**4. Dispersion: How to get there?**

Pioneer invertebrate species must be good dispersers, but our knowledge in this field is limited. The easiest dispersion would be by air, either by active flight, by passive wind transport, or a combination. On Svalbard, areal dispersal of invertebrates over the foreland of Midtre Lovénbreen glacier was studied by Coulson et al. (2003). Large numbers of Diptera, Hymenoptera and Araneae were caught in water and sticky traps, some only 15 m

Primary Succession in Glacier Forelands:

large Opiliones of the genus *Mitopus,* mentioned above.

pioneer species have to be ecologically similar.

**5.1 Specialists or generalists?** 

*Bourletiella hortensis* (Hågvar, 2010).

**5.2 Parthenogenetic or bisexual?** 

extra flexibility.

dispersal of individuals.

How Small Animals Conquer New Land Around Melting Glaciers 159

Most beetles can fly, but not all. Not surprisingly, typical pioneer beetles are generally fully winged, while non-flying species are late colonisers in glacier foreland (Gobbi et al., 2007). However, dispersing by foot may be efficient in certain very active invertebrates, as the

The combined knowledge today indicates that pioneer communities in glacier forelands are rather predictable, and that dispersal may not seriously restrict community development (e.g. Hodkinson et al., 2004; Kaufmann, 2001; Vater, 2006). However, this does not mean that

Pioneer invertebrates in European glacier forelands comprise both specialists and generalists. Even specialists represent a heterogeneous group, depending on their speciality. Some are "cold-loving", represented by the springtail *Agrenia bidenticulata* (Hågvar, 2010) and certain carabid beetle species of the genus *Nebria*, for instance *Nebria nivalis* (Bråten & Flø, 2009; Gobbi et al., 2007; Kaufmann, 2001; Vater, 2006). Such cold-adapted species may increase their distribution if the area of pioneer ground increases due to an increased melting rate, but may eventually disappear locally if the glacier or snow field melts away. A second group of specialists are those preferring open, barren ground. Some of these, both among microarthropods, beetles and spiders, have an alpine and/or arctic distribution. However, some also occur in lowland areas on various sandy, gravely or stony habitats, for instance carabid species of the genus *Bembidion* (Bråten & Flø, 2009) or the springtail species

Ecological generalists from several taxonomic groups are found in pioneer communities. These species tolerate a wide range of habitats, both in the lowland and in mountains. Examples from European glacier forelands are the carabid beetle *Amara quenseli*, the harvestman *Mitopus morio* , the springtail *Isotoma viridis,* and the oribatid mite *Tectocepheus velatus* (Bråten & Flø, 2009; Hågvar, 2010; Hågvar et al., 2009; Hodkinson et al., 2004; Kaufmann, 2002; Vater, 2009). An interesting point is that is rather predictable which "generalists" are present among the pioneers, in the same way as the specialists are predictable. Clearly, only a few "generalists" can extend their ecological niche far enough to thrive on pioneer ground close to a glacier – including the ability to arrive there. Later successional stages may contain several other "generalist species", but they do not have this

Some springtails and mites are parthenogenetic, which means that one single individual can start a local population. This ability is an obvious advantage for a pioneer species if dispersion is a limiting factor. In a glacier foreland in south Norway, Hågvar et al. (2009) found that the two characteristic pioneer mites were parthenogenetic. However, parthenogenetic mites were found along the whole foreland gradient, including some slowdispersing species. Among springtails in the same foreland, the pioneer species were mainly bisexual (Hågvar, 2010). Therefore, among microarthropods, parthenogenesis is not more typical among pioneer species than among later colonisers. This may indicate efficient

**5. Ecological similarities and differences between pioneer invertebrates** 

from the glacier snout on 2 year old ground. Sticky traps were placed either just above ground level, or at a height of 1 m. It was concluded that spiders caught 1 m above ground must have been aerially dispersed. The actual spider family, Linyphiidae, is known for their ability to fly by wire, called "ballooning". By raising the abdomen and gradually releasing a thread in the breeze, the spider is finally lifted upwards and can be blown very far away. Holm (1958) suggested that many spider species on Svalbard had originally arrived from Greenland as aerial plankton. The airborne Diptera and Hymenoptera in the glacier foreland represented a food source for the spiders. Another interesting observation was that more than 95 % of the animals caught in sticky traps were taken close to the ground, and very few at 1 m height. The vast majority of animals were dispersing at, or below, 0.25 m. Furthermore, animals were trapped from all directions, despite some prevailing wind directions during the study. It was concluded that these arctic insects appear to make flights of short duration and remain close to the ground where wind velocities are considerably reduced and air temperatures elevated. This behaviour enables them to perform directional flight largely independent of wind direction (Coulson et al., 2003).

Although several springtails and mites are early colonisers on Svalbard, Coulson et al. (2003) did not catch these groups in the sticky traps. Later, Mangnussen (2010) achieved some springtails and mites in water traps on Svalbard outside glaier forelands. His traps had a sticky rim to avoid crawling into the trap. Interestingly, the airborne transport of springtails seemed to occur at low wind speeds and in periods with high air humidity, indicating a high surface activity during such conditions. Since springtails can jump, they may be taken further by air currents. In Alaska, wind-blown springtails and mites have been collected in suspended plankton nets (Gressitt and Yoshimoto 1974). Elsewhere, they have even been taken as aerial plankton at altitudes of 1,500 m (Glick, 1939; Riley et al., 1995), so local dispersion by wind seems likely.

Fig. 6. Sticky trap illustrating activity of various Diptera on a 3 year old moraine at Midtdalsbreen glacier snout, Norway. Photo: Sigmund Hågvar.

Most beetles can fly, but not all. Not surprisingly, typical pioneer beetles are generally fully winged, while non-flying species are late colonisers in glacier foreland (Gobbi et al., 2007). However, dispersing by foot may be efficient in certain very active invertebrates, as the large Opiliones of the genus *Mitopus,* mentioned above.

The combined knowledge today indicates that pioneer communities in glacier forelands are rather predictable, and that dispersal may not seriously restrict community development (e.g. Hodkinson et al., 2004; Kaufmann, 2001; Vater, 2006). However, this does not mean that pioneer species have to be ecologically similar.
