On the distinction of Dactylorhiza baltica and D-pardalina (Orchidaceae) and the systematic affinities of geographically intermediate populations

The eastern European Dactylorhiza baltica (Klinge) N. I. Orlova and the western European D. pardalina (Pugsl.) Aver. (= D. praetermissa var. junialis (Verm.) Sengh) are usually considered to have non-overlapping geographic distributions, for which reason it has rarely been realized that they are morphologically similar. They have not previously been thoroughly compared by molecular methods, and no existing flora or revision has convincingly demonstrated that they can be distinguished by morphological characters. In reality, they might be 'political' rather than natural taxa. Prompted by the recent discovery of geographically intermediate populations (in eastern Denmark), originally identified as D. baltica, we have addressed this problem by analysis of morphometric data as well as molecular data from allozyme markers, plastid haplotypes, nuclear ITS alleles and nuclear microsatellites. Dactylorhiza baltica and D. pardalina turned out to be clearly distinguished genetically, and although they are morphologically similar, a few characters were identified that distinguish with 81-85% certainty between the two taxa. Molecular and morphometric data place the geographically intermediate populations in D. pardalina. Both taxa were confirmed to be allotetraploids combining diploid genomes from the D. incarnata s.l. and D. maculata s.l. lineages, and they should therefore be recognized as infraspecic taxa under D. majalis s.l. Thus, D. baltica should be called D. majalis subsp. baltica; D. pardalina is identical with D. praetermissa var. junialis, but the nomenclatural consequences for D. praetermissa, if treated as subspecies under D. majalis, are still unresolved.