reptilia 35(3): 376 –381.

I conducted this work as a Newton International Fellow.

During my work on the crested newt traditionally referred to as Triturus karelinii it soon became clear that more than one species is involved (although sorting out the details took a bit of time and is still not quite finished). The name arntzeni had been used before to refer to newts from the Balkan part of the range. This name was proposed in honor of my mentor Pim Arntzen. I never bought it. Not the part of Pim deserving his own newt – he most certainly does. But I did not believe that the type locality, near the small village Vrtovać in Serbia, belonged to Triturus karelinii sensu lato. It was located in such an inconvenient place, in a region where the distribution was not well understood but where I strongly suspected Triturus macedonicus would occur. We went and visit the type locality during a field trip in 2010 with Jelka Crnobrnja-Isailović. We had been joking about the identification of these newts beforehand and when we walked up to the pond judgement day had arrived. As usual there was a bit of competition of who would catch the first crested newt and as usual it was Pim who netted the first newt. I don’t remember the exact words he used when he took it out of the net but it was pretty dry and went something like “Well… that’s macedonicus…”. I have to admit I thought it all pretty funny.

Pim at the pond

These sedated newts just don’t look like karelinii

Last year we published a paper in Zootaxa where we reviewed all the available data and made clear that newts from the type locality of arntzeni strongly resembled macedonicus. So basically Pim burned his own name! Spartak Litvinchuk provided some tissue from the actual type material but this was too degraded to sequence DNA. Although DNA data from fresh material pointed towards macedonicus, arguably the amount of markers used was a bit low (and note that mitochondrial DNA is pretty useless in this region due to wide scale introgression). However, with the new Triturus Ion Torrent protocol sequencing a lot of markers for poorly preserved material suddenly became possible. In a follow-up paper, again in Zootaxa, we compared high quality genetic profiles for nine newts from the arntzeni type locality (including the holotype to which, according to the rules of taxonomy, the name is officially attached) to a comprehensive sample including all crested newt species. We got an unambiguous picture of the genetic ancestry of the Vrtovać newts: as suspected they were mostly macedonicus, but there was also a considerable amount of karelinii genes present, evidencing hybridization. These finding strongly support that the name arntzeni is not valid. Instead we introduced the name ivanbureschi and this time we made sure the type locality is nowhere near a contact zone with another species of crested newt!

This figure might look a bit complex but sums up the genetic identity of the Vrtovać newts. The bottom left corner represents a pure macedonicus genotype and the bottom right pure karelinii (now ivanbureschi). The top of the triangle represents pure F1 (first generation) hybrids. This plot is made with the HIest R package by Ben Fitzpatrick.

By the way, this pond is the type locality for ivanbureschi. We didn’t put this picture in the original paper, not sure why actually.

Reference: Wielstra, B., Arntzen, J.W. (2014). Kicking Triturus arntzeni when it’s down: large-scale nuclear genetic data confirm that newts from the type locality are genetically admixed. Zootaxa 3802(3): 381-388.

Reference: Wielstra, B., Litvinchuk, S.N., Naumov, B., Tzankov, N., Arntzen, J.W. (2013). A revised taxonomy of crested newts in the Triturus karelinii group (Amphibia: Caudata: Salamandridae), with the description of a new species. Zootaxa 3682(3): 441-453.

I conducted this work as a Newton International Fellow.

You can only learn so much about a study system if you have few genetic markers available. Particularly if that study system has an extensive history of hybridization, as is the case for Triturus. Because salamanders have massive and complex genomes it is not possible to simply sequence one. Not yet at least. Luckily we had some Triturus transcriptome data laying around. The transcriptome contains messenger DNA – the transcripts of functional genes – but not the long introns and endless repeats that torment salamander genomes. Hence it provides relatively simple, genome-wide reference data for marker design. After designing a large set of markers, I tested which of these worked for all crested newt species and multiplexed the successful ones for a large set of newts. Next the whole bunch was sequenced at Naturalis’ next-generation sequencing facility on an Ion Torrent machine and subsequently the huge amount of genetic data was sorted out with a bioinformatics pipeline and converted it to a workable format. It sounds easy, but it was quite an undertaking.

The top picture shows the localities sampled (circled numbers, three newts each) and the bottom picture shows the probability with which they belong to their own species (all 1, as would be expected if our method works).

Using data for a set of newts representing all species and also some putative hybrids we showed that the protocol can be used to allocate individuals to the proper species and pick out genetically admixed newts. Although the amount of markers is still relatively modest, the data provide a very detailed picture on the genetic composition of crested newts. A paper describing the methodology has recently been published in Molecular Ecology Resources. Basically we can now provide a detailed picture of the distribution of the different species and the genetic distinction of and gene flow between these species: an important next step in my research. Meanwhile I have sequenced about 1500 individuals throughout the range of Triturus as the basis for quite some papers to come. Stay tuned!

These are a couple of newts from the population marked with an H on the map above.  You can click on the picture to see a larger version. The throat and belly pattern is very variable in this populations. Some animals look more like macedonicus or ivanbureschi and the ones depicted here look especially messy. Based on the Ion Torrent data these newts indeed show mixed macedonicus and ivanbureschi genetic ancestry and almost all are identified as backcrosses towards ivanbureschi (with the remainder being F2 hybrids).

By the way, many thanks to Wieslaw Babik, Michał Stuglik and Piotr Zieliński from Jagiellonian University, Krakow, Poland for helping with the design of this protocol!

Reference: Wielstra, B., Duijm, E., Lagler, P., Lammers, Y., Meilink, W., Ziermann, J.M., Arntzen, J.W. (2014). Parallel tagged amplicon sequencing of transcriptome-based genetic markers for Triturus newts with the Ion Torrent next-generation sequencing platform. Molecular Ecology Resources 14(5): 1080-1089.

I conducted this work as a Newton International Fellow.

The different crested newt species typically have very distinct mitochondrial DNA. However, as these guys hybridize like rabbits, often the mitochondrial DNA of one species has locally been transferred (introgressed) into the wrong species. ‘Locally’ can refer to quite an extensive range actually. Species distribution modelling of distinct geographical populations, only recognizable based on DNA divergence, is a good way to test whether they represent cryptic species – multiple species erroneously classified as a single one because no one previously realized they were distinct. However, if you would use mitochondrial DNA to identify the putative species, introgression would cause misidentifications and could lead to under- or overestimation of the ecological niche of the putative species. We explore this problem in a paper just out in PLoS ONE.

This figure shows the setting in which we test the effect of introgressed mitochondrial DNA on niche estimation. The two areas with red dot on the wrongly colored background are causing trouble. If using mitochondrial DNA as a guidance, you would interpret these localities as beloning to the red species, while excluding them from the green or blue species.

Reference: Wielstra, B., Arntzen, J.W. (2014). Exploring the effect of asymmetric mitochondrial DNA introgression on estimating niche divergence in morphologically cryptic species. PLoS ONE 9(4): e95504.

I conducted this work as a Newton International Fellow.