OUP user menu

Thermal Adaptations of the Leaf Beetle Chrysomela lapponica (Coleoptera: Chrysomelidae) to Different Climes of Central and Northern Europe

(CC)
Jürgen Gross, Erik Schmolz, Monika Hilker
DOI: http://dx.doi.org/10.1603/0046-225X-33.4.799 799-806 First published online: 1 August 2004

Abstract

This study investigated the physiologically adaptive significance of elytral coloring and other thermal adaptations to different climes of Central and Northern Europe in a leaf beetle. Adults of Chrysomela lapponica L. differ in the coloring pattern of their red and black elytra when comparing populations from Finnish Lapland (Northern Europe) and from the Czech Republic (Central Europe). While the elytra of adults of the Czech population were bright red in color with a proportion of 51% of black marks, the relative size of the black marks of the Finnish beetle’s elytra was significantly higher (68%). When exposed to light, the dark morphs from Finland increased their body temperature within a shorter time and reached higher mean body temperatures than the lighter morphs from the Czech Republic. The walking speeds of the two different morphs did not differ at temperatures of 20, 15, and 10°C. However, at 5°C, the walking speed of the melanic beetles from Finland was significantly higher. The specific metabolic rate during overwintering of the adults was significantly higher in the Finnish morph at a temperature of 5°C compared with the Czech morph. The average and maximum temperatures during overwintering in the leaf litter layer in the Czech Republic were more variable and on average higher than under the snow cover in Finland. These results on the thermoregulation in C. lapponica are discussed with special respect to melanic elytral coloring as adaptation toward lower ambient temperatures.

  • temperature
  • Chrysomela lapponica
  • color morph
  • thermal balance
  • calorimetry

Introduction

Melanism is widespread in insects living at low temperatures. Increased pigmentation has been observed in arctic, antarctic, and alpine insects (Somme and Block 1991, Lopatin 1996, Medvedev 1996, Gross 1997, Jolivet and Verma 2002). The highest levels of melanisation in leaf beetles were observed among high altitude species whose counterparts in lower altitudes are paler with dark spots and stripes (Lopatin 1996). Muggleton (1978) suggested that elytral coloring influences thermoregulation in beetles, because melanic morphs are expected to absorb moreradiation and heat up faster to higher body temperatures. Hence, they should have a selective advantage at lower temperatures(Lusis 1961). In fact, the difference in coloring between melanic and nonmelanic morphs seemed to be the principal factor influencingthe activity of the coccinellid Adalia bipunctata (De Jong et al. 1996). Additionally, melanism may protect against the transmission of UV-β radiation (Strathdee and Bale 1998). However, the effect of pigmentation on the thermoregulation and UV-β protection has only been tested in few insectspecies, which were mainly Lepidoptera (Watt 1968, 1969, Brakefield and Willmer 1985, Kingsolver 1985, Fields and McNeil 1988, Kingsolver 1988, Kingsolver and Wiernasz 1991, Goulson 1994, De Jong et al. 1996), of which the melanic morphs did not live in arctic or mountainous environments.

The leaf beetle Chrysome lalapponica L. is distributed in central and northern Europe, forming distinct allopatric populations, which are specialized either onbirch or willow (Gross and Hilker 1995, Zvereva et al. 1995, Gross 1997). Some populations differ in several chemical, morphological, and ecological traits (Hilker and Schulz 1994, Gross and Hilker 1995, Gross 2001, Gross et al. 2002). A very striking difference is evident when considering the elytral coloring. Large black marks on the elytra are characteristicfor adults of willow-feeding populations from northern Europe (Finland, Norway, and Russia) and from a higher mountainousregion of western Europe (Massive Central, France), where temperatures during overwintering are very low but stable (Zvereva et al. 1995, 1998, 2003, Gross 1997, 2001, Milyashevich 2000). In contrast to this dark morph, adults from birch-feeding populations in central Europe (Germany, Czech Republic, and Poland),which are distributed at lower altitudes, have brightly red elytra with only smaller black marks (Gross and Hilker 1995, Gross 1997, Gross et al. 2002). This light morph needs to cope with very variable temperatures during overwintering.

The different populations of C. lapponica from Finland and the Czech Republic live in environments that differ in their temperature conditions and in the length oftheir vegetation period. In this study, we hypothesized that the melanic elytral coloring is advantageous for those C. lapponica who live at very low temperatures and need to optimize the use of warming radiation. Thus, we compared the increase of bodytemperatures of a melanic Finnish morph and a nonmelanic Czech morph after exposure to light. To gain insight into the translationof body temperature into activity, we measured walking activities of the melanic and nonmelanic beetles at different temperatures.Furthermore, specific metabolic rates of the two morphs were compared to elucidate whether they also adapt their metabolismto the climate conditions of their habitats.

Materials and Methods

Leaf Beetles.

Adults of C. lapponica were collected from the willow Salix borealis in northern Europe (69°45' N; 27°01' E) in the surroundings of the Kevo Subarctic Research Institute,Finland. The area was located ≈60 km north of the continuous pine forest line and 80 m above sea level. Individuals ofa second population were collected on the birch Betula pubescens in central Europe (50°00' N; 12°40' E) in the Czech Republic. The collection site was in the surroundingsof Marianske Lazne and located 850 m above sea level. Field-collected individuals of each population were reared separatelyin the laboratory in plastic boxes (15 by 13 by 6 cm), which were lined at the bottom with moistened paper towels at 15°C,70% RH, and a 16/8 light/dark cycle (Czech beetles) or constant light (Finnish beetles). The beetles were fed with leavesof their natural host plants.

Coloring of the Elytra.

To quantify the degree of melanization, the elytra of 40 randomly chosen adults of each C. lapponica population were examined for their coloration. The areas of the red and black spots were determined by covering the elytrawith a transparent plastic foil with an imprinted grid (0.1 by 0.1 mm). A binocular (M5A; Wild Heerbrugg, Switzerland) wasused for measuring the areas. The degree of melanization was calculated as percentage area with black coloring of the elytralsurface.

Body Temperature.

To determine body temperature with respect to body weights, adult beetles of the melanic population from Finland and of thelighter-colored population from the Czech Republic were weighed using a precision balance (Sartorius MC 210 S, Sartorius,Göttingen, Germany).

Measurements of Beetle Body Temperature in the Laboratory.

After weighing, the beetles were killed by freezing them at -25°C. A small hole was pierced into the ventral sideof the thorax with a sharpened needle. The needle was immediately replaced by a thermocouple (Ni-CrNi, diameter 0.2 mm; Linseis,Selb, Germany), which was inserted 2 mm into the thorax. To minimize measurement errors caused by heat conduction, the beetlewas held only by the thermocouple, which was glued into the tip of a micropipette. The dorsal side of the beetle was irradiatedat a distance of exactly 10 mm by two light-conducting fibers (Olympus, Hamburg, Germany) connected to a halogen lamp (150W, 6428; Philips, Hamburg, Germany). Constant distance between the beetle and source of radiation was achieved through a frameonto which the ends of the light-conducting fibers and the thermocouple were attached. The beetle was surrounded by a verticalplastic tube (60 by 25 mm), which was also fixed to the frame to prevent artifacts caused by heat convection. Temperatureswere measured by a digital thermometer (T-DLIN-U; Linseis) coupled with a chart recorder (L6532; Linseis). All measurementswere conducted in a climate chamber at 10.0°C and 90% RH to prevent desiccation of the beetles during the measurements.When the initial body temperature had stabilized at 10.0°C, the halogen lamp was switched on, and the temperature wasrecorded continuously for 5 min. Body temperatures of nine adults of each population were measured. In contrast to the otherexperiments, where adults were selected by chance to represent the possible range of variation, for these experiments we triedto select adults of similar size to get information on body temperature independent of size and weight.

Measurements of Beetle Body Temperature in the Field.

To assess body temperatures at field conditions (light/shadow), measurements were conducted during a sunny day in a publicpark in Berlin. Body and air temperature were measured simultaneously using a hand-held digital thermometer (Omega HH23 OmegaNewport, Deckenpfronn, Germany) equipped with two T-Type Teflon-insulated thermocouples (Omega 5 SC-TT-T-36). To measure theadult body temperatures of laboratory-reared C. lapponica, a thermocouple was inserted ≈3 mm into the abdomen of the live beetle. A sheet of paper shadowed the beetle during thefirst measurement while it was sitting on another sheet of paper. After reaching a stable temperature for 30 s, the beetlewas exposed to direct sunlight, and a second measurement of the body temperature was taken.

Walking Speed.

To assess differences in walking speed of adult leaf beetles, a Plexiglas rod (length 50 cm), with marks at intervals of 5cm, was mounted vertically in a frame. Experiments were conducted in a climate cabinet (Schütt, Göttingen, Germany)at 5.0°C, 10.0°C, 15.0°C, and 20.0°C, respectively. The cabinet was illuminated from above with eight fluorescenttubes (Lumilux daylight 18 W; Osram, Munich, Germany). Individuals of both populations of C. lapponica were allowed to acclimate for 2 h to each test temperature. Beetles were kept in plastic boxes (15 by 13 by 6 cm) and allowedto feed on their natural host plants. Any feeding behavior was noted. After acclimation, a single individual was placed withforceps at the base of the rod. As soon as the beetle had passed the first mark on the rod, the time needed for covering adistance of 20 cm was recorded. Beetles that stopped walking or returned were excluded. Each individual was used only forone experiment.

Calorimetry.

Heat production rates of overwintering adults of C. lapponica were determined at 5 and 10°C by means of an isoperibolic batch microcalorimeter (Biocalorimeter B.C.P.; ÉlectroniqueArion, Orsay, France) connected to a chart recorder (BD 111900750; Kipp and Zonen, Delft, The Netherlands). The calorimetercontained a measuring and a reference vessel with volumes of 15 ml each. The sensitivity of the calorimeter was 45.8 μV/mW.The calorimeter was set into a climate cabinet (Heraeus BK 500 E, Heraeus, Hanau, Germany) with a constant temperature of2.0°C. The temperature gradient between the inside of the calorimeter and the surrounding air was at least 3°C toimprove the sensitivity of the instrument. In addition, the calorimeter was placed into a styropore box, thus further stabilizingthe device against temperature fluctuations.

To prevent walking activity during the experiments, all beetles overwintered separately in small plastic vials (diameter,66 by 112 mm; Plano, Wetzlar, Germany) with five openings at each side for ventilation in a climate cabinet (Schütt)at 2°C and 80% RH. Groups of 10 plastic vials were set into petri dishes lined with moistened filter paper. The weightsof the vial and beetle were determined separately at the beginning and the end of the overwintering experiment using a precisionbalance (MC 210 S; Sartorius). In addition, all beetles were weighed in their vials before and after the calorimetric experiments.Body mass was calculated by subtracting the weight of the vials from the total weight.

To determine the heat production rates, beetles in their plastic vials were placed inside the calorimeter measuring vessel.In each experiment, four individuals were placed in one vessel, because the heat production rates of single individuals weretoo low for a reliable determination. Heat production was monitored for at least 40 min. Before and after each experiment,the baseline (i.e., the calorimeter signal without animals) was recorded. The mean heat production rates were determined byelectronic integration (Digikon, Kontron, Germany) of the P(t) curves. Tests were replicated between 7 and 10 times for eachpopulation and each temperature regimen (see Results). All calorimetric investigations were performed between 10 October 1998and 25 March 1999.

Recording Field Air Temperature Fluctuations During Overwintering.

To assess the temperature fluctuations that C. lapponica must cope with, small temperature data-loggers (-20.0 to +70.0°C, Hobo H01 and H8 Pro; Onset Computer, Bourne,MA, U.S.A.) were placed into the beetles' overwintering quarters in both habitats in central and northern Europe duringthe overwintering of the beetles. All beetles overwinter as adults in leaf litter layer. The data loggers recorded the temperatureevery 12 h. Measurements were conducted from October 1998 to May 1999. The collected temperature data were evaluated usingthe Boxcar Pro Program Version 3.51 (Onset Computer).

Statistical Analyses.

All data are presented as mean values ± SD. Data from the elytral coloration measurements, body temperature measurementsin the laboratory, and body mass were evaluated using the Mann-Whitney U test. The differences between the internal body temperature (Tb) and the air temperature (Ta) of the two morphs in the field experiment, which were measured in shadow or direct sunlight, were analyzed by two-way analysisof variance (ANOVA), followed by least significant difference (LSD) post hoc tests. Data from walking speed measurements andfrom the comparison between the specific heat production rates (calorimetry) were statistically evaluated by t-test for independent samples. Correlations between body weights and heat production rates were compared using the Spearmanrank correlation test. All statistical analyses were conducted using Statistica 5.5 software (StatSoft 1999).

Results

Coloring of the Elytra.

Elytra of the Finnish beetles were significantly darker than those of the beetles from the Czech population (Fig. 1A).While the proportion of the black color was 51 ± 7.3% in individuals from the Czech population, the black elytral proportionof the melanic Finnish morphs was 68 ± 9.2% (Mann-Whitney U test, P < 0.001, n = 40). The typical pattern of the elytra of the morphs from the Czech and the Finnish population are shown in Fig. 1B.

Fig 1.

Proportion of the black coloration of the elytra of randomized selected adult C. lapponica from two populations (Czech Republic, on Betula pubescens; Finland, on Salix borealis). (A) Mean values and SDs. (Mann-Whitney U test; P < 0.001; n = 40 of each population). (B) Elytral pattern of an adult from the Czech Republic (left) and Finland (right). White areas of the elytra in drawings are naturally red colored.

Body Temperature and Heating Time.

At the beginning of the laboratory experiments, each adult had a body temperature of 10°C. As soon as beetles were exposedto artificial light for a period of 5 min, they started heating up. Figure 2Ashows the body temperatures of adults from the Czech and Finnish population at 30-s intervals of the 5-min heating period.Adults of the melanic morphs from Finland reached a significantly higher internal body temperature than the light morphs fromthe Czech Republic (dark morphs, 21.0 ± 1.7°C; light morphs, 18.9 ± 1.1°C; Mann-Whitney U test, P < 0.05) in significantly shorter time (dark morphs, 2.9 ± 0.48 min; light morphs, 4.0 ± 0.50 min; Mann-WhitneyU test, P < 0.01; indicated by the arrows in Fig. 2A). The mean body weights of the tested beetles did not differ significantly between populations (dark morphs, 24.0 ±6.6 mg; light morphs, 24.3 ± 5.8 mg; Mann-Whitney U test, P > 0.05).

Fig 2.

(A) Body temperature of adult C. lapponica from two populations under laboratory conditions. White bars, light morphs from the Czech Republic; gray bars, dark morphs from Finland. Mean values and SDs of the internal body temperature increasing under irradiation for a period of 5 min are given at 30-s intervals. The initial body temperature at the beginning was 10°C. The time until the maximum temperature was reached is indicated by arrows. It is significantly shorter (P < 0.01), and the maximum body temperature is significantly higher (P < 0.05) in melanic beetles from Finland. Mann-Whitney U test (n = 9 for each population). (B) Differences in body temperature of adult C. lapponica fromtwo populations under field conditions. White bars, light morphs from Czech Republic; gray bars, dark morphs from Finland. Mean values and SDs of the differences between the internal body temperature (Tb) and the air temperature (Ta) of beetles measured in shadow or exposed to direct sunlight are given. Two-way ANOVA (n = 20). All effects (LE: F =191.9, P = 0.000; P: F =10.2, P = 0.002; LE ×P: F =9.3, P = 0.003). Different letters show significant differences (post hoc LSD, P < 0.001).

In the field experiment (Fig. 2B), the internal body temperature differed significantly between the two morphs when they were exposed to direct sunlight. Internalbody temperatures also differed when comparing adults exposed to shadow and sun temperatures (two-way ANOVA: light exposure[LE]: F = 191.9, df = 1, P < 0.001; population [P]: F = 10.2, df = 1, P < 0.01; LE × P: F = 9.3, df = 1, P < 0.01). No differences were found between the body temperatures of the two morphs sitting in the shadow. The differencesof air and body temperatures of beetles exposed to sunlight (dark morphs, 3.1 ± 0.62 K; light morphs, 2.1 ±0.96 K) were significantly higher in the dark morphs of the Finnish population (post hoc LSD tests, P < 0.001). The mean body weights between both morphs differed significantly (dark morphs, 21.3 ± 5.66 mg; light morphs,28.2 ± 5.43 mg; Mann-Whitney U test, P < 0.05), but no correlation between body temperature and body weight within each group was detectable (Spearman rank correlationtest, P > 0.05).

Walking Speed.

As shown in Fig. 3,the walking speed of C. lapponica in both populations depended on the temperature. Whereas at higher temperatures the walking speed of beetles of both populationswas faster than at lower temperatures, no significant differences between individuals of both populations were found at 20(t = 1.44, df = 43, P > 0.05), 15 (t = 1.19, df = 48, P > 0.05), and 10°C (t = 1.15, df = 74, P > 0.05). However, in the experiment conducted at 5°C, beetles from the Finnish population reached a significantly higherwalking speed than the Czech beetles (t = 3.14, df = 48, P < 0.01). Adults of the Finnish population showed feeding activity at all tested temperatures, whereas Czech adults did notfeed at an ambient temperature of 5°C.

Fig 3.

Walking speeds of adult C. lapponica from two populations at different temperatures. White bars, light morphs from Czech Republic; gray bars, dark morphs from Finland. 20°C: n = 20/25 (Czech Republic/Finland); 15°C: n = 24/25; 10°C: n = 26/50; 5°C: n = 25/25. t-test for independent samples, not significant. n.s.P >0.05; **P < 0.01.

Specific Heat Production Rate During Overwintering.

The specific heat production rates during overwintering did not differ at 10°C between adults of C. lapponica from the Czech and Finnish populations (Fig. 4;t-test: t = 1.20, df = 15, P > 0.05). At 5°C, the beetles from the Finnish population had a significantly higher heat production rate than the Czechbeetles (t-test: t = 2.49, df = 14, P < 0.01). Although the body weights of the adults differed significantly between both populations at 5°C (t-test: 10°C: t = 0.65, df = 15, P > 0.05; 5°C: t = 5.29, df = 14, P < 0.001), no correlation between body weight and specific heat production rate within each group was detectable (Spearmanrank correlation test, P > 0.05).

Fig 4.

Mean values and SDs of the specific heat production rates (µW/ mg) during overwintering of adults of C. lapponica from two populations at two different temperatures. White bars, light morphs from Czech Republic; gray bars, dark morphs from Finland. t-test for independent samples, not significant. nsP > 0.05; **P < 0.01. All comparisons conducted within each temperature group.

Overwintering Temperature in the Field.

The temperature in the overwintering quarters differed greatly between the two distribution areas of both populations: becauseof a permanent snow cover during winter, a nearly constant temperature of -0.5°C, varying merely between -1.0and -0.2°C, was found in the overwintering quarter in Finland. Temperature in the leaf litter layer was independentfrom air temperature. In the Czech Republic, the temperature in the leaf litter layer varied between -1.4 and 9.4°C.The mean temperature here was 2.1 ± 2.5°C (Fig. 5).

Fig 5.

Mean temperatures in leaf litter layer in Finland (FI) and the Czech Republic (CR) during overwintering of C. lapponica in the winter 1998/99. Temperatures were measured by data loggers.

Discussion

Even though there is some variation in elytral coloring within the populations of C. lapponica (Gross 1997, Milyashevich 2000, Zvereva et al. 2003), the elytra of the Finnish adults were significantly darker than the elytra of Czech beetles. Whereas in some other insectgroups variations of specific developmental parameters can cause different phenotypic morphs (Watt 1969), the relationship between black and red color in C. lapponica is heritable, which was shown by crossing experiments between different populations from Finland (Zvereva et al. 2003) and between beetles from Finland and the Czech Republic (J.G., unpublished data). Heritable color polymorphism is also knownin ladybird species (Muggleton 1979, Osawa and Nishida 1992).

The melanic morphs of the Finnish population of C. lapponica attained a higher body temperature in shorter time than the nonmelanic Czech morphs when irradiated with artificial lightin laboratory experiments or when exposed to direct sunlight in field. These results indicate that, under standardized conditions,the elytral coloring affects the internal body temperature of C. lapponica, as has been shown in a former study for melanic and nonmelanic morphs of the ladybird beetle Adalia bipunctata (De Jong et al. 1996).

While the ecological significance of the melanic and nonmelanic morphs of A. bipunctata remained the subject of some controversy (Muggleton 1978, 1979, Müller 1985, Brakefield and Willmer 1985, De Jong et al. 1996, De Jong and Brakefield 1998), the thermoregulative abilities of the two differently colored C. lapponica morphs may be well interpreted as adaptations toward the climatic conditions of either morph.

When reared under standardized temperature conditions, larvae of the Finnish population develop significantly faster thanlarvae of the Czech beetles (Gross et al. 2003). After emerging from the pupa, an adult beetle has to feed until the end ofthe vegetation period when the leaf quality decreases to maintain or increase its energy reserves. Thus, a higher internalbody temperature will increase the feeding activity of the adults and benefit the melanic beetles of C. lapponica living in a colder environment, as it is the case in melanic morphs of Adalia bipunctata, where activity was significantly higher during illumination compared with nonmelanic morphs (De Jong et al. 1996).

Feeding and walking activities of C. lapponica might not only be affected by color adaptations toward environmental temperature, but also by other physiologically adaptiveprocesses. At low temperatures, the Czech beetles did not feed and walked only slowly. In contrast, the Finnish beetles showedhigher activities, which might be considered an adaptation to start the season early in spring, even when temperatures arestill low in the Finnish habitat. Differences in enzymatic processes at low temperatures might have caused the higher activitiesof the cold-living Finnish beetles. Temperature may act as a selective factor, favoring different allozymes in different environments(Dahlhoff and Somero 1993, Dahlhoff and Rank 2000).

During overwintering, the physiology of an insect changes considerably, e.g., when it experiences sub-zero temperatures (Bale et al. 2001, Sinclair et al. 2003). Its metabolism is greatly reduced to decrease energy requirements (Keister and Buck 1974, Schmolz and Lamprecht 2000). Adults of the Czech population of C. lapponica spent ≈230 d in their overwintering quarters, whereas overwintering in Finland lasted an extra 45 d (J.G., unpublisheddata; N. E. Fatouros, personal communication). For overwintering, adults of both populations walk ≈5 cm deep into the leaflitter layer close to the host plant. The results of the calorimetric investigations show that the specific heat productionrates of the Finnish beetles, which have to stay in diapause for a longer time, were higher than the specific heat productionrates of the Czech beetles. The lower heat production rates of the Czech beetles might be advantageous to cope with the fluctuatingtemperatures during overwintering. Our results show that overwintering beetles are threatened less by very low temperaturesthan by higher and fluctuating ones, which are common in central Europe during wintertime (Fig. 5), because of the lack of a permanently insulating snow cover (Sinclair et al. 2003). The higher temperatures that the Czech beetles encounter during overwintering could be detrimental, because an increaseof the environmental temperature means an increase of body temperature, which could lead to an increase of metabolic costsfor the beetle. The low specific metabolic rate found in the Czech population of C. lapponica might be an adaptation to the climatic conditions in their habitat to reduce costs at raising overwintering temperatures.

In insects, as in all animals, the specific heat production rate is generally negatively correlated with body mass (Kittel 1941, for beetles, Coelho and Moore 1989, for cockroaches; Lehmann et al. 2000, for Drosophila). Even though we could not detect any correlation between body mass and specific heat production rate in C. lapponica, the larger body mass of the Czech beetles may lead to a lower specific heat production rate and thus a further reductionof their energy expenditure. This might be advantageous at the Czech overwintering quarters, where the exposure to fluctuatingand mild temperatures might require more energy that needs to be spent economically.

Interestingly, the differences in body size between both populations do not yield differences in reproductive success andfitness. The number of offspring is not different between Czech and Finnish beetles (Gross et al. 2003). Because the heatproduction rates within the experimental groups did not change significantly with body mass but only between the groups, wepresume that the different heat production rates are adaptive for both populations, with a specific level of metabolism forboth of them. The higher specific heat production rate of Finnish beetles can be explained with their smaller body size/bodymass.

The results of this study show that the melanic morphs living in colder environments of Finland are well adapted toward lowtemperatures. Our study suggests that the morphological and physiological differences between the willow- and birch-feedingpopulations of C. lapponica studied here provide a good adaptation to the respective climate conditions in their natural habitats.

Acknowledgments

We thank S. Neuvonen (Kevo Subarctic Research Institute, Finland) for the temperature measurements in the overwintering quartersof C.lapponica; N. E. Fatouros, M. Homann, and K. Täger (Freie Universität Berlin) for help in the laboratory; and I. Lamprecht(Freie Universität Berlin) for assistance in the calorimetric experiments. We also thank the Ministry of Ecology in theCzech Republic for permission to collect leaf beetles in protected areas. This study was supported by the Deutsche Forschungsgemeinschaft(Hi 416/9-1,2) and Berlin for a NaFöG fellowship.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial reuse, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions{at}oup.com

References Cited

View Abstract