Background and aim

Not only has Beringia provided a landbridge connecting Alaska and Siberia, but it was also a glacial refugium during the last glaciation. Beringia is therefore central for our understanding of present day arctic ecosystems. However, since this event the ecosystems on each side of the Bering Strait have diverged, and the Beringia 2005 expedition provides an unique opportunity to compare the genetic, taxonomic and functional diversity in ecosystems on both sides of the strait.

Our knowledge regarding the distribution, diversity and regulation of willows and insects in the Arctic is generally very limited. To understand ecosystem function in the Arctic we need more basic knowledge on factors influencing the distribution and regulation of populations in the Arctic. The aim of this project is to compare the diversity, distribution and trophic interactions in willow–insectparasite systems on both sides of the Bering Strait and along a latitudinal gradient.

Willows are dominant plants in the Arctic and one of the few deciduous wood plants that occur up to the High Arctic (Danks 1981, Hjältén et al., 2003). Because of this they are a very important host for many herbivores in the Arctic (Danks, 1986). Many willows have a circumpolar distribution and may differ in plant characteristics in different parts of the distribution range (e.g. Nearctic, a zoogeographical region comprising Greenland and North America, and Palaearctic, a zoogeographical region comprising Eurasia north of the Himalayas, North Africa and the temperate part of the Arabian Peninsula) (Argus et al., 1999, Skvortsov 1999). These differences are likely to have led to local adaptations in some insect herbivores, since host-associated evolutionary changes can be very specific and occur rapidly in some species, such as sawflies (Hymentoptera tenthredinidae) (Nyman, 2000). Galling sawflies are very suitable for studying plantherbivore–parasite interactions due to a high degree of specialisation and host specificity (Zinojev 1998, Roininen et al., 2005) and because it is very simple to determine growth and survival of larvae inside the galls and levels of parasitism by dissection of the galls.

Experience from the expedition Tundra Northwest 1999 (TNW99) suggests that galling sawflies are essentially absent, or at least very rare, in the High Arctic (above 70°N) in the Nearctic, despite the fact that their main host willows occurred up to the High Arctic (Hjältén et al., 2003). There are indications that sawflies occur further north in the Palaearctic (Roininen et al., 2002). The Beringia 2005 expedition gave an excellent opportunity to study biogeographical patterns in the distribution and regulation of sawflies in a latitudinal gradient on both sides of the Bering Strait and from Japan (data sampled earlier) to the Chukotka Peninsula and Alaska.

Another important arctic herbivore is the woolly bears, Gynaephora groenlandica and G. rossii, which belong to the Lymantriidae family and are real arctic specialists. The developmental time of the larva is long, up to 14 years, and they seem to be able to survive bad summers almost without eating and with very little weight loss (Danks, 1983, 1986). They sometimes reach extremely high densities, up to 26.4/m² or 264 000/ha (Danell et al., 1999), making them the completely dominant insect herbivores in some habitats. However, our knowledge of their food preferences is limited. They are generally regarded as polyphagous (Danks, 1981) but their food preferences have rarely been examined (however see MacLean and Jensen, 1985).

The importance of parasitoids/predators for regulation of insect populations in the Arctic has been poorly investigated. It has been hypothesized that sawflies are mainly regulated by bottom up effects (Price et al., 1994, 1996). However, recent studies from the Palaearctic indicate that parasitoids are important mortality factors in populations of galling sawflies (Roininen et al., 2002). The Beringia 2005 expedition provided a unique opportunity to address questions regarding the regulation of sawfly populations in the Palaearctic and Nearctic.

Preliminary results and discussion

1.The distribution and diversity of willows, insects and parasitoids

To determine the biogeographical patterns in the distribution of willows and gallers, we identified the willow species present, determined their abundance and the densities of sawflies on each field site (for methods see Roininen et al. 2002, Hjältén et al., 2003). We were able to collect data from nine sites, three in Alaska (Denali Highway and Barrow (two sites)) and six in Russia (Novoye Chaplino, Penkigney Bay, Yanrakinot, Lavrentia, Toygunen and Wrangel Island). Willows were found at all these sites and we collected material for analyses of DNA and chemistry from eleven willow species (Salix chamissonis, S. reticulata, S. polaris, S. arctica, S. fuscescens, S. ovalifolia, S. lanata, S. glauca, S. rotundifolia, S. phlebophylla, S. hastata). Galling sawflies were found on all sites except Wrangel Island (but earlier observations suggest that gallers are present there). This suggests that both willows and their associated gallers are widely, albeit in the case of gallers (in our experience) unevenly and patchily distributed in Beringia. Parasites on galling sawflies were also present at all sites where the sawflies were found, seemingly more common and diverse at lower latitudes. However these data have not yet been analysed. Nevertheless it is likely that these unique data will provide further insight in species distribution and associations of willows, galling sawflies and their parasitoids in Beringia.

2. Host plant use of the arctic woolly bears, Gynaephora rossii

We collected information regarding the host plant use of Gynaephora rossii in the field. In addition we collected 400 individuals and did no-choice feeding experiments using 27 different plant species collected in the field. We identified three responses of the larvae: No feeding=0, tasting=1 (few bites consumed) and full acceptance=2 (most or all of the leaves consumed). The highest density of woolly bears was found on Wrangel Island and we therefore used that population in the experiment. We used 27 plant species belonging to 17 plant families in the test and presented leaves from each plant species to 10 woolly bear larvae. The results showed that the woolly bears were less generalistic than expected. They only fully accepted 8 plant species (a majority – 5 species – being willows) as food, and these came from 3 plant families (Salicaceae, Rosaceae, Crassulaceae) (photo). This suggests that the arctic woolly bear probably utilizes less plant species than earlier expected. However, more detailed analyses of e.g. plant chemistry and performance of woolly bears on different plant species is needed to find the proximate reason for this host acceptance pattern.

3. Regulation of arctic sawfly populations in the Palaearctic and Nearctic

To determine regulatory/mortality factors and the diversity of parasites and predators, 100–600 galls were collected on around 25–100 randomly selected individuals per species (sample size depends on the size of the willows and the density of the gallers). The galls were opened and the survival rate and relative importance of different mortality factors, e.g. plant induced mortality and mortality due to different types of parasites and inquilines were determined (for methods see Roininen et al. 2002, Hjältén et al. 2005 manuscript). We collected and dissected 4 000–5 000 galls during the expedition to determine mortality factors and parasitism on the sawfly larva. We identified 12 different mortality factors for the larvae, 3 being related to plant responses and the rest due to parasites. The preliminary results show that the degree of parasitism is very high – around 80–90% in several of these galler populations – suggesting that parasites could play a central role in the regulation of these sawfly populations. The diversity of parasites was also high, but due to limited knowledge on these groups of insects in the Arctic we have not yet determined how many parasitoid species were utilizing sawfly gallers. However future analyses of these data will give more information on the role of parasitoids existing in populations of arctic galling sawflies.

4. Genetic diversity and chemotaxonomy and plant–animal interactions

To determine the biogeographical patterns in plant genetic and chemical diversity, we randomly selected 10 willow individuals per species on each site. However, for this purpose we specifically use species with a circumpolar distribution (distributed on both sides of the Bering Strait), e.g. Salix polaris, S. arctica, S. phlebophylla, S. pulchra and S. reticulata (Porsild and Cody, 1980, Skvortsov, 1999). From each individual we collected leaf and stem material for DNA analyses, analyses of plant chemistry and morphology. We will soon start conducting the molecular and chemical analyses on the 11 willow species collected during the expedition. This material, together with data collected from the Swedish-Russian Tundra Ecology expedition 1994 and the TNW 99 expedition, provides a unique opportunity to study the genetic and chemotaxonomic divergence between willow communities on both sides of the Bering Strait.

Conclusions

Despite changes in the route and logistic problems during the expedition, it provided an excellent opportunity to study ecological and evolutionary questions. Our results so far show that both willows and galling sawflies are widely distributed in the Beringian tundra biome. It also shows that the parasitic community utilizing galling sawflies is highly diverse and probably plays an important role in the regulation of sawfly communities. We also found that the arctic woolly bear probably utilizes less plant species than earlier assumed.