Larval outbreaks in West Greenland: Instant and subsequent effects on tundra ecosystem productivity and CO2 exchange
Insect outbreaks can have important consequences for tundra ecosystems. In this study, we synthesise available information on outbreaks of larvae of the noctuid moth Eurois occulta in Greenland. Based on an extensive dataset from a monitoring programme in Kobbefjord, West Greenland, we demonstrate effects of a larval outbreak in 2011 on vegetation productivity and CO2 exchange. We estimate a decreased carbon (C) sink strength in the order of 118–143 g C m−2, corresponding to 1210–1470 tonnes C at the Kobbefjord catchment scale. The decreased C sink was, however, counteracted the following years by increased primary production, probably facilitated by the larval outbreak increasing nutrient turnover rates. Furthermore, we demonstrate for the first time in tundra ecosystems, the potential for using remote sensing to detect and map insect outbreak events.
KeywordsArctic Carbon Disturbance Ecosystem productivity Eurois occulta Insect outbreak
Arctic tundra ecosystems cover ca. 8% of the global land area. Yet, the vast stocks of organic carbon (C) stored in their soils make them especially important in a climate change context (McGuire et al. 2012), since increasing temperatures may result in increased emissions of carbon dioxide (CO2) and methane (CH4). The occurrence of periodic disturbances, such as fires, pathogens and insect outbreaks, which to a varying temporal and spatial extent damage vegetation and affect C cycling, are also likely to change in the future (Callaghan et al. 2004; Post et al. 2009). There are, however, large gaps in our understanding of how extreme events affect ecosystem functioning and they are as such generally underrepresented in process-based ecosystem models (McGuire et al. 2012).
Insect outbreaks can have extensive consequences for ecosystem productivity and functioning in subarctic and arctic biomes (Callaghan et al. 2004; Post et al. 2009). The outbreaks may lead to local and regional canopy defoliation (Tenow and Nilssen 1990; Callaghan et al. 2004; Bjerke et al. 2014), decreased vegetation biomass (Pedersen and Post 2008; Post and Pedersen 2008), shifts in vegetation composition (Karlsen et al. 2013; Jepsen et al. 2013), decreased C uptake (Heliasz et al. 2011) as well as cascading impacts through other food web compartments (Jepsen et al. 2013). The prevalence and intensity of these disturbances are expected to increase with a warmer climate (Neuvonen et al. 1999; Callaghan et al. 2004; Chapin et al. 2004). The strong warming observed in northern high latitudes (Stocker et al. 2013) has been associated with a northward extension of outbreaks of moths and their leaf-defoliating larvae in northern Fennoscandia (Post et al. 2009; Jepsen et al. 2013), likely related to enhanced survival of overwintering eggs due to warmer winters (Callaghan et al. 2004).
There are several reports of outbreaks of the autumnal moth Epirrita autumnata and the winter moth Operophtera brumata from northern Fennoscandia, occurring at roughly decadal intervals (Tenow and Nilssen 1990; Callaghan et al. 2004; Heliasz et al. 2011; Jepsen et al. 2013; Karlsen et al. 2013). The larvae of these moth species not only defoliate forests of mountain birch Betula pubescens, but have also been found to feed on understorey vegetation including dwarf birch Betula nana and bilberry Vaccinium myrtillus (Karlsen et al. 2013). During an extensive outbreak in the lake Torneträsk catchment in subarctic Sweden in 2004, the mountain birch forest was a much smaller C sink during the growing season compared with a reference year, most likely due to lower gross primary production (Heliasz et al. 2011). Furthermore, changes in light conditions caused by defoliation and nutrient additions from larval faeces and carcasses (Karlsen et al. 2013) may alter the conditions for plant species not directly affected by defoliation.
In Greenland, outbreaks of larvae of the noctuid moth Eurois occulta have occasionally been reported (see "Background" section). During the 2004–2005 outbreak in Kangerlussuaq, West Greenland, the above ground biomass of all plant functional groups was reduced by up to 90% as a result of intense defoliation (Post and Pedersen 2008). However, little is known about the frequency, timing and extent of the outbreaks of E. occulta in Greenland. The purpose of this study is therefore to synthesize available knowledge on E. occulta outbreaks in Greenland and their effects on ecosystem functioning and productivity. We were fortunate to document an outbreak of E. occulta larvae in 2011 in Kobbefjord, West Greenland, where an extensive monitoring programme has been ongoing since 2008. We aim to quantify the effects of the larval outbreak on the ecosystem productivity by analyses of monitoring data on land–atmosphere exchange of CO2 and vegetation greenness derived from an automatic camera setup. We study the effects of the larval outbreak over a longer time period including three years following the outbreak, allowing for an investigation of how the tundra ecosystem responds to the larval attack in subsequent years. Furthermore, we use satellite imagery to investigate possible historical outbreaks in the Kobbefjord catchment.
Reported outbreaks of Eurois occulta in Greenland
Fox et al. (1987)
Pedersen and Post (2008)
Pedersen and Post (2008)
Avery and Post (2013)
Avery and Post (2013)
Mølgaard et al. (2013)
Mølgaard et al. (2013)
Materials and methods
This study was conducted in Kobbefjord/Kangerluarsunnguaq in low Arctic West Greenland (64°08′N, 51°23′W, ca. 25 m a.s.l.), located ca. 20 km from Nuuk, the capital of Greenland (Fig. 1). This area is subjected to extensive monitoring and long-term research activities within the Greenland ecosystem monitoring (GEM) programme. The area is part of a valley system surrounded by mountains that reach up to ca. 1300 m a.s.l. The monitoring area covers 32 km2 and is characterised by dwarf shrub heaths intersected with dry south-facing slopes and smaller fen areas. The heaths are dominated by Salix glauca, Betula nana and Empetrum nigrum (Bay et al. 2008). Long-term (1961–1990) mean annual temperature and precipitation sum for Nuuk are −1.4 °C and 750 mm, respectively (Cappelen 2012).
Three terrestrial monitoring sub-programmes are operational in the Kobbefjord valley, namely BioBasis, GeoBasis and ClimateBasis (cf. Jensen and Rasch 2008); data from these programmes form the basis of this study. In 2008, an experiment was set up consisting of 18 control plots, six open-top ITEX chambers that increase temperature (cf. Henry and Molau 1997) and six plots with Hessian tents that reduce incoming light (Aastrup et al. 2015). In this study, only data from control plots were used.
Measurements of CO2 exchange were conducted weekly to biweekly during the snow-free season 2008–2014 using the closed chamber technique. A plexiglas measuring chamber (0.33 × 0.33 × 0.34 m), equipped with a fan for air mixing and a HTR-2 probe logging photosynthetic photon flux density and air temperature, was placed on top of a fixed metal frame for three minutes and air was analysed for CO2 concentrations using an infrared gas analyser EGM4 (PP Systems, USA). The linear change in CO2 concentration in the transparent chamber was used to calculate net ecosystem exchange (NEE), whereas a subsequent measurement in a dark chamber was used to represent ecosystem respiration (Reco). Gross primary production (GPP) was calculated as the difference between light and dark measurements (GPP = NEE − Reco).
All taxonomic groups of arthropods were sampled on a weekly basis at four sites located within a few hundred metres of the experimental plots, each with eight pitfall traps as specified by Aastrup et al. (2015). The traps contained ca. 200 ml water with one teaspoon of salt and two drops of detergent. The number of E. occulta larvae was counted at the department of Bioscience, Aarhus University, Denmark. For the purpose of this paper, samples from one site (arthropod plot 3) with vegetation composition and coverage resembling the CO2 flux plots were included in the analyses.
Soil temperatures (ST) from a depth of 1, 5, 10 and 30 cm were measured with T107 temperature probes (Campbell Sci., UK) approximately 500 m from the experimental plots. Incoming photosynthetic photon flux density (PPFD; Li-190SA, LICOR, USA) and air temperature (AT; Vaisala HMP 45D, Finland) were obtained from a weather station located ca. 2 km from the experimental plots. Daily imagery of the valley was derived from a HP E427 digital camera housed inside a weatherproof box. The box was mounted at 500 m above sea level in September 2009, and daily images were taken at noon local time (Westergaard-Nielsen et al. 2013).
The MOD13Q1 vegetation index (VI) product was used to assess the spatial and temporal differences of VI signals at locations with reported larvae outbreaks. The VI product is derived from the MODIS sensor on board the Terra satellite platform. MOD13Q1 is a 16-day composite at 250 m spatial resolution based on cloud-free observations and includes, e.g. normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) measurements (Huete et al. 2002). The data were pre-processed based on the quality assessment layer, to include only observations with VI quality down to bit 1000, however, no further processing steps to adjust for possible differences in the number of observations within each composite were taken.
The impact of larval outbreaks was examined using a window of 3 × 3 satellite pixels, with the centre pixel covering the geographical coordinates of reported outbreaks (for Kobbefjord, we used the location of the experimental plots, whereas for Kangerlussuaq, we used the coordinates for site 1 in Young et al. 2016). Time-integrated NDVI and EVI, which have been found to correlate significantly with the aboveground phytomass in the Arctic (Westergaard-Nielsen et al. 2015), were calculated using five 16-day composites during DOY 177–241 in each year. Values below 0.2 and 0.1 for NDVI and EVI, respectively, were considered erroneous and replaced by linear interpolation. This was done for Kobbefjord NDVI and EVI on DOY 193, 2004; DOY 209, 2012 and DOY 241, 2013. No values were below the thresholds for Kangerlussuaq.
Meteorological characteristics during the study period in Kobbefjord including snow characteristics (Max depth, m; DOY of melt in the CO2 flux plots, day of year), annual means (AT, air temperature, °C at 2 m; Precipitation, mm) and summer means from June, July and August (AT; Precip; PPFD, photosynthetic photon flux density, µmol m−2 s−1)
Summer (JJA) values
DOY of melt
Mean ± standard error (spatial replication) of measured transparent (net ecosystem exchange, NEE) and dark (ecosystem respiration, Reco) fluxes and estimated budgets of gross primary production (GPP), Reco and NEE during June–August 2008–2014 in Kobbefjord. Superscript letters for the budgets columns, derived from a Tukey’s HSD multiple comparison, indicate significant differences among years. Numbers in parentheses reflect sample size (i.e. number of plots)
Measurements (mg CO2 m−2 h−1)
Budgets (g C m−2)
−358 ± 81
262 ± 12
– ± –
– ± –
– ± –
−497 ± 88
216 ± 13
−252 ± 39abc (14)
100 ± 5a (16)
−150 ± 40abc (12)
−496 ± 77
288 ± 14
−244 ± 21ab (18)
150 ± 7b (18)
−94 ± 17ab (18)
−86 ± 32
236 ± 9
−104 ± 14a (10)
110 ± 3a (17)
7 ± 16a (10)
−1023 ± 143
406 ± 21
−399 ± 49bcd (13)
200 ± 10 cd (18)
−216 ± 44bc (13)
−1164 ± 178
425 ± 27
−424 ± 44 cd (18)
182 ± 10bc (18)
−242 ± 37bc (18)
−1119 ± 160
479 ± 28
−505 ± 65d (18)
231 ± 13d (18)
−274 ± 54c (18)
Estimated flux components (GPP; gross primary production, Reco; ecosystem respiration, NEE; net ecosystem exchange, g C m−2) in 2011 using meteorological data (photosynthetic photon flux density, PPFD and soil temperature, ST) from 2011 and parameterisations of Eqs. 1 and 2 from other years (2009, 2010, 2012–2014)
−196 ± 31
84 ± 5
−111 ± 31
−180 ± 16
119 ± 6
−61 ± 12
−280 ± 34
158 ± 10
−136 ± 30
−366 ± 37
170 ± 10
−196 ± 31
−430 ± 57
205 ± 12
−226 ± 46
Area coverage from camera-based classification in Kobbefjord 2011
Total area coverage (km2)
Fraction of classified area
Number of segments
Average area per segment (m2)
The outbreak of the noctuid moth E. occulta in Kobbefjord in 2011 had a strong and extensive impact on the vegetation. The production of leaves, buds and catkins or flowers in all species monitored in the area was heavily impacted and the vegetation reproduction was seriously reduced (Bay et al. 2012). Also, in the year following the outbreak, 2012, the total catkin and flower production was low (Bay et al. 2013), indicating that the plants focused their resources on establishing new leaves; a compensatory growth as a response to herbivory (McNaughton 1983). The excess energy stored in 2012 from not producing catkins resulted in a record amount of catkins in 2013 (e.g. Fig. 4).
The larvae did not forage upon leaves from all plant species, e.g. Empetrum nigrum was generally left untouched, although larvae feeding on E. nigrum flowers were observed in the field. E. nigrum is generally unpalatable to herbivores and not directly defoliated during moth outbreaks (Jepsen et al. 2013; Karlsen et al. 2013); however, previous studies have discussed the possibility that starving larvae attempt to eat their leaves making them more susceptible for desiccation or infection (Jepsen et al. 2013; Karlsen et al. 2013; Olofsson et al. 2013). Plot-scale NDVI measurements showed that E. nigrum was notably less green in 2011 than in other years (Olsen et al. 2014), which could thus be explained by the larval outbreak. However, other adverse effects such as frost damage (Bjerke et al. 2014) during early winter 2010/2011 cannot be excluded.
In the following years, E. nigrum was greener than in previous years, also before the outbreak (Olsen et al. 2014). This pertained to all plant species as seen by generally higher NDVI values measured after the outbreak; an indication of good health, which may stem from higher levels of nutrients made available for the plants from decomposed larvae. Arctic tundra vegetation is generally nutrient limited (cf. Chapin and Shaver 1985) and larvae faeces and carcasses can provide a nutrient pulse to the system (Kagata and Ohgushi 2012). Also, reduced plant nutrient uptake during 2011, as a consequence of reduced growth, may have resulted in excess nutrient availability in the following years. Post and Pedersen (2008) report a fourfold increase in nitrogen (N) concentration in leaf tissues of S. glauca and B. nana at the peak of the larval outbreak in Kangerlussuaq (Table 1) along with a rapid biomass recovery following the outbreak.
There was a marked decrease in CO2 fluxes in 2011, both in terms of instantaneous, measured fluxes (Fig. 5) and estimated summertime budgets (Table 3). Net ecosystem exchange was close to zero during June–August in 2011 (Table 3), indicating that the heath ecosystem, represented by the experimental plots, did not accumulate C during this period. The reduction in CO2 accumulation during 2011 was caused by a significant decrease in GPP to less than half of that in other years (Table 3). However, fluxes were higher in the years following the outbreak, again indicating a rapid ecosystem recovery after the larvae attack. The increase in GPP in 2012–2014 may be explained by an increase in nutrient availability due to the larval outbreak, as discussed above, favouring vegetation growth in subsequent years.
The rapid regrowth and the increase in primary productivity indicate that the tundra ecosystem may not be as vulnerable as anticipated with respect to these outbreaks. The ecosystem may have developed a high degree of resiliency as a response to outbreak events occurring at certain intervals. Our findings, that the years following the outbreak (2012–2014) had higher GPP and stronger C sink strengths compared with the years preceding the outbreak (2009–2010), correspond to the strong biomass recovery observed in Kangerlussuaq following a larval outbreak (Post and Pedersen 2008). This indicates that the effects of outbreaks may be counterbalanced by increased primary production in the following years. Further, the larvae appear to play a significant role by influencing nutrient dynamics and accelerating N turnover.
By using the parameterisations from other years, we estimated what the CO2 exchange in 2011 would have been in the absence of the larval outbreak (Table 4). This approach takes inter-annual variation in meteorological characteristics into account, e.g. the impact of long-lasting snow cover in 2011 is included in these estimates. Parameters from all other years resulted in higher (i.e. more negative) GPP and stronger C sink strength, whereas the effect on Reco was less consistent. The parameters from 2010 resulted in lower (less negative) GPP sum and weaker C sink strength compared with other years, which can be associated with the modest number of moth larvae affecting ecosystem productivity also in 2010 (Fig. 3). The parameters from 2013 and 2014 resulted in high GPP sums and strong C sink strengths, indicating a strong recovery from the larval attack and a potential switch to a more productive state as discussed above. As such, when assessing the effect of the larvae on vegetation productivity and C budget in 2011, it seems most reasonable to use parameters from 2009 and 2012. Although a simplification, it results in a decrease in C sink strength in the order of 118 to 143 g C m−2, with an associated uncertainty (combined standard error from the spatial replication) of approximately ± 47 g C m−2. Scaling to the field of view of the camera (Fig. 2), taking the whole area of attacked heath into account (Table 5), results in a C loss of 98–119 tonnes C. However, the camera covers <10% of the entire Kobbefjord catchment (2.58 out of 32 km2) so the catchment scale C loss may be one order of magnitude larger (approximately 1210–1470 tonnes C). This approximation can be compared with a study in northern Sweden by Heliasz et al. (2011), who estimated a C loss of 29 000 tonnes C for a mountain birch forest area of 316 km2 exposed to an outbreak of larvae in 2004 of the autumnal moth E. autumnata.
We demonstrate the potential for using satellite imagery to detect and map insect outbreaks in West Greenland. To our knowledge, this is the first time satellite data have been used to observe effects of insect outbreaks in tundra ecosystems. Both NDVI and EVI show a clear decrease in time-integrated values during 2011 in Kobbefjord, with a decrease of 15 and 26%, respectively, compared with the 2000–2014 mean (Fig. 6). A decrease of 16% in time-integrated GCC in 2011 matches this range. These estimates also match a 16–27% decrease in peak NDVI in 2012 in northern Fennoscandia during a moth larvae outbreak, compared with a 2000–2011 average (Bjerke et al. 2014). The outbreaks in Kangerlussuaq in 2004–2005 and 2010–2011 (Pedersen and Post 2008; Avery and Post 2013; Table 1) are also visible through low values of time-integrated indices (Fig. 6); however, the picture is less clear for this site with low values also in other years.
It might be suggested that outbreaks of E. occulta in West Greenland occur in synchrony since the satellite data indicate low NDVI and EVI in Kobbefjord also in 2004; a year with documented outbreak in Kangerlussuaq (Pedersen and Post 2008). However, as the current study is limited to the near vicinity of field investigations in the respective site, impacts on a larger scale can only be speculated upon. Nevertheless, spatial synchrony in outbreak events indicates that climatic variations play a key role in triggering outbreak events (Klemola et al. 2006; Young et al. 2014). Outbreaks of Epirrita autumnata and Operophtera brumata in northern Fennoscandia have been associated with reduced egg mortality during warm winters (Tenow and Nilssen 1990; Callaghan et al. 2004; Chapin et al. 2004; Young et al. 2014) as well as with decreased parasitoid and predator activity because of low spring and summer temperatures (Virtanen and Neuvonen 1999; Callaghan et al. 2004). However, other studies have found a positive relationship between moth outbreaks and spring and summer temperatures (Klemola et al. 2003; Young et al. 2014). In our study in Kobbefjord, the winters (December–February) of 2009/2010 and 2010/2011 were indeed the warmest on record with mean temperatures of −2.7 and −4.5 °C, respectively, compared with a 2008–2014 mean of −6.7 °C (Table 2). Also, the summer of 2011 was relatively cold. It is also worth mentioning the importance of snow, which plays a key role in regulating Arctic ecosystem functioning (cf. Callaghan et al. 2012). There was a thick and long-lasting snow pack in the winter 2010/2011 in Kobbefjord (Table 2), which insulated and protected overwintering E. occulta larvae from low winter temperatures. In line with this argumentation, it can be noticed that heath vegetation on mountain slopes does not appear to be affected by the larvae (Fig. 2); these areas are generally colder and covered by less snow than the lowlands.
Our results indicate a marked decline in summertime C uptake during an outbreak of the larvae of E. occulta in 2011 in Kobbefjord. However, the years following the outbreak (2012–2014) were characterised by stronger C uptake compared with the years preceding the outbreak. This indicates that the ecosystem is well adapted to these outbreaks and that they presumably occur at certain intervals if a number of environmental conditions are fulfilled. As a consequence of the outbreaks, nutrient turnover rates increase and growth is favoured in subsequent years. As such, the outbreaks may facilitate ecosystem rejuvenation (Tenow et al. 2004).
Future studies should focus on developing tools based on remote sensing products such as the vegetation indices used here for mapping larval outbreak events in West Greenland. A spatially distributed dataset of outbreak events, as opposed to occasional observations, would be highly useful for comparisons with gridded climate data. However, satellite data can only provide landscape scale information on net effects. There is thus an urgent need to continue and expand upon in situ environmental monitoring efforts in the Arctic, in order to improve upon the process-based understanding of how climate change and associated changes in extreme events such as insect outbreaks may affect tundra ecosystem functioning and dynamics. Predicting extreme events, e.g. larval outbreaks, is difficult so in order to capture the events continuous, long-term monitoring programmes are required.
Data for this study were provided by the Greenland Ecosystem Monitoring Programme. A special thanks to Maia Olsen and other BioBasis’ field assistants for their support in the field. This work was supported by the Danish Energy Agency, Danish Environmental Protection Agency and Nordic Center of Excellence eSTICC (eScience Tools for Investigating Climate Change in northern high latitudes) funded by Nordforsk (Grant 57001).
- Aastrup, P., J. Nymand, K. Raundrup, M. Olsen, T.L. Lauridsen, P.H. Krogh, N.M. Schmidt, L. Illeris, et al. 2015. Conceptual design and sampling procedures of the biological programme of NuukBasic. 2nd edition. Aarhus University: DCE—Danish Centre for Environment and Energy.Google Scholar
- Bay, C., P. Aastrup, and J. Nymand. 2008. The NERO line. A vegetation transect in Kobbefjord, West Greenland. Aarhus University: National Environmental Research Institute.Google Scholar
- Bay, C., K. Raundrup, J. Nymand, P. Aastrup, P.H. Krogh, T.L. Lauridsen, L.S. Johansson, M. Lund, et al. 2012. Nuuk Basic: The BioBasis programme. In Nuuk Ecological Research Operations, 5th Annual Report, ed. Jensen, L.M. 33–46. Aarhus University: DCE—Danish Centre for Environment and Energy.Google Scholar
- Bay, C., J. Nymand, P. Aastrup, K. Raundrup, P.H. Krogh, T.L. Lauridsen, M. Lund, K. Albert et al. 2013. Nuuk Basic: The BioBasis programme. In Nuuk ecological research operations, 6th Annual Report, ed. Jensen, L.M. and M. Rasch, 33–46. Aarhus University: DCE—Danish Centre for Environment and Energy.Google Scholar
- Bjerke, J.W., S.R. Karlsen, K.A. Høgda, E. Malnes, J.U. Jepsen, S. Lovibond, D. Vikhamar-Schuler, and H. Tømmervik. 2014. Record-low primary productivity and high plant damage in the Nordic Arctic Region in 2012 caused by multiple weather events and pest outbreaks. Environmental Research Letters. doi:10.1088/1748-9326/9/8/084006.Google Scholar
- Cappelen J., P.R. Wang, M. Scharling, R.S. Thomsen, L. Boas, K. Vilic and Stendel M. 2012. Danmarks klima 2011 med Tórshavn, Færøerne og Nuuk, Grønland. Teknisk rapport 12-01. Copenhagen: Danish Meteorological Institute (in Danish, English summary).Google Scholar
- Heliasz, M., T. Johansson, A. Lindroth, M. Mölder, M. Mastepanov, T. Friborg, T.V. Callaghan, and T.R. Christensen. 2011. Quantification of C uptake in subarctic birch forest after setback by an extreme insect outbreak. Geophysical Research Letters 38: L01704. doi:10.1029/2010GL044733.CrossRefGoogle Scholar
- Iversen, J. 1934. Moorgeologische untersuchungen auf Grönland. Meddelelser fra Dansk Geologisk Forening 8: 341–358. (In German).Google Scholar
- Jensen, D.B. 2003. The biodiversity of Greenland - a country study. Technical Report No. 55. Nuuk: Pinngortitaleriffik, Grønlands Naturinstitut.Google Scholar
- Jensen, L.M., and M. Rasch. 2008. Nuuk ecological research operations, 1st Annual Report, 2007. Copenhagen: Danish Polar Centre.Google Scholar
- Karsholt, O., N.P. Kristensen, T.J. Simonsen, and M. Ahola. 2015. Lepidoptera (Moths and butterflies). In The Greenland entomofauna: An identification manual of insects, spiders and their allies, ed. J. Böcher, N.P. Kristensen, T. Pape, and L. Vilhelmsen, 302–352. Leiden: Brill.CrossRefGoogle Scholar
- McGuire, A.D., T.R. Christensen, D. Hayes, A. Heroult, E. Euskirchen, J.S. Kimball, C. Koven, P. Lafleur, et al. 2012. An assessment of the carbon balance of Arctic tundra: Comparisons among observations, process models, and atmospheric inversions. Biogeosciences 9: 3185–3204. doi:10.5194/bg-9-3185-2012.CrossRefGoogle Scholar
- Mølgaard, P., K. Christensen, M.S. Vöge, G.B. Christensen, and K.V. Ommundsen. 2013. Confirm the presence of Eurois occulta larvae on Disko. In Arctic Station—Annual Report 2013, 25–26, ed. T.W. Perlt, and K. Christoffersen. Copenhagen: Faculty of Science, University of Copenhagen.Google Scholar
- Neuvonen, S., P. Niemelä, and T. Virtanen. 1999. Climatic change and insect outbreaks in boreal forests: The role of winter temperatures. Ecological Bulletins 47: 63–67.Google Scholar
- Olsen, M., J. Nymand, K. Raundrup, P. Aastrup, P.H. Krogh, T.L. Lauridsen, M. Lund and K. Albert. 2014. Nuuk Basic: The BioBasis programme. In Nuuk Ecological Research Operations, 7th Annual Report, ed. Jensen, L.M. and T.R. Christensen, 34-45. Aarhus University: DCE—Danish Centre for Environment and Energy.Google Scholar
- Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, et al. 2013. Climate change 2013: The physical science basis. Cambridge: Cambridge University Press.Google Scholar
- Vibe, C. 1971. Lavere dyr i Grønland [Smaller animals in Greenland]. In Danmarks Natur, ed. T.W. Böcher, C.O. Nielsen, and A. Schou, 444–452. Copenhagen: Politikens Forlag. (in Danish).Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.