Glowing in the dark: discriminating patterns of bioluminescence from different taxa during the Arctic polar night
Research since 2009 has shown that despite almost total darkness during the Arctic polar night, there is much more biological activity than previously assumed, both at the sea surface, water column and sea floor. Here, we describe in situ monitoring of the bioluminescent fraction of the zooplankton community (dinoflagellates, copepods, krill and ctenophores) as a function of time and space. In order to examine the relative contribution of each selected taxon and any diurnal patterns in the relative signals, a time series platform capable of detecting in situ bioluminescent flashes was established in Kongsfjord, Svalbard, during the polar night in January 2013. Combined with laboratory-controlled measurements of animals collected next to the time series platform, we present both taxon-specific and community characteristics of the bioluminescence signal from a location at 79°N and from the middle of the polar night. Based on this 51-h time series, we conclude that the bioluminescent fraction of the zooplankton does not maintain a diurnal signal. Rather, the frequency of bioluminescence flashes from the entire bioluminescent community remained steady throughout the sampling period. Furthermore, we conclude that bioluminescence flash kinetic characteristics have a strong potential for in situ taxa recognition of zooplankton.
KeywordsArctic Polar night Marine zooplankton Bioluminescence Copepods Krill Ctenophores
During the last few decades, there has been a consistent and strong reduction in the sea ice extent in the Arctic Ocean (Sakshaug et al. 2009; Polyak et al. 2010). For its unique ecosystem, the sea ice is an important habitat for more than one thousand species of algae and invertebrates that have been recorded living in or on the underside of the ice, as well as the marine mammals and seabirds that find their main food there (Bluhm et al. 2011). In addition, the sea ice acts as a key regulator for important processes such as light penetration, water mass formation and temperature regulation. Hence, the Arctic sea ice extent and thickness is a “canary in the coal mine” with respect to global climate change in the high north (Polyak et al. 2010). Today we see an expanding activity and international interest in the marine environment in the Arctic with respect to natural resources such as oil, gas, minerals and fisheries. New activities and sea-transport routes require new knowledge about ecosystem structure and dynamics for sound management and decision making (Sakshaug et al. 2009; Wang and Overland 2009; Hirche and Kosobokova 2011).
There has been a common assumption from the governmental authorities, nature management and the scientific community that during the polar night, there is low biological activity and biomass at all trophic levels throughout the water column and on the sea floor. Lately, however, the use of new underwater technology such as autonomous underwater vehicles (AUVs) and optical sensors have proven to be important for elucidating life and environmental conditions during the polar night (Berge et al. 2009, 2012; Båtnes et al. 2013). Also, we now know that several species of seabirds (Brünnichs guillemot, black guillemot, kittiwake, fulmar, little auk and glaucous gull) are overwintering in Arctic fjords (Weslawski et al. 1991; Berge et al. 2012, own observations) and actively feeding in the upper water column. Their prey comprises of different species of zooplankton (copepods, amphipods, ctenophores, krill) and fish (e.g., polar cod) indicating key trophic linkages persist at these locations.
BL kinetic parameters used to discriminate between taxa
Maximum intensity of BL
×109 photons s−1
Average BL intensity
×109 photons s−1
Cumulative sum of BL0 % until BLmax
×109 photons s−1
Time to reach maximum BL
High intensity duration, BL20 % rise and 20 % decay
Time from BL20 % decay to BL0 %
Materials and methods
The bathyphotometer (UBAT, Wet Labs, USA) consist of a water inlet and outlet using an impeller to pump the water into a detection chamber equipped with a photon multiplier tube (PMT) counting bioluminescence as photons s−1 after calibration according to Herren et al. (2005). The UBAT flow rate was 0.373 L s−1. The impeller in the detection chamber of the bathyphotometer was used to induce bioluminescence. The controller software (BLINC, Wet Labs) was used as graphic user interface and for control of instrument (inserting calibration coefficient, pump RPM, flow RPM, PMT input/output voltage and raw counts).
The software was used to schedule BL recording times, i.e., schedule for repeating samples and sample length adjusting for battery power needed for 51-h TS at in situ temperatures.
The bathyphotometer provides the measurements as raw 1–60 A/D counts per second that were converted to bioluminescence values (BL, photons s−1) after correction of instrument-specific photomultiplier tube (PMT) calibration coefficient provided by Wet Labs. The BP utilizes an impeller to continuously draw a measured volume of water into a chamber where bioluminescence was measured by the PMT at 60 Hz (see Herren et al. 2005; Berge et al. 2012; Moline et al. 2013 for details). The PMT calibration coefficient was different depending on gain setting.
Taxa-specific bioluminescence: laboratory experiments
The UBAT was used for detecting and characterizing the flash kinetics on living specimens in a laboratory aquarium on board the R/V Helmer Hanssen (University of Tromsø) January 9–21, 2013. The UBAT was connected to the DH4 data handler, logger and power unit (both Wet Labs) with baud rate PC-DH4 set to 115 kHz using the DH4 software. The DH4 was powered with 12 V (input power 12–15 V) and a power cord connection with 12 V was connected from DH4 to UBAT.
In situ collection of zooplankton for onboard laboratory BL work was performed using WP2 zooplankton net (180 μm, 0.25 m2 opening) and Mik-net (1,000 μm, 3 m2 opening) close to the UBAT time series station (in situ temperature ranging from 0.54 to 0.89 °C and salinity 34.62–34.64 (recorded at 1 Hz).
Taxa (still in suspension in net flask) were immediately separated into polyethylene trays and stored in pre-filtered saltwater in a dark cooling room at 2–4 °C before transferred to 50 L aquarium filled with pre-filtered (180 μm) seawater (25 L of pre-washed polyethylene tanks) at in situ temperature. Prior to each specimen examined for BL, new and fresh, pre-filtered seawater at in situ temperature was used. A pre-washed bathyphotometer was submerged into the aquarium and examined species for species.
In situ measurements of bioluminescence
Bioluminescence kinetics of the four taxa producing BL in the laboratory using BL kinetic characteristic
BL groups (× 109)
×109 photons s−1
×109 photons s−1
×109 photons s−1
Taxa-specific BL kinetic signatures
In particular four parameters; maximum bioluminescence (BLmax), time to maximum bioluminescence (Tmax), BL cumulative sum (Σmax) and average bioluminescence (BLmean) provided a BL kinetic signature that could be used to identify the species responsible (Figs. 1, 2; Table 1).
The first step was to automatically identify (see Eqs. 1–6) and store the BL kinetic signature of the taxa identified in situ (Figs. 3, 4, 5). A manual and visual analysis has been done to eliminate those signatures that were clearly representative of more than one taxa, and therefore difficult to identify and classify (Fig. 2a, b). All the selected data are then further processed and identified. The eliminated data will still take part in the statistical analysis to provide information about the “unidentified data.” The BL kinetic signatures in Fig. 2 has been calculated for all the peaks measured during the time series at 20 m depth (Figs. 3, 4) and compared with the species-specific BL characteristics (Fig. 1; Table 1 elucidated during the laboratory experiments (5 taxa, Table 2). By putting specimens of different taxa into the UBAT, one by one, we detected BL and non-BL zooplankton.
Equation 2 is the normalized error between the maximum bioluminescence (BLmax) of the BL kinetic signature of the unknown taxa measured in situ and the parameter of the kinetic BL signature of the taxa measured in laboratory. In this, we considered each parameter to have the same importance (the same weight in Eq. 1), that means: w1,w2,w3,w4 = 1.
All unclassified flashes have deleted manually (using the approach shown in Fig. 2) from further analysis. The unrecognized BL signatures need further detailed experiments under controlled conditions, but is beyond the scope of this paper.
In the plot, the cumulative sum of BL related to the identified peaks has been integrated over the hours, divided between taxa and compared (Eqs. 5–6). Unclassified peaks have not been included in Fig. 5, but they will be considered in a further refinement to have a more complete overview about the total amount of light present in a certain hour.
The integrating factor khour depends by the amount of sampled minutes over the hour, basically the integrating factor extrapolates the measured data per hour.
Specific-specific characteristics in the laboratory
The four dominant bioluminescent taxa in the laboratory study included (with number of specimens in parentheses) M. longa (12), M. ovum (11), B. cucumis (5) and Meganychtiphanes norvegica (7), all of which were collected and identified from samples taken next to the in situ platform. All bioluminescent species (BL) gave BLmax ranging from 2 to 100 × 109 photons s−1. Bioluminescence characteristics from monocultures of these were measured according to the following BL characteristics by means of BL maximum value, average value, maximum value time and rising time (see Fig. 1a; Table 2 for details).
Seven taxa were found to be non-bioluminescent in this study (number of individuals in parentheses): The pelagic amphipods Themisto libellula (n = 14) and T. abyssorum (n = 8), the arrow worms Parasagitta elegans (n = 15) and Eukrohnia hamata (n = 15), the calanoid copepod Calanus glacialis (n > 100), the krill Thysanoessa sp. (n = 50), and the pteropod Clione limacina (14).
These species gave a BL signal up to 3 × 106 photons s−1, i.e., equivalent to the background signal of filtered water. In contrast, the BLmax values for the true bioluminescent taxa ranged from 2–10 × 109 photons s−1, 1,000 times the background signal.
Identification of in situ bioluminescence peaks
Based upon the taxon-specific bioluminescence characteristics (see Fig. 1; Table 2), five identified taxa were identified as active throughout the period. In total, more than 50 % of the peaks could unequivocally be assigned as being produced by M. longa, whereas dinoflagellates accounted for one-third of all identified peaks. The ctenophore B. cucumis, although numerically the least dominant, had a BL signal one order of magnitude stronger than any of the other taxa (Lapota et al. 1992a, b).
The total BL signal from the zooplankton community was measured in situ during a 51-h deployment of the bathyphotometer at 20 m depth (Figs. 2, 3, 5), covering three noon and two midnight periods. The variation in total number of measured peaks varied throughout the sampling period, no clear diurnal signal could be detected.
A total of 2,981 BL flashes were detected during the 51-h time series. Two thousand seven hundred and fifteen different BL peaks were identified, based on the species-specific BL characteristics (see below) comprising M. ovum (20 flashes); M. norvegica (286 flashes), B. cucumis (30 flashes), M. longa (1,344 flashes) and dinoflagellates (1,031 flashes), see Fig. 2. During the second part of the deployment (January 16–17), at total of 266 were identified; M. ovum (10), M. norvegica (44), B. cucumis (6), M. longa (121) and dinoflagellates (81), Fig. 2. The latter group was identified by the bioluminescent characteristics of dinoflagellates (Table 1; Fig. 1).
Identification of biolum taxa based on the laboratory measurements
Major bioluminescent taxa during TS measurements at 20 m depth in Kongsfjorden
15–16 January specimens (% of total)a
16–17 January specimens (% of total)b
Copepod (M. longa)
Krill (M. norvegica)
Ctenophore 1 (M. ovum)
Ctenophore 2 (B. cucumis)
The bioluminescent fraction of the zooplankton community has previously been examined from the same site and at the same time of year, but using a different method deployed herein. Berge et al. (2012) reported on the bioluminescence measured by the same UBAT mounted on an AUV, concluding that bioluminescent dinoflagellates were conducting vertical migrations within the upper 30 m of the water column. Later, Moline et al. (in press) reanalyzed the same dataset and distinguished three main clusters of flashes, arguing that they represented three different groups of bioluminescent taxa. The BL flash kinetics from dinoflagellates (data from Kongsfjorden in January 2010) belongs the low-intensity BLmax group called G1 (Lapota et al. 1992a, b) which was the largest BL group recorded in these waters during polar night. Two other BL groups were recorded (BL group G2 and G3) in Moline et al. (in press), but these groups were not identified. In Table 3, BLmax indicates that B. cucumis can be put into BL G3, and the rest in BL group G2. However, Moline et al. (in press) also identified two other clusters, named group 1 and 2, respectively. Based on the laboratory-controlled bioluminescence characteristics of four other taxa defined herein (Table 2), we conclude that the group 3 in Moline et al. (in press) most likely represent B. cucumis. Especially the BLmax and BLmean correspond with those herein described as characteristic of this large and common ctenophore (Majaneva et al. 2013). The second group (Group 2), however, does not unequivocally correspond to any of the four taxa characterized in Table 2, although the high Σmax might indicate that this group is also a ctenophore—the smaller but more numerous M. ovum
In conclusion, we argue that the flash kinetics of bioluminescent zooplankton is a useful and easy way of characterizing zooplankton communities, particularly so at high latitudes during the otherwise challenging winter conditions. In general, bioluminescence of arctic zooplankton is in itself a poorly studied topic, one that may provide new and vital information concerning both ecological interactions and physical forcing of biological processes during the arctic winter. Specifically, our study demonstrates that the bioluminescence flash kinetics of arctic zooplankton has clear species-specific characteristics that allow for in situ identification of species.
The work is part of the two NFR projects Circa (Project Number 214271/F20) and Marine Night (Project Number 226417/E10). Furthermore, the work was supported by the CoE AMOS at NTNU (NFR 223254). Thanks are given two three anonymous reviewers for constructive comments.
- Båtnes AS, Miljeteig C, Berge J, Greenacre M, Johnsen G (2013) Quantifying the light sensitivity of Calanus spp. during the polar night: potential for orchestrated migrations conducted by ambient light from the sun, moon, or aurora borealis? Polar Biol DOI 10.1007/s00300-013-1415-4, Electronic supplementary material doi:10.1007/s00300-013-1415-4
- Berge J, Cottier F, Last K, Varpe Ø, Leu E, Søreide J, Eiane K, Falk-Petersen S, Willis K, Nygård H, Voegedes D, Griffiths C, Johnsen G, Lorenzen D, Brierley AS (2009) A diel vertical migration of Arctic zooplankton during the polar night. Biol Lett. doi:10.1098/rsbl.2008.0484 PubMedCentralPubMedGoogle Scholar
- Moline MA, Berge J, Johnsen G, Båtnes A, Blackwell S (in press) Bioluminescence flash kinetics characterize pelagic community structure. J Plankton ResGoogle Scholar
- Sakshaug E, Johnsen G, Kovacs K (2009) Ecosystem Barents Sea. Tapir Academic Press, Trondheim 587 ppGoogle Scholar