• In a microgravity experiment onboard the International Space Station, circumnutations of Arabidopsis thaliana were studied. Plants were cultivated on rotors under a light:dark (LD) cycle of 16 : 8 h, and it was possible to apply controlled centrifugation pulses. Time-lapse images of inflorescence stems (primary, primary axillary and lateral inflorescences) documented the effect of microgravity on the circumnutations.
• Self-sustained circumnutations of side stems were present in microgravity but amplitudes were mostly very small. In darkness, centrifugation at 0.8 g increased the amplitude by a factor of five to ten. The period at 0.8 g was c. 85 min, in microgravity roughly of the same magnitude. In white light the period decreased to c. 60 min at 0.8 g (microgravity value not measurable). Three-dimensional data showed that under 0.8 g side stems rotated in both clockwise and counter-clockwise directions.
• Circumnutation data for the main stem in light showed a doubling of the amplitude and a longer period at 0.8 g than in microgravity (c. 80 vs 60 min).
• For the first time, the importance of gravity in amplifying minute oscillatory movements in microgravity into high-amplitude circumnutations was unequivocally demonstrated. The importance of these findings for the modelling of gravity effects on self-sustained oscillatory movements is discussed.
Self-sustained oscillations in biological systems attract wide interest both from theoretical and from experimental points of view (Turing, 1952; Winfree, 1980). In some cases the oscillations at the cellular level are transformed into variables that are easily observed and measured. Helical, rotational growth movements are abundant among plant species and constitute a striking example of growth oscillations. Charles Darwin (Darwin, 1880) was instrumental in visual studies of such plant movements and denoted them ‘circumnutations’. Aspects of growth processes in plants are of fundamental research interest and the arousal of rhythmic growth movements make them fascinating but difficult to explain. Darwin emphasized the endogenous origin of the circumnutations and suggested that environmental factors – such as gravity – could have a modifying effect. However, ever since Darwin's publication more than a century ago, it has been debated whether environmental signals, such as the Earth's gravity, could generate some types of circumnutation. We report unequivocal results from a study of such movements in Arabidopsis thaliana on the International Space Station (ISS).
Circumnutation movements, exemplified in Fig. 1(a) for A. thaliana ecotype Columbia wild type, cover a wide frequency range in different plant species (a rotational cycle of minutes to some hours; Brown, 1991; Johnsson, 1997; Schuster & Engelmann, 1997; Mugnai et al., 2007). The considerable deviations from the plumb line shown in Fig. 1(a) would lead to gravitropic reactions in a tropistically sensitive plant. Of course, reactions to the mechanical bending and compression of the tissue also occur. Thus, gravitropic stimulation or induced reactions due to bending (e.g. Brown, 1991; Baskin, 2007) can be implied and might ‘interfere’ with rhythmic growth movements or even generate the rhythms of plant organs. Overshooting in feedback (or compensation) systems comprising gravitropic or bending reactions could, in fact, be the origin of the circumnutations. It has indeed been demonstrated that gravity influences large-amplitude circumnutations in several plants (e.g. Helianthus and Pharbitis). Evidence comes, for example, from centrifuge experiments (Zachariassen et al., 1987), clinostat experiments (Johnsson, 1974) and experiments using mutants (Hatakeda et al., 2003; Kitazawa et al., 2005; Tanimoto et al., 2008). The SCARECROW (SCR) and SGR5 genes are reported to regulate gravitropism as well as circumnutations in A. thaliana (Kitazawa et al., 2005; Tanimoto et al., 2008), but some A. thaliana mutants show circumnutations but no gravitropic reactions (B. G. B. Solheim, pers. comm.). Experimental results over the last few decades have shown that different types of plant cells and tissues often exhibit endogenous rhythmic ion and water transport (Shabala, 2003; Manusco & Shabala, 2007). Such endogenous oscillations in tissue could lead to circumnutations (Shabala & Newman, 1997a).
In the early literature it was suggested (Johnsson & Heathcote, 1973) that detailed studies of circumnutations in weightlessness would help to elucidate the role of gravity in the generation of these movements. In the first Spacelab flight, Brown & Chapman (1984) recorded circumnutations of Helianthus annuus hypocotyls in an experiment of c. 1-wk duration (Brown & Chapman, 1984; Brown et al., 1990). The movements did not appear as frequently in weightlessness (40% of the time as compared with 1 g controls), the amplitude of the movements was reduced (by a factor of two) and the period of the movements was shorter (87.6 min instead of 105 min at 1 g). Darwin's concept of an endogenous origin of the circumnutations was thus supported in this experiment on H. annuus hypocotyls (Brown & Chapman, 1984, 1990; Kiss, 2006). However, detailed studies of the connection between gravitropic reactions and rhythmic, cellular processes were still to be carried out both in hypocotyls and in other parts of the plant.
The present A. thaliana experiment was performed on the ISS. Our aim was to obtain direct confirmation of the importance of gravity for the generation of the circumnutations. The use of rotors in the experiment allowed the application of hypogravitational g-‘pulses’ to A. thaliana plants growing (otherwise) in weightlessness, and thus each plant provided its own control. CCD cameras were used to follow plant movements for c. 70 d. The experiment reveals new facets of circumnutations and their acceleration dependence.
The main focus of this paper will be the movements of the primary axillary and lateral inflorescence stems (terminology: Jouve et al., 1998), but the main stem (primary inflorescence) movements will also be discussed. For the sake of brevity, the primary axillary and lateral inflorescences stems will together collectively be referred to as ‘side stems’.
Materials and Methods
The experiment was carried out in the European Modular Cultivation System (EMCS), located in the US Destiny module of the ISS. The EMCS hardware (Brinckmann, 2005) provided the facilities for a long-term experiment (the circumnutation experiment was part of the seed-to-seed experiment Multiple Generations 1 (MULTIGEN-1)).
Plant material and growth conditions
Plants were grown in experiment containers (ECs) on two separate rotors (allowing g values of between 0 and 2 g). A water reservoir was located on each rotor. The EMCS is an incubator where watering, automatic control of air exchange, medium and air humidity, as well as illumination from the light array could be achieved in the ECs by interfacing with the EMCS. Details of the EMCS design, the rotors and the ECs have been published previously (Brinckmann, 2005; Fossum et al., 2005; Solheim et al., 2006), but an overview of the rotor configuration is given for the convenience of the reader in Fig. 2(a). Four ECs were mounted on each rotor. Images of plants in the ECs were acquired from two turnable cameras per rotor and one mirror per EC. Further details are given in the caption of Fig. 2(a) and in the subsection ‘Rotors’ below. Figure 1(b) shows one plant developing in weightlessness on the ISS and a video image acquired 2 d after visual emergence of the main inflorescence stalk. A light array is situated above the plant, and when the rotor is started later in the experiment, the centrifugal force is directed ‘downwards’, as indicated by the arrow (cf. Fig. 1d). Positions in the video image are specified by x- and y-coordinates, as indicated by the axes. The ECs, which had dimensions of 60 × 60 × 160 mm and were fitted with plant cultivation chambers (PCCs), allowed A. thaliana to germinate and grow and finally to flower.
Arabidopsis thaliana L. ecotype Columbia wild-type seeds (purchased from Lehle Seeds, Round Rock, TX, USA) were cultivated under a light:dark (LD) cycle of 16 : 8 h. The growth medium was based on zeolite (ZeoPro, ZeoponiX; Digitech AB, Älandsbro, Sweden) in combination with nutrient enrichment Murashige and Skoog (MS) (M-5519; Sigma-Aldrich Norway AS, Oslo, Norway). Seeds germinated in 26 out of the 32 holes provided in the experiment. Depending on the configuration, there was either one or three seeds in each hole. In the 12 holes that only contained one seed per hole, seven seeds, or c. 60%, germinated (germination was difficult to determine for the three seeds per hole configuration, but one or more seeds germinated in 85% of these holes). The ground experiments had c. 95% germination.
Light Light-emitting diodes (LEDs) in the light array produced 75 W m−2 photosynthetically active radiation (PAR) in the centre of the EC. LEDs used were white (Nichia NSPW 500 BSbS, Nichia, Tokushima, Japan), Red (Rohm SLA-570-JT-3F-XP/XQ, Rohm co., Kyoto, Japan) and infrared (IR) (for dark imaging; Honeywell SE5455, Honeywell International Inc., Morristown, NJ, USA), providing a continuous spectrum from approx. 400 to 730 nm with a peak at 465 nm and a lower peak at approx. 660 nm. The IR LEDs had zero intensity below 790 nm.
Temperature The temperature was nominally kept at 23°C.
Rotors Rotors were either kept at rest to provide micro-g conditions or pulsed to provide a 1-g environment at the level of the seed position. The distance to the rotor axis diminished for plant parts growing ‘upwards’ and the g-level accordingly diminished to c. 0.8 g in the middle of the EC and 0.5 g at the very top of the EC.
Overall experimental set-up
Each PCC was kept in dry mode until the experiment was actuated and initialized on 24 August 2007, after transport of the plant material to the ISS on 8 August. The number of holes in the PCCs, three or five depending on the configuration, was carefully considered with respect to the numbers of seeds in every hole and their germination percentage, in order to achieve optimal experimental conditions for a complete MULTIGEN-1 experiment, but also to provide suitable conditions for studying the leaf and stalk movements of individual plants. A reference background and a reference pin were installed in the ECs to allow exact determination of the positions of plant parts from the camera pictures (seen in Fig. 1b–d).
The MULTIGEN-1 experiment was stopped on 5 November 2007 and plants were dried and conserved in an incubator (at a temperature of 23°C) for later analysis.
The growth of the plants and the image retrieval were interrupted by unforeseen problems, reducing the number of image sequences suitable for analyses. A short timeline of the relevant parts of the experiment is given in Fig. 2(b). Power outages occurred and interfered sometimes with planned activities (cf. Fig. 2b) or with the LD schedule. Predicted difficulties were related to power outages caused by space crafts visiting the ISS. The most serious problem was that the air flow was too high in the ECs early in the experiment, causing drying out of several plants after 7 d; the remaining plants recovered but were delayed in development. Out of the seeds activated by watering, c. four plants allowed determination of rosette leaf movements (to be reported) and only one plant achieved the flowering stage. However, the plant that reached the flowering stage was alone in the EC and therefore provided with optimal conditions for circumnutation determinations. During the emergence and growth of the inflorescence stem and side stems, the growth rate recovered and the development of the plant progressed at a normal rate.
These details are presented in order to convey some of the challenges that might be met by users of the equipment. Our experiment was, however, successful and provides unique photographic sequences of circumnutations in microgravity and under centrifugation.
Four video cameras (Sony FCB-IX470) were mounted on the two rotors (cf. Fig. 2a), allowing recording of plants in the ECs both in visible light and in IR light (i.e. during the whole LD cycle). Images were sampled at preset intervals, in most recordings every 5 min. By adjusting and rotating the mirrors used for the imaging (see Fig. 2a), pictures could be taken from slightly different angles of the same plant in an EC. This allowed 3D pictures to be constructed in addition to the 2D recordings (B. G. B. Solheim, manuscript in preparation). Sampled images were stored temporarily on a mass memory unit and were down-linked. Some of the files containing the large amounts of data produced allowed automatic scanning for more rapid, convenient determination of positional changes; others required cumbersome, detailed individual calculations.
All plant movements are seen as movements in the x–y plane (the plane defined in Fig. 1) because of the 2D limitation of the camera; thus only x- and y-coordinates are readily available. If a movement in the z-direction occurred (i.e. the plant moved towards the camera or away from it; cf. Fig. 1c and d) a camera in a fixed position could not record such a movement. The technique of recording two images with a position change of the mirror between exposures could then be used to determine the z-position in addition to the x- and y-positions, thus giving 3D coordinates. Playing sequences of movements also helped to estimate directions of circumnutation, rotation, etc.
The technique used to construct 3D pictures instead of simple 2D recordings required additional control of the mirror motor and the frequency of imaging. The 3D technique was developed to extend the capabilities of the EMCS imaging system and is only briefly mentioned in the present report.
Every image retrieved was interpreted and the positions of parts of shoots, leaf tips etc. were electronically specified in a 2D grid, representing the x- and y-coordinates as a function of time. Experimental time was documented on every image. Movements of the plant organs in the x- and y-directions (horizontally and vertically in images, respectively) were determined also when different levels of camera zoom were used during plant growth. The background pattern in the ECs was a necessary aid in this procedure.
Subpixel accuracy was difficult to obtain, in part because of compression artefacts, and resulted in a resolution in 2D positions of, in principle, 1 pixel, corresponding to c. 0.25 mm. Repeated measurements on the same sequences (also by different persons) confirmed that the resolution was of this order.
The EMCS hardware provided a facility that allowed the A. thaliana plants to germinate and to mature on the ISS. The camera system and downloading routines functioned most of the time, but power-down periods as well as loss of images caused disruptions in some time sequences. Planned experiments could not, therefore, always be carried out.
Germination after imbibition of the seeds was at least c. 60%. Growth was, however, initially delayed as a result of drying out of the medium (caused by excessive air flow which was not adjusted until the wilting process had started in most plants). At the onset of the experiment, the rosette leaves demonstrated clear short-period movements (with periods of c. 45 and c. 80 min) in microgravity and will be the topic of a future paper (B. G. B. Solheim et al., unpublished).
The same plant as in Fig. 1(b) is also shown in Fig. 1(c), still in microgravity but 3 d older than in Fig. 1(b). Rotor acceleration was applied on the same day (cf. Fig. 2b) and the plant reactions can be seen in Fig. 1(d). The plant was the only one in this EC and it could grow without touching other plants (and was not restricted by the walls of the container during the period studied). The inflorescence stem as well as the primary axillary and lateral inflorescences stems were studied only when they could move freely.
Circumnutational movements of side stems
Figure 3 illustrates a unique sequence of movements of the five different side stems of the plant in Fig. 1(c) and (d). The application of g-force is indicated at the top of the figure and LD sequences are indicated by the white/shaded areas. The x-positions of the tips of the shoots were determined under the following conditions (as seen when moving stepwise from left to right in the figure): (a) microgravity in darkness; (b) c. 0.8 g on the centrifuge in darkness (average g value in relevant growth volume); (c) 0.8 g on the centrifuge in the light; (d) 0.8 g in darkness; (e) 0.8 g in the light; (f) microgravity (rotor stopped), with the plant first in the light and then in darkness. In one stem (the fourth curve from the top) the x-coordinate did not change much when the rotor was started because the shoot moved in the camera direction. The movement in the y-coordinate is, therefore, documented in an inset. The sequence illustrates the following features: minute movements in microgravity and darkness, pronounced gravitropic reactions and circumnutations at 0.8 g in darkness or light, and finally almost complete absence of circumnutations when the centrifuge was stopped again. A final transition to darkness caused a slight lowering of the plagiotropic pointing direction for some shoots.
Circumnutations in microgravity
Movements of side stems were usually of a random nature, but a few rhythmic movements were present also in microgravity. Examples can be discerned before the rotor started; see curves 3 and 4 from the top of Fig. 3. The peak to peak values were then of the order of 2 mm and the cycle period was difficult to determine, but was estimated at c. 90 min (varying from 115 to 60 min, as shown in Fig. 3).
Circumnutations during application of g-force
The period of the circumnutation cycles under the initial 0.8-g treatment varied with the light conditions and decreased from 83 min (± 2 min, standard deviation of the mean; n = 25 measured cycles) in darkness to 62 min (± 1.5 min; n = 24)) in the light (simple averages calculated over cycles present even if trends could be discerned). The light effect is in accordance with results in the literature on A.thaliana circumnutations of inflorescence stems at 1 g (Someya et al., 2006).
Amplitudes at 0.8 g were smaller in darkness than in the light (estimated from the recordings to be c. 6.6 and 8.8 mm, respectively; i.e. a ratio of c. 1.3). The circumnutations were damped (possibly as a result of circadian effects; cf. Buda et al. 2003; Niinuma et al. 2005) in the second light period, resulting in a decrease in amplitude (see Fig. 3), a behaviour known from both our earlier laboratory experiments on Earth and the ground-based MULTIGEN-1 control experiment running in the Experiment Reference Model (ERM).
The oscillations did not continue in microgravity after the termination of the rotor stimulation (Fig. 3), although the control plants were capable of oscillation (Fig. 5). Therefore, the effects of the accelerations (and gravity) do not seem to require the long-term build-up of conditions favourable for after-effects promoting circumnutations.
The gravitropic reactions of the side stems at the start of an acceleration force were clearly visible (cf. Fig. 3). The shoots evidently moved from their plagiotropic set point direction in weightlessness, as demonstrated for the first time here, and were able to react to the applied hypogravity acceleration with a gravitropic reaction time of c. 30 min. They then moved, reversibly, back towards the original direction after the end of the acceleration pulse (cf. Fig. 3). The plagiotropic set point direction seems to be only slightly affected by the light in the experiment (the tip positions of the side stems slightly lowered when the light was switched off in microgravity).
Direction of circumnutation rotations
3D analysis of the movements in 0.8 g (B. G. B. Solheim, unpublished) demonstrated that the tips showed rotational movements in both clockwise and counter-clockwise directions. Two results from the 3D analysis of the movements of side stem tips are shown in Fig. 4(a). As seen from ‘above’ (the direction of light and centrifugal acceleration), the tips of the side stems moved in both counter-clockwise and clockwise directions, as shown by the projections onto the horizontal plane.
The structure of the movements can be made clearer. First, the scatter in the data can be decreased by filtering the data (a running average filter, weighing three subsequent data in the times series with weights ¼, ½, and ¼, was used). This has been done with the data in Fig. 4(b). Secondly, because the side stems are often obliquely oriented with respect to the main stem, at least in microgravity, a horizontal projection is not always easily interpreted. The movements were therefore projected in such a way as to be seen from ‘above’, but along the pointing direction of the side stem. Figure 4(b) demonstrates such a projection (in 0.8 g). The more circular nature of the circumnutations when seen from this angle is useful and facilitates precise studies of, for example, amplitudes. Finally, a demonstration of the movement of the same tip in the x-, y- and z-coordinate system is given in Fig. 4(c).
Circumnutation movements of the main stem
Only very slight tendencies towards periodic movements of the main stem could be ascertained before an acceleration pulse (0.8 g) was given in darkness (the length of cycles was estimated at c. 110 min). After a gravitropic reaction time of c. 25 min, oscillations built up in 0.8 g, but unforeseen power outages coupled with a transient temperature increase and loss of data during parts of that period meant that only a few observations were possible. The amplitude acquired tended to be much larger than 5 mm and the period was c. 90 min.
During a later light sequence (30–31 October; see Fig. 2b), the main stem showed clear circumnutations with small amplitudes in microgravity but these were amplified and regular when acceleration was applied. Although few complete cycles were available, the ratio between the amplitudes with and without acceleration was estimated at c. 2.0. The period length of the cycles studied (under light conditions) was estimated to be c. 60 min in microgravity and 78 min under 0.8 g, but few data were available.
Circumnutational movements of ground controls
Figure 5 is a record from a control in the ERM at 1 g on Earth. It illustrates the typical oscillatory movements of a lateral inflorescence throughout 2 d of recording (the same stage of development as for shoots in Fig. 3) under LD 16 : 8 h. The growing side stem is continuously oscillating both in darkness and in the light, and the amplitude of its movements has the same magnitude as the amplitudes of movements shown by the stems in Fig. 3 (at 0.8 g). It also exemplifies the complicated amplitude effects under LD conditions (e.g. as seen in Someya et al., 2006). In the present case, a characteristic decrease in amplitude after the transition to light, and a decrease in amplitude just before the onset of darkness were observed. Characteristically, the plant was clearly circumnutating in 1 g both in darkness and in the light (and with a shorter period in the light than in darkness, as also shown by the ISS stems in Fig. 3).
The ISS allows studies of biological objects for long periods in weightlessness (perturbations and interruptions must, however, be carefully monitored to avoid erroneous interpretations of the data). The ISS in general, and the EMCS in particular, will facilitate studies and interpretations of many tropistic reactions in plants as well as of growth processes in general. It can be foreseen that, to understand large parts of plant sensory physiology, experiments in weightlessness and under hypogravity will have to be performed.
The recordings of circumnutation reported here in microgravity and under acceleration pulses are unique, although earlier studies of seed-to-seed growth in A.thaliana have been performed in space (Link et al., 2003). The present results extend the observations by Brown et al. (1990) on H. annuus hypocotyl circumnutations and confirm the existence of nutational movements in microgravity, although (as was the case for H. annuus) nutational cycles are much fewer in microgravity and the amplitudes smaller. The present recordings demonstrate not only that endogenous nutations, as Darwin predicted, can be present in the side stems, but also that there is a generating and amplifying effect of acceleration forces, and thus of gravity, on this type of movement (Fig. 3). The side stems of a single plant that moved freely in a container showed this in a striking way: minute circumnutations in weightlessness increased in amplitude by a factor of c. five to ten at 0.8 g (cf. Fig. 3). The stems also showed gravitropic reactions (although cultivated in microgravity) with a gravitropic response time of c. 30 min.
The amplitudes of circumnutations also vary under conditions found on Earth – high-amplitude nutations are often correlated with longer periods of, typically, 90 min length, and low-amplitude circumnutations (often denoted micronutations; Heathcote, 1966; Barlow et al., 1994) with shorter periods. Small-amplitude movements as seen under microgravity might be difficult to separate from random movements (cf. the fractal analysis by Mandelbrot, as cited by Barlow et al. 1994). Reported nutational movements are sometimes as small as parts of an angular degree and are indeed difficult to separate from random movements. For a stem/shoot of 1 mm diameter, a length increase of c. 20 µm on one side with respect to the other one would result in a bending of about 1 angular degree. We estimate the resolution in our experiment to be of this order (1 pixel corresponding to c. 0.25 degrees when the camera zoomed to cover the entire growth area in the EC).
The gravitropic participation in the circumnutation movements leads to the problem of how the acceleration signals or the gravity field of the Earth affects internal oscillators that are capable of generating cycling movements. A correlation has been shown to exist between endogenous, rhythmic ion pumps (and/or ion fluxes) and circumnutations in roots and twining shoots (Millet et al., 1988; Shabala & Newman, 1997a). Osmotically driven water fluxes must ultimately be responsible for the bending and the rotational movements. Neither these osmotic water fluxes nor circumnutation movements must necessarily be coupled to growth (Berg & Peacock, 1992; Baskin, 2007), although they occurred during growth in the present experiment.
Focusing on endogenous oscillations, the coupling between individual cells around the stem periphery necessary to cause the circumnutation ‘waving’ might consist of diffusion coupling, electrical coupling, etc. (Turing, 1952; Brown, 1991; Johnsson, 1997). It may also involve coupling between endogenous rhythms in the individual cells. The calcium oscillators documented in many plant cell systems (e.g. guard cells; Blatt et al., 2007) might be part of such rhythms and have a relevant period length. However, if this is the case, a direct correlation between calcium oscillations and circumnutations must be demonstrated experimentally. Further features that must be explained are simultaneous appearances of two different nutation frequencies, as found, for example, in A.thaliana hypocotyls (Schuster & Engelmann, 1997), and two oscillator approaches have been used to explain the resulting movements (Shabala & Newman, 1997b; Johnsson et al., 1999).
The gravitropic reaction chain involves hormone translocations, ion transport and osmotic gradients which can provide links between the gravitropic system and the endogenous circumnutations in shoots discussed above. Oscillatory tendencies of the tissue can be triggered by stimulations from gravitropic reactions (cf. gravitopic stimulations of the plant in Fig. 1) or from compression/extension reactions in the tissues and result in large-amplitude circumnutations. Such stimulations might be looked upon as ‘resonance pulses’, exciting the system.
Deviations from a set point direction determined tropistically, or from tissue morphology, can be involved in a feedback system able to perform oscillations. This process might be identical to (or difficult to separate from) the process denoted ‘autotropism’ in the literature (‘straightening’ of a curvature). The origin of the role of gravity in circumnutations must be found in effects other than those in a gravitropic ‘overshoot’ model (Israelsson & Johnsson, 1967) but must be compatible with results from experiments that show the amplifying role of gravity in circumnuations of at least two species tested in weightlessness, namely, the present studies of the stems of A.thaliana and the hypocotyl recordings of H. annuus (Brown et al., 1990). Modelling of self-sustained oscillations combined with the necessary gravity effects is still in its infancy (Antonsen, 1998).
The circumnutation results obtained in the present microgravity study unequivocally reveal several new features of movements in freely moving orthotropic and plagiotropic stems of A.thaliana, imbibed and cultivated under microgravity. In the experiments the plant tissue was exposed to acceleration pulses, and the plant was able to serve as its own control when the effects of pulses were recorded. The results of this experiment should allow research in this field to take one step further, allowing future experiments to focus on how gravity amplifies circumnutations.
We gratefully acknowledge support from the technical staff at our departments, economic support from the Norwegian Space Centre and ESA, and technical collaboration with Prototech AS (Bergen, Norway) and Astrium (Friedrichshafen, Germany). We are also grateful for support from the Norwegian User Support and Operation Centre, Trondheim (N-USOC) before and during the operations.