Solitary herb with a central root, the main parts of the stem a short, ground-level, mostly branched caudex with leaf rosettes at the ends of the branches, and lateral from the base of the rosettes one or mostly more ascending flowering stems reaching 10–30 cm or more.

Leaves alternate, mostly basal, with long petioles and (3)4–5 leaflets. Petioles 1.2–3.5(6.0) cm, longer than blade, mostly densely covered by 1–2 mm long, appressed or ascending, verrucose hairs (a VERY strong lens is needed to see that they are verrucose, preferably a microscope), appressed or semi-appressed, never with floccose hairs (see Comparison for explanation of ‘floccose’). Blade 0.9–2.0(2.3) × 0.9–2.0(2.2) cm. Upper blade surface subglabrous and dark green or tinged reddish, or covered with long, straight, white hairs. Lower blade surface pale grey or white due to a pubescence of crisp hairs between veins and long, straight, semi-appressed hairs on veins. Always at least some leaves with 4–5 leaflets. The two pairs of lateral leaflets sessile and with a small distance between the pairs, making the leaf ‘semi-digitate’ or ‘semi-pinnate’ rather than true digitate (as in, e.g., Potentilla nivea) or true pinnate (as in, e.g., P. pulchella); the terminal leaflet has a very short stalk (petiolule) or is narrowly cuneate towards base. Leaflets obovate or oblong in outline, non-overlapping or nearly overlapping, terminal leaflet 0.6–1.5(1.9) × 0.4–1.2(1.5) cm, upper lateral leaflets 0.5–1.0(1.2) × 0.5–0.8(1.0) cm, margins with 3–5 pairs of teeth or lobes, mostly with tufts of long hairs in the apices of teeth or lobes. Stem leaves reduced, simple or with 2–3 small leaflets.

Inflorescence a cyme with (1)2–5(7) bracteate flowers on long pedicels.

Flowers radially symmetric with 5 epicalyx bractlets, sepals and petals. Epicalyx bractlets 3.5–4.5 × 0.6–1.0 mm, shorter and much narrower than the sepals, narrowly elliptic. Sepals 4.5–6 × 2.5–3.5 mm, triangular. Petals 6–9 × 5.5–7.5 mm, 1.5–2 times as long as sepals, obcordate, more or less emarginate, yellow. Stamens numerous. Carpels free, numerous. Styles apical, slender cylindrical, ca. 1 mm, with several distinct papillae at base.

Fruit a nutlet, up to 20–30 or more from each flower.

Reproduction by seeds, sexual and asexual (agamospermy); no vegetative reproduction. The flowers are adapted to insect pollination but the plant is facultatively agamospermic, that is, able to form seed both sexually and asexually (without fertilization of the egg cell). However, asexual seed development depends upon pollination for fertilization and development of the endosperm (the nutrient tissue for ovule development), a phenomenon named pseudogamy (Müntzing 1928; Gentscheff & Gustafsson 1940). Such endosperm fertilization is often more efficient with pollen from a relative than from the same species (Asker & Jerling 1992; Nyléhn 2002). This makes for interesting possibilities in plants with a potential for hybridization (see Elven et al. 2011). Investigated Svalbard populations of P. insularis are partly agamospermic (Nyléhn 2002). Fruit production is quite regular in Svalbard (R. Elven observ.).

There is no special adaptation to dispersal.

Potentilla insularis, P. arenosa ssp. chamissonis and P. nivea are the three taxa of Potentilla in Svalbard that combine the two characters of leaves with 3(5) leaflets attached at the same or nearly the same point (digitate or semi-digitate), and lower leaf surface white pubescent. Potentilla nivea differs from the two others in petioles with exclusively, or at least a total dominance of, floccose hairs (flat, crinkly, and intertwined); the others lack floccose hairs altogether; P. arenosa ssp. chamissonis has long, stiff, often patent, verrucose hairs on the petioles and regularly three leaflets; whereas P. insularis mostly have a denser layer of subappressed, indistinctly verrucose hairs on both petioles and blades and nearly always at least some leaves with more than three leaflets. Soják (1989) stated the hairs of P. insularis to be nearly smooth but those of the type specimens, in the Oslo herbarium (O), are distinctly verrucose.

Potentilla pulchella and P. lyngei are also often white-pubescent but are distinguished by having pinnate leaves, with two or more pairs of lateral leaflets at some distance from each other. Potentilla crantzii and P. hyparctica have digitate leaves with 3(5) leaflets but do not have the white pubescence on the lower leaf surface.

Almost confined to cliff ledges, rocky outcrops, and scree in slopes with high insolation rates, but also observed (albeit rarely) in dry gravelly or stony heaths or on loamy plains. On well drained, mixed or coarse substrates with circumneutral or basic soil reaction (pH). Requires moderate snow protection during winter but only occurs in sites where the soil is relatively early exposed in spring. Probably little grazed by reindeer or geese.

Thermophilous. Confined to the middle arctic tundra zone. Found in the clearly and weakly continental sections. Restricted to the central and northern parts of Spitsbergen where there are several sites along the inner and middle parts of Isfjorden, Kongsfjorden and Krossfjorden and a very few at Wijdefjorden and Bockfjorden.

Potentilla insularis is restricted to Svalbard (endemic), at least as currently circumscribed. It is, however, rather similar to P. pedersenii, a species widespread in Greenland and arctic Canada.

Potentilla insularis is the only species of Potentilla in Svalbard with semi-digitate or semi-pinnate leaves, i.e., with an origin from hybridization(s) between sect. Niveae and sect. Pensylvanicae (previously called sect. Multifidae) as stated by Soják (1986). It was reported by Soják (1986) as a hybridogeneous species developed from the hybrid P. arenosa ssp. chamissonis × P. lyngei. Plants with semi-digitate or semi-pinnate leaves occur as populations in numerous localities in Svalbard, mostly associated with P. pulchella, often with P. arenosa and sometimes with P. nivea, but never (until very recently) found together with anything that could be assigned to P. lyngei (see that species). The Svalbard material of P. insularis was commented on by Elven & Elvebakk (1996) and subsequently studied by Eriksen & Nyléhn (1999), Hamre (2000), Hansen et al. (2000), and Nyléhn & Hamre (2002).

Yurtsev (in Elven et al. 2011) argued in favour of Soják's hybrid hypothesis but Elven (in Elven et al. 2011) argued that Soják's hybrid theory, at least as concerns the type material (from the cliff Hyperithatten by inner Isfjorden, Nordenskiöld Land), is improbable. This hypothesis, of a hybrid origin from P. arenosa ssp. chamissonis × P. lyngei, is countered by: (a) The proposed parent from sect. Pensylvanicae (P. lyngei), even if present in one known site in Svalbard, is absent from the type site of P. insularis where this species grows together with P. pulchella and P. arenosa ssp. chamissonis (R. Elven observ.) (b) The common species in Svalbard of sect. Pensylvanicae is P. pulchella, and this is not a relevant parent from morphological evidence (this also supported by Yurtsev) or from molecular data. (c) In molecular markers, P. insularis is more similar to P. arenosa ssp. chamissonis than to P. nivea (the two other more or less widespread, related taxa in Svalbard) but consistently different. And (d), in isoenzymes, P. insularis is not more varied than, e.g., P. arenosa ssp. chamissonis or P. nivea (see Nyléhn & Hamre 2002). It does not include markers indicative of a hybrid origin with partaking of a more distant parent than from the P. nivea–arenosa group. The molecular evidence currently points towards P. insularis, a morphologically characteristic member of Soják's "Multifidae [=Pensylvanicae] × Niveae" intersectional hybrids, genetically being a member of sect. Niveae. Section Niveae may have to be extended to include some of the semi-pinnate leaved plants. Our third assumption is that the hypothesis of Soják and Yurtsev of inter-section hybridity should not be accepted as a general explanation until it is experimentally supported in several cases. This is also stated by Eriksen & Yurtsev (1999: 216): "In most other cases ... however, it is not obvious, [and] even unlikely, that the taxa described as hybrids descend from the suggested parents".

Whether P. insularis is a hybrid species or not, it has a range in Svalbard independent of its relatives and deserves recognition. The plants currently assigned to P. insularis may be recurrent hybrids forming populations by agamospermy after hybridization, an independent taxon or a combination of these (we suspect the latter). However, as P. pulchella is not a probable partner, we have no idea what other species could be a pinnate-leaved parent in such hybrids. In spite of several studies dedicated to just this species, it is still a mystery. The similarity with Greenland plants with semi-digitate or semi-pinnate leaves is, however, strong, and there P. pedersenii Rydb. is an obvious parallel. This is a rather common species in Greenland and the Canadian Arctic Archipelago. It may be that Svalbard P. insularis is a heterogeneous mixtum of P. pedersenii and plants of other hybrids. We may have to merge P. insularis in the strict (type) meaning into the western hemisphere P. pedersenii and loose one of the very few Svalbard endemics, for which Norway has a special arctic and global responsibility.

Rønning (1961, 1964, 1979) and authors who have followed him have treated P. insularis (in a mixtum with tall-grown P. pulchella) under the name P. rubricaulis. The true P. rubricaulis Lehm. is not a candidate in Svalbard as it is a rather different species, restricted to NW Canada and perhaps Alaska, and with a W North American species (P. bimundorum Soják) in its assumed parentage (see Ertter et al. 2014).

Asker, S. & Jerling, L. 1992. Apomixis in plants. – CRC Press, Boca Raton.

Elven, R. & Elvebakk, A. 1996. Part 1. Vascular plants. – In: Elvebakk, A. & Prestrud, P. (eds.), A catalogue of Svalbard plants, fungi, algae, and cyanobacteria. – Norsk Polarinstitutts Skrifter 198: 9–55.

Elven, R., Murray, D.F., Razzhivin, V. & Yurtsev, B.A. (eds.) 2011. Annotated Checklist of the Panarctic Flora (PAF) Vascular plants. http://panarcticflora.org/

Eriksen, B. & Nyléhn, J. 1999. Cases of phenotypic plasiticity in leaves of Potentilla L. (Rosaceae). – Skrifter Norske Videnskaps-Akadademi I. Matematisk–Naturvitenskapelig Klasse, n. s. 38: 191–199.

Eriksen, B. & Yurtsev, B.A. 1999. Hair types in Potentilla sect. Niveae (Rosaceae) and related taxa – terminology and systematic distribution. – Skrifter Norske Videnskaps-Akadademi I. Matematisk–Naturvitenskapelig Klasse, n. s. 38: 201–221.

Ertter, B., Elven, R., Reveal, J.L. & Murray, D.F. 2014 (2015). Potentilla Linnaeus. – In: Flora of North America Editorial Committee (eds.), Flora of North America north of Mexico. 9. Magnoliophyta: Picramniaceae to Rosaceae: 121–218.

Gentscheff, G. & Gustafsson, Å. 1940. Parthenogenesis and pseudogamy in Potentilla. – Botaniska Notiser 1940: 109–132.

Hamre, E. 2000. Variation in the genus Potentilla in Svalbard – Analysis of isozymes, micro? and macromorphometry. – Cand. scient. Thesis, Univ. Oslo, Oslo.

Hansen, K.T., Elven, R. & Brochmann, C. 2000. Molecules and morphology in concert: tests of some hypothesis in arctic Potentilla (Rosaceae). – American Journal of Botany 87: 1466–1479.

Müntzing, A. 1928. Pseudogamie in der Gattung Potentilla. – Hereditas (Lund) 11: 267–283.

Nyléhn, J. 2002. Predominant cross pollination in an alpine population of the facultative apomict, Potentilla crantzii (Crantz) G.Beck ex Fritsch, Rosaceae. – In: Nyléhn's Dr. scient. Thesis, Univ. Oslo, Oslo.

Nyléhn, J. & Hamre, E. 2002. Facultative apomixis and hybridisation in arctic Potentilla section Niveae (Rosaceae), contributions to an intricate taxonomy. – In: Nyléhn's Dr. scient. Thesis, Univ. Oslo, Oslo.

Rønning, O.I. 1961. Some new contributions to the flora of Svalbard. – Norsk Polarinstitutts Skrifter 124. 20 pp.

Rønning, O.I. 1964. Svalbards flora. – Norsk Polarinstitutt, Oslo.

Rønning, O.I. 1979. Svalbards flora. Ed. 2. – Norsk Polarinstitutt, Oslo.

Soják, J. 1986. Notes on Potentilla. I. Hybridogenous species derived from intersectional hybrids of sect. Niveae and sect. Multifidae. – Botanische Jahrbücher für Systematik 106: 145–210.

Soják, J. 1989. Notes on Potentilla (Rosaceae). VIII. P. nivea L. agg. – Candollea 44: 741–762.