Seasonal dynamics of the vertical migration behaviour of mesopelagic fish
MetadataShow full item record
This thesis focuses on the behavioural dynamics of mesopelagic fish in a fjordic environment. Acoustic data recorded with a stationary echosounder was used to primarily investigate the diel vertical migration behaviour of the mesopelagic fish Maurolicus muelleri. Particular emphasis was on the effect of daily and seasonal changes in surface light irradiance and the vertical distribution of large predator-like fish on M. muelleri. In addition the behaviour of the deeper distributed mesopelagic lantern fish Benthosema glaciale and large, individual fish were also investigated. Diel vertical migrations (DVM) are vital in maintaining the energy flow in marine ecosystem food webs, and migrating organisms thus directly and indirectly provide food/nourishment to many non migrating animals. Typically, DVM are characterised by an ascent to surface waters from deeper day time depths at dusk, followed by a nocturnal period close to the surface and a subsequent descent to daytime depths at dawn. M. muelleri and B. glaciale are abundant and ecologically important fish in several marine systems. They are part of the mesopelagic animal community that performs DVM - nature’s largest migration in terms of biomass. In the Norwegian Sea and fjords M. muelleri forms acoustically visible scattering layers (SLs), while the deeper living B. glaciale appears acoustically more scattered and less dense, not aggregating in distinct SLs. Both species are challenged with seasonally changing gradients of environmental variables including light, temperature, salinity, food abundance and piscivorous predators. Although several aspects of the behaviour and biology of M. muelleri and to a lesser extent of B. glaciale have been studied, our knowledge of short term and seasonal variation of individual fish and SL behaviour, as well as behavioural responses to seasonal changes in surface light irradiance and predator distribution is limited. To address these shortcomings we therefore utilised the fairly recent developments of stationary echosounder assemblages, located on fjord bottoms, which made it possible to collect continuous acoustic records of the entire water column over a 15 month period. Here I present 4 papers covering the description of a stationary acoustic platform and its application in studying the temporal and spatial behavioural dynamics of B. glaciale individuals and layers as well as large individual fish (Paper 1), the monthly and seasonal variations in DVM behaviour of M. muelleri SL (Paper 2), the relationship between daily and monthly changes in surface light intensity and M. muelleri SL depth (Paper 3), and the vertical distribution of M. muelleri SL in relation to seasonally changing predator distribution (Paper 4). By collecting continuous acoustic data with an upward facing split beam echosounder located on the fjord bottom it was possible to observe biotic changes in the water column within the course of days, weeks, and months (Paper 1-4), as well as investigate the vertical distribution and swimming behaviour of individual fish by applying a target strength threshold and target tracking (Paper 1 and 3). Paper 1 specifically describes the reverse diel migrations during day time of the deeper distributed (> 200 m) lanternfish B. glaciale and its swimming behaviour, which was mostly passive and more active during stepwise vertical excursions. Paper 1 also deals with depth related differences in large fish swimming behaviour and target strength. Shallower distributed large fish (< 300 m) had both higher target strengths and swimming speeds than deeper distributed fish (> 200 m) lanternfish B. glaciale and its swimming behaviour, which was mostly passive and more active during stepwise vertical excursions. Paper 1 also deals with depth related differences in large fish swimming behaviour and target strength. Shallower distributed large fish (< 300 m) had both higher target strengths and swimming speeds than deeper distributed fish (> 300 m). Paper 2 reports on novel behaviours of M. muelleri ontogenetic layers; early morning ascents, reverse diel migrations, and arrested migrations. Previous studies have shown that juvenile M. muelleri perform midnight sinking, i.e. the relocation to deeper waters after dusk, in January in addition to ordinary DVM between January and June. Adult fish on the other hand do not display DVM in January, and stayed in deep waters throughout the diel period, while migrating vertically between the surface and deeper day time depths in spring (May and June). In addition to these previous observations we found that a proportion of ascending juvenile M. muelleri arrested their migration between January and April. Further, adult fish displayed early morning ascents to the surface (in the absence of light) in September and October, and reverse migrations in November until January by migrating approximately 20-30 m upwards at dawn instead of descending. Adult fish ‘overwintered’ for approximately 6 months, whereas younger fish displayed midnight sinking between August and beginning of May. Paper 2 also demonstrated that when three SLs were present, the shallowest SL was composed of post-larvae 10mm long. The observed behavioural patterns are interpreted as responses to individual physiological state, hunger-satiation and condition, prey availability and predation risk at alternative depths. Light affects many aspects of fish behaviour. The preferendum hypothesis states that organism change their depth during DVM because they follow a constant light level, i.e. an isolume. In Paper 3 the upper border depth of M. muelleri SLs was positively, linearly correlated to surface irradiance, which could suggest that M. muelleri follows an isolume. However, this assumes an invariant light attenuation coefficient. On the contrary we found that the light intensity of the isolume was not constant. It tended to vary with migration phase, whether fish were ascending or descending and it also changed with month. We therefore suggest that M. muelleri follow a preferred light range instead of a strict isolume, which reflects a more dynamic relationship. Differences in preferred light range also appear to be related to ontogeny and physiological state. In paper 4 we applied target tracking to study the depth distribution of large, putative predatory fish. High numbers of predator-like fish at the surface at night during autumn and winter month, October – March, appear to cause juvenile M. muelleri to relocate deeper after dusk in order to avoid these predators. However, juveniles continued to decent at dawn in autumn and winter months despite the absence of pelagic daytime predators at this time. This suggests that M. muelleri has an innate (genetically fixed) behavioural response synchronised with the light cycle and triggered by a change in surface light intensity. The presence of high numbers of predators in shallow waters between dawn and dusk may explain why M. muelleri avoided the shallow waters at day time in spring and summer months. An increasingly deeper daytime distribution of predator-like fish in the course of summer may also have influenced the vertical distribution especially of the deeper SLs. The overall findings in this thesis show that vessel independent acoustic observation platforms are well suited to detect and observe both short term and seasonal changes in the behaviour of both layers of fish and individuals. By combining acoustic data with environmental, biological and catch data, I have shown that the DVM behaviour of mesopelagic individuals and SLs is much more complex than previously believed, with fish displaying hitherto undescribed temporally varying behavioural patterns. My studies corroborate and supplement the current understanding of ecological interactions and processes that persist in fjordic environments. Behavioural aspects of M. muelleri have been used to model responses to environmental constraints and the life history of this species, and results presented in this thesis thus have the potential to improve such models. Due to their vital role as prey for commercially exploited fish species in marine ecosystems, mesopelagic fish species have in recent years been the focus of acoustic surveys to estimate their abundance and determine their distribution. A better understanding of their behaviour can aid in improving such estimates and contribute to the improved ecosystem based management of marine resources.
Paper 1: Marine Ecology Progress Series 395, Kaartvedt, S.; Røstad, A.; Klevjer, T.; Staby, A., Use of bottom-mounted echo sounders in exploring behavior of mesopelagic fishes, pp. 109–118. Published version. Copyright 2009 Inter-Research. The published version is also available at: http://dx.doi.org/10.3354/meps08174Paper 2: Staby, A.; Røstad, A.; Kaartvedt, S. A full year study of diel vertical migration in the mesopelagic fish Maurolicus muelleri reveals novel and varied behavioural patterns. Draft version.Paper 3: Staby, A.; Aksnes, D. L., Follow the light – diurnal and seasonal variations in the vertical distribution of the mesopelagic fish Maurolicus muelleri. Draft version. Published in Marine Ecology Progress Series 422: 265-273. Published by Inter-Research The published version is available at: http://dx.doi.org/10.3354/meps08938Paper 4: Staby, A.; Røstad, A.; Salvanes, A. G. V.; Kaartvedt, S., Seasonal occurrence and vertical migration of potential predators of Maurolicus muelleri. Draft version.