The taxa in the Potentilla nivea complex (section Niveae) are closely related and morphologically similar, they are mainly separated on indumentum types and number of leaflets: P. chamissonis Hulten has long, straight, verrucose hairs and ternate leaves; P. nivea L. has short, floccose hairs and ternate leaves; and P. insularis Soják has long, straight hairs less verrucose than P. chamissonis and digitate leaves. Despite the close relationship, the taxa have been subdivided into a series of subspecies, varieties, forms and postulated hybrids.

The subdivisions might reflect local races that have arisen through the predominantly apomictic mode of reproduction in these plants. Many species of Potentilla reproduce by pseudogamy, i.e. agamospermy with a demand for pollen to initiate seed set (endosperm formation). Furthermore, the apomixis is often facultative, i.e. some seed production might be sexual. The reproduction system may vary among and within taxa and populations, see e.g. Eriksen (1996), Holm (1996), Asker (1977).

In this study taxa in the Potentilla nivea complex from Svalbard and Fennoscandia have been analysed. Collections were made of Potentilla nivea ssp subquinata (Lange) Hulten, P. chamissonis, and P. insularis from Svalbard, P. chamissonis and P. nivea ssp nivea from Northern and Southern Norway, and P. chamissonis from Northern Finland. The material is or will be analysed by means of isozymes, RAPDs (Random Amplified Polymorphic DNAs) and morphometry. Crossing experiments are carried out within and between the taxa from Svalbard. Offspring from the crossing experiments will be analysed by means of isozymes and morphometry.

The taxa have slightly different habitats. At Svalbard, P. chamissonis occurs in bird manured scree slopes and flat ground close to the shoreline; P. insularis occurs in south facing cliffs, often bird manured scree slopes; and P. nivea occurs in south facing cliffs and on mountain plateaus. In mainland Norway, P. chamissonis was found mainly in subalpine cliffs, and P. nivea mainly on alpine mountain plateaus. However, in a few localities the two taxa occurred together.

Soják (1986) proposed that P. insularis, a Svalbard endemism, had arisen from P. chamissonis x P. lyngei Yurtsev & Sojak (the latter in section Multifidae). Potentilla lyngei has been reported once from Svalbard, but this collection has been questioned (Elven & Elvebakk 1996). It probably represents a manured specimen of P. insularis (cf. Eriksen & Nyléhn 1999). Elven & Elvebakk (1996) proposed that rather P. chamissonis and P. pulchella R. Br. (the latter in section Multifidae) might be the parent species of P. insularis. Furthermore, Soják (1986) proposed that P. nivea ssp subquinata is the hybrid between P. nivea ssp nivea and P. pulchella. These hypothese of hybrid origin for the Svalbard taxa are not supported by molecular data.

Analyses of RAPDs (Hansen 1998), morphometry, and isozymes (Hamre in prep.) have revealed that P. insularis, P. nivea ssp subquinata and P. chamissonis on Svalbard are closely related, and distant from P. pulchella. P. pulchella is neither a parent species of P. insularis nor for P. nivea ssp. subquinata.

When material from Fennoscandia was included in the RAPD analysis of taxa in the P. nivea complex, three different groups were established in the PCO plot (Fig. 1). P. chamissonis from Finland and Southern Norway is grouped together with P. chamissonis from Svalbard. P. nivea ssp. subquinata from Svalbard and P. nivea ssp nivea from Southern Norway form separate groups, as distant from each other as from P. chamissonis. The collections from Northern Norway are not analysed at the moment. Thus, the so far analysed material (P. chamissonis, P. nivea subsp. subquinata, P. nivea subsp. nivea and P. insularis) are closely related, but should probably be regarded as separate taxa. Whether the taxa will be given rank of species or subspecies will await further research (especially analysis of the material from Northern Norway).

Although hybrids, hybrid swarms and complexes are proposed to be common in these taxa, only few putative hybrids were found in our collections. No hybrids between different sections were found. In two populations in Southern Norway where P. nivea and P. chamissonis co-occured, no certain hybrids were found. The position of the two lowermost South Norwegian P. chamissonis individuals, originated in a mixed population, might indicate some introgression. Only a few plants in the Svalbard material were probably of hybrid origin, possibly between P. nivea and P. insularis (Hansen, 1998).