Morphological measurements

3.2.1.1. Herring swimbladder volumes

Data on the volume of Baltic herring swimbladders were collected at Forsmark nuclear power plant (60º24′N 18º10′E), on the Swedish east coast north of Stockholm in 1988. In order to prevent damage to internal organs the fish were caught close to the surface by land seines and were allowed to swim freely to a 5 m3, 1 m deep holding tank at the experimental site. Total lengths of the herring ranged from 17.2 to 32.6 cm. The fish were held in the holding tanks where they were allowed to adapt to surface pressures for 7-10 days. Immediately before taking swimbladder volume measurements, the herring (n = 104) were scooped over to a smaller tank in batches of five, where they were anaesthetised with 300‰ benzocaine. The swimbladders of the fish were then emptied by gentle ventral massage from beneath the pelvic fins

towards the anal opening. Swimbladder gas was collected with an inverted funnel suspended beneath the water with a top mounted glass burette. Swimbladders were then emptied completely by underwater dissection and their volume measured to the nearest 0.1 ml. No residual gas was found in the main chamber or the anal duct after inspection. Each fish was weighed to the nearest 1 g and its length measured to the nearest 0.1 cm. Finally, all fish samples were analysed for percentage fat content by standard methods using sodium sulphate grinding and ethyl ether extraction (see Brawn 1969). These measurements were compared to swimbladder volumes obtained from Norwegian spring-spawning herring (Ona 1990) collected in 1983 at Skogsvåg (60º15′N 5º05′E) in western Norway using the same methods.

Swimbladder volumes were also calculated assuming that herring aim to minimise energy expenditure by maintaining a state of neutral buoyancy (i.e. the density of the whole fish has to equal the sea water density). For the fish to gain neutral buoyancy, the swimbladder has to acquire a certain volume (Vsb, in %)

relative to the whole fish volume in order to balance the density of the surrounding sea water (ρsw) and fish body (ρfish):

sb fish sw fish sb ρ ρ ρ ρ V − − = (3.1)

The density of the swimbladder gas (ρsb) was assumed to be 0.0013 g cm-3 (Brawn

1969). Sea water density was calculated according to an algorithm developed by Fofonoff and Millard (1983) using an equation of the form ρsw = ρ(T,S,z), where the

density is a function of temperature (T), salinity (S) and depth (z). Uniform values for temperature of 10 ºC and salinity of 35 were assumed for the Norwegian Sea. Values of 10 ºC and 7 respectively were chosen for the Baltic Sea.

Density of the fish body (ρfish, i.e. whole fish excluding the swimbladder) was

calculated using an adapted proportion key for the volume proportions of various body components (Vf, Vsc, Vb, Vr for fat, scales, bones and the ‘rest’ of the body,

Vf : Vsc : Vb : Vr = [Vf] : [0.5] : [1.2] : [98.3 - Vf] (3.2)

where fat contents (Vf) were based on measurements made on the collected herring

samples (Norwegian spring-spawning herring: mean = 18.4%; Baltic herring: mean = 7.3%). Considering the steadily decreasing trend in fat content of herring in the Baltic proper since the early 1980s (Cardinale and Arrhenius 2000), present values were based on recent findings of Bignert et al. (2007) and were assumed to be 2.1%. Swimbladder volume estimates for Baltic herring were performed using both the 1980s and current fat content values.

The density of the fish body was calculated by dividing the sum of the weighed densities (the product of volume proportion and density) of each body component by the sum of all body volume proportions:

r b sc f r r b b sc sc f f fish V V V V V V V V + + + ρ + ρ + ρ + ρ = ρ (3.3)

The following density values were used for both herring stocks according to data published by Brawn (1969): fat, ρf = 0.926 g cm-3; scales, ρsc = 1.966 g cm-3; bones,

ρb = 1.993 g cm-3; and the ‘rest’ (= other body tissues), ρr = 1.057 g cm-3. Mean

values of ρfish were 1.049 g cm−3 for Norwegian spring-spawning and 1.063 g cm−3

for Baltic herring. By inserting ρfish into Equation (3.1), swimbladder proportions of

total fish volume could be estimated theoretically for both Norwegian spring- spawning and Baltic herring cases. Conversion to absolute volumes was achieved by calculating the volume of the fish body by dividing the weight by the density (ρfish).

Predicted model and observed empirical values were then compared using correlation and linear regression analysis.

3.2.1.2. Baltic clupeid body and swimbladder dimensions

Baltic herring and sprat samples were collected in October 2002 on the Swedish component of the Baltic International Acoustic Survey (BIAS) in the Baltic Sea (ICES Subdivisions 25, 27 and 29). Live fish were selected from the catch and placed in a tank with seawater immediately after they were hauled on board. Fish that were still swimming horizontally after 2-5 minutes were carefully transferred into an anaesthetic bath (4-6‰ clove oil solution) with a small net. The fish were left in the anaesthetic bath for 5 minutes and then measured (total length, to the nearest 0.5 cm; maximum height and width, to the nearest 0.1 mm). Afterwards, the fish were frozen and X-rayed (at 20 kV and 900 mA). Maximum dimensions (length, height and width) of the swimbladder were measured using good quality X-ray images of the herring (n = 25; length: 13 - 24.5 cm) and sprat (n = 21; length: 7 - 13.5 cm). The narrow extensions at the anterior and posterior ends of the swimbladder were excluded from the measurements.