Growth, metabolism and growth efficiency of a ... - Oxford Journals

Journal of Plankton Research Vol.17 no.9 pp.1757-1769,1995

Growth, metabolism and growth efficiency of a euphausiid crustacean Euphausia pacifica in the southern Japan Sea, as influenced by temperature Naoki Iguchi and Tsutomu Ikeda1 Japan Sea National Fisheries Research Institute, 1 Suido-cho, Niigata 951 and 'Seikai National Fisheries Research Institute, 49 Kokubu-machi, Nagasaki 850, Japan Abstract Growth (assessed from intermolt period and molt increment) and metabolism (oxygen consumption) of the post-larva of Euphausia pacifica from the southern Japan Sea were determined at seven graded temperatures ranging from 1 to 25°C. The intermolt period shortened progressively as temperature increased from 1 to 20°C, but an effect of temperature on molt-to-molt growth increment was not seen. Oxygen consumption rates were accelerated by the increase in temperature up to 20°C. Beyond 20°C, E.pacifica exhibited reduced oxygen consumption and died within 1 day without molting. After removing the effect of body size, the relationships between growth rate and temperature, and between oxygen consumption rate and temperature, were established. The carbon budget was calculated as a function of temperature. Because of differential effects of temperature on growth and metabolism, the net growth efficiency [K{. growth x 100/(growth + metabolism)] changed with temperature. The optimum temperature at which E.pacifica attained the maximum K2 was 11.4°C, which was derived from calculation of cumulative carbon invested in growth and metabolism in this animal. In an alternative method, the optimum temperature was obtained mathematically by solving a set of differential equations. The biological and ecological significance of the optimum temperature which leads to the

Introduction The growth of oceanic zooplankton is considered to be governed primarily by temperature and food (Huntley and Boyed, 1984). To date, numerous laboratory experiments have been carried out on various zooplankton groups to evaluate the effect of food on the growth rate and growth efficiency [see Paffenhofer and Harris (1979) for a review]. In contrast, relatively less is known about the effect of temperature. Temperature is known to exert differential effects on various physiological rate processes of zooplankton, i.e. feeding and oxygen consumption rates of various copepods (Anraku, 1964), ingestion, growth and oxygen consumption rates of euphausiids (Sameoto, 1976; Ross, 1982a) and a mysid (Toda et al, 1987), and growth and oxygen consumption rates of an amphipod (Ikeda, 1991). This suggests that growth efficiency, as a consequence of complex physiological rate processes, varies with temperature. However, detailed information about the optimum temperature to maximize net growth efficiency at unlimited feeding is currently limited to a mysid (Toda etal., 1987) and an amphipod (Ikeda, 1991). The euphausiid Euphausia pacifica is distributed widely in northern North Pacific and its marginal seas, including the Bering Sea, Okhotsk Sea and Japan Sea (Brinton, 1962). From seasonal distribution and life history data for this euphausiid in the southern Japan Sea (Iguchi et al., 1993), the entire temperature range which E.pacifica could encounter during its lifetime is estimated to be as wide as 1 © Oxford University Press

1757

N.Iguchi and T.Ikeda

to >20°C. Lasker (1966) and Ross (1982a) carried out laboratory experiments on E.pacifica and determined its net growth efficiency, but these previous data were obtained at only one (Lasker, 1966) or two temperature (Ross, 1982a), and are insufficient to generalize the effect of temperature on the material budget of the post-larva of E.pacifica in the Japan Sea. In this study, we determined the growth and metabolism of E.pacifica under various temperature regimes that cover its entire natural range. We discuss the optimum temperature at which the net growth efficiency of this animal is maximal. Materials and methods

Animals Euphausia pacifica (juveniles and adults) were collected from an offshore station in Toyama Bay (37°0'N, 137°14'E), southern Japan Sea, in December 1991 and April 1992. Collections were made with oblique hauls of a fish-larva net (0.5 mm mesh) or a 2 m Issac-Kidd Midwater Trawl (1.5 mm mesh) from 200 m depth to the surface at night. On board ship, live specimens of E.pacifica were quickly sorted from the catches and used for immediate oxygen-consumption experiments. Some of the specimens collected in December 1991 were maintained at ~10°C on board ship and transported to the land laboratory for growth experiments. At the same sampling site, seawater was collected from 250 m depth with Niskin bottles and kept in 20 1 plastic containers for use in the experiments. Body allometry The body length (BL: the maximum distance between the tip of the rostrum and the distal end of the telson excluding spines) and the exopodite length of the uropod (EL) of freshly collected specimens were measured under a dissecting microscope to the nearest 0.025 mm. The specimens were rinsed briefly with a small amount of distilled water, blotted on filter papers and kept frozen (-20°C). Later, frozen specimens were freeze-dried for the determination of dry weights (DW). Growth Growth of euphausiids is achieved through regular moltings, hence the rate of growth in length (GL) of E.pacifica was estimated by determining molt interval period (IP) and molt increment (MI) in this study (i.e. GL = MI/IP). Specimens were placed individually into 1000 ml glass containers filled with seawater. A pennate diatom, Phaeodactylum tricornutum, cultured in f/2 media at 10°C, was provided as food at concentrations of 3-6 x 104 cells ml-1, which are equivalent to 300-600 u.g C I 1 . The cell size (maximum 19 ^m) of P.tricornutum falls in the 12-64 \xm range effectively grazed by E.pacifica (Parsons et al., 1967), and its concentration in terms of carbon is close to or greater than the maximum reported from the euphotic zone of the Japan Sea and northern North Pacific ( 0.05). For this reason, MI data for the specimens with 8-14 mm BL were pooled at each temperature and the effect of temperature was analyzed (Table II). The analysis of variance indicated no significant difference among Af/data for different temperatures (F = 0.843, df = 4,110, P > 0.25), yielding a grand mean MI of 0.16 mm. On the basis that MI is independent of T, more precise MI can be obtained from the data on natural population growth and natural habitat T, combined with equation (3) of this study, i.e. MI = GL x IP. For the population of E.pacifica in Toyama Bay, the natural growth rate was estimated as 0.102 mm day 1 for juveniles at a mean temperature regime of 6.2°C (Iguchi etal., 1993). Then: Ml = 0.102 X

(4)

We used MI derived from natural growth rate in the following calculations [the choice of MI experimentally obtained (0.16) or that derived from natural growth rates (0.1 x Mfumu"- * o^snj does not affect the estimate of optimum temperature to maximize K^\. Growth rate in length (GJ or in dry weight (Gw) at given size (BL) and temperature is now predicted by a combination of equations (3) and (4), or (2), (3) and (4). 1760

Temperature and growth efficiency of ELpacifica

• This study

o Jerde and Lasker (1966) • Paranjape(1967) A Fowler era/. (1971)

o.o -0.2 -0.4

s

-0.6 -0.8 -1.0 -1.2 10 T(°C)

20

15

Fig. 2. Relationship between intercepts (£>) of the IP-T relationship in Figure 1 and temperature (7". °C) for E.pacifica (Iog106 = 0.0340 - 0.04357, r = -0.999, N = 5). The figure includes data reported by Jerde and Lasker (1966), Paranjape (1967) and Fowler el at. (1971) on the same species. These data by other workers were re-calculated assuming the same v (= 0.0321) for the IP-BL relationship established in this study. Table I. Regression analysis of intermolt period (IP, days) on pre-molt BL for E.pacifica. Regression model is logT0 lP = vBL + b (v and b are constants). N, is number of specimens and A', is number of molts collected. Since 95% CIs of v overlapped each other, a common v was calculated (excluding fewer data sets at 1 and 20°C), and the resulting b was shown in the last column Expt T

1 5 10 15 20

N,

4 5 5 5 4

N.

9 27 37 47 15

w(95%CI)

0.0267 (0.0030-0.2352) 0.0182 (0.0034-0.0970) 0.0303 (0.0076-0.1206) 0.0415 (0.0108-0.1598) 0.2505(0.0315-1.9900)

r

0.664 0.582 0.716 0.651 0.433

bfor v = 0.0321

Mean BL

log,./*'

11.303 11.045 12.643 13.056 9.557

1.344 1.008 0.796 0.670 0.450

0.9807 0.6534 0.3905 0.2480 0.1435

Oxygen consumption versus temperature Oxygen consumption (R: |il O2 individual"1 Ir1) data were normalized as specific oxygen consumption (R^) of the specimens weighing 1 mg DW, dividing by mg DW*' after Ross (1982a). This was done because the DW range of the specimens used in this study (1.4-19.8 mg DW) was too narrow to resolve the effect of DW. Preliminary semi-log plots of R^ obtained in December 1991 and April 1992 against temperature (7°, C) separately indicated that R^ increased exponentially with the increase in temperature up to 20°C in both cases. Beyond 20°C, the R^ decreased progressively in both cases. The results of a covariance test revealed that 1761

N.Igucbi and T.lkeda Table II. Variance analysis of molt increment (MI, mm) obtained at five different temperatures for E.pacifica (8.0-14.0 mm BL). Means ± 1SD. See the text for details.

Exptr

N

1 5 10 15 20

9 25 30 37 14

Ml

Test for the effect of T

(mean ± 1SD)

CO

F= 0.843 df = 4,110 />>0.25NS

0.08 ± 0.13 0.11 ±0.20 0.18 ± 0.28 0.18 ± 0.25 0.22 ± 0.36

the regression lines from December data and April data were essentially the same (for slope, F=0.033, df = 1,66, P > 0.25; for intercept, F= 2.806, df = 1,67, P > 0.05). With these results, both data sets were pooled to analyze the relationship between R^ and T (Figure 3), producing: 'ognAdj = 0.0347(±0.0051, 95%CI)T + 0.0036 (r = 0.853, N = 70)

(5)

Data for the 22 and 25CC experiments were omitted in this regression analysis because these temperatures were known to be outside the tolerance limit of E.pacifica from growth experiment data mentioned above. Conversion to carbon units Since the carbon content (% of DW) of juvenile and adult E.pacifica in Toyama Bay is rather stable over various seasons of the year (41.0-45.1%; N.Iguchi and T.lkeda, unpublished data), without an appreciable difference between males and

24

Fig. 3. Relationship between adjusted oxygen consumption rate (R^, (jj O ; mg~' DW-aii rr') and temperature (T, °C) (log^.,, = 0.0347 T+ 0.0036, r = 0.853, N = 70). Closed circles and vertical lines crossing the circles are means and their 95% CI. Numerals alongside the solid circles denote the number of data.

1762

Temperature and growth efficiency of E.pacifica

females, a grand mean value of 43% was used in this study. Hence, the daily growth in D W (Gw) is expressed in carbon units as: G c = 0.43Gv

(6)

The adjusted oxygen consumption rate at 1 mg DW (R^) was converted to a carbon unit per day basis (Rc) assuming protein metabolism (RQ = 0.97; cf. Gnaiger, 1983): Rc = 0.97 x 12/22.4 x 24 x R^ = 12.47 R^,

(7)

where 12/22.4 is the carbon mass in 1 mol of CO2 and 24 is hours in a day. Net growth efficiency versus temperature Net growth efficiency was defined as growth x 100/(growth + metabolism) in this study. The amount of carbon associated with cast molts was ignored because of its minor contribution to the lifetime carbon budget of E.pacifica (5% of the sum of assimilated carbon; Ross, 1982a). Further, the fraction of carbon in each molt is known to be proportional to body size, not to temperature in E.pacifica (Ross, 1982b). Carbon invested in reproduction was also not considered in this calculation, as little is known of the effect of temperature on reproduction of this animal. Cumulative Gc and Rc were computed for a specimen growing from 6 to 23.2 mm BL (an arbitrarily selected range) through 17 consecutive molts at each 1°C interval between 1 and 20°C. The K2 at each temperature is illustrated in Figure 4. The K2 increased with increasing Tup to 36% at around 11°C, then decreased.

T

i

i

i

i

i

i

i

i

>

i

i

i

i

i

i

i

i

i

i

Fig. 4. Relationship between net growth efficiency (ATJ calculated from carbon budget and temperature

(r,°C)for E.pacifica.

1763

N.lguchi and T.lkeda

The precise optimum T to maximize K2 can be solved mathematically. As defined above: /f2 = C c (G c + /?c)-' = (l + /? c /C c )- 1

(8)

Since Rc > 0 and G c > 0, it is evident from equation (8) that K2 becomes the maximum when RfJGc is the minimum. Both functions Rc and G c are continuous against T, so that the optimum 7" can be obtained from the derivative of i.e. d(Rc/Gc)/dT = {(dRc/dTjGc- RddGc/dT^yCc2 = 0 This leads to (dRJdT) Gc - flc(dGc/dr) = 0

(9)

Daily growth in BL (ABL) is defined already as MI/IP. Differentiating equation (2), daily growth in DW (= &DW) is: ADW= (9.954 x lfr^) x 3.156 x BL2156ABL = A x ASL Substituting from equations (3) and (4): A D W = A X 10-0*102 _ 10-0W357- • 0.0340

where A = (9.954 x KH) x 3.156 x BL1156, A' =Ax 10-04102, B = -0.0435 and C = 0.0340. D W is converted to carbon units by multiplying by 0.43 [see equation (6)], then we get: G c = 0.43 x DW = 0.43 x A' x io- loSr + c = 0.43x/l'xe-^7"+c

(10)

where a = 2.3026, B' = 2.30265 and C = 2.3026C. Then

From equations (5) and (7), Rc is: Rc = 12.47 x /? = 12.47 x li (12) 1764

Temperature and growth efficiency of ELpadfica

where a = 12.47 x lfF**, p = 0.0347 and p' = 2.3026p. Then: p'Rc

(13)

Substituting equations (10), (11), (12) and (13) into equation (9): Gc-flc(dGc/dr)

= RcGc(p> + aB'eBT*c) = 0

This leads to e r T + c = -p/afl'. Solving this equation for T, we obtain: T=(\n(-$'/aB')-C')/B' = 11.365 (=* 11.4°C).

(14)

It is now clear that the optimum temperature is determined by only three parameters (p in the R-T relationship and B and C in the IP-T relationship). A sensitivity test, in which these three parameter values in equation (14) were manipulated in the order of ± 10%, showed that the effect of P was ± 1°C, whereas that of B or C was only ±0.1°C, indicating p as the most important parameter to determine the optimum temperature. When 95% CI associated with p are considered, equation (14) yielded an optimum temperature ranging from 10.0 to 13.0°C. Discussion The life history pattern of E.pacifica is known to vary geographically. Their breeding season is February to May in Toyama Bay, southern Japan Sea (Iguchi et al, 1993), June in Okhotsk Sea (Ponomareva, 1963), September off Oregon (Smiles and Pearcy, 1971) and all seasons in Southern California (Brinton, 1976). Life span ranges from as short as 8-12 months off Southern California to 2 years in the Okhotsk Sea and waters around the southern Aleutians (Nemoto, 1957; Ponomareva, 1963). In light of this plasticity of life history patterns, a comparison of physiological parameters (P, B and C) used for the calculation of optimum temperature between E.pacifica inhabiting different geographical locations is a prerequisite to generalizing the present results. The IP to T relationship, which includes parameters B (slope) and C (intercept) for E.pacifica, given by previous workers is compared with the present results in Figure 2. Data sets of both Jerde and Lasker (1966) from off California and of Paranjape (1967) from around Vancouver Island lie very close to our regression line. Fowler et al. (1971) gave IP data of E.pacifica off Oregon at three temperatures (5,10 and 15°C). Among these three data sets of Fowler etal. (1971), the 5 and 10°C data can be compared favorably with our results, but his 15°C data depart appreciably from our regression line. As is also seen in the R to T relationship (discussed below), E.pacifica living in the eastern North Pacific appear to have a narrower range of temperature tolerance than those in the Japan Sea. Lasker (1966) considered that temperatures >12°C did not accelerate molts further, and 3 days is the minimum IP for E.pacifica off California. Clearly, this is not the case for 1765

N.Igucfai and T.Ikeda

E.pacifica in the southern Japan Sea (see Figure 1). However, after taking into account this geographical difference in the range of temperature tolerance, the IP-Trelationships of E.pacifica reported from the eastern North Pacific agree well with the present results. There are no comparable eastern region IP data either below 5°C or above 15°C for E.pacifica. The other component of the growth of E.pacifica, MI, was not affected by temperature in this study (cf. Table II). Temperature independence of Ml was also reported for the mysids Leptomysis linguura and Hemimysis speluncola over the temperature range from 10 to 22°C (Gaudy and Guerin, 1979), and the amphipod Themisto japonica over the temperature range from 1 to 12°C (Ikeda, 1990). According to the review of Hartnoll (1982), the effect of temperature on MI is highly variable, and no consistent pattern was seen over many crustaceans. The MI of E.pacifica recorded from the readings of molts was less than that estimated from its natural growth rate in this study. The quality of food used may be considered as of prime importance for this reduced MI. From this review, malnutrition of E.pacifica due to the use of P.tricornutum as a staple food in this study is unlikely, since this diatom has been used successfully to raise a euphausiid (Euphausia superba) from larvae to mature adults in the laboratory (Ikeda, 1987). Since the reduction of MI was not seen in the euphausiids grown from eggs in the laboratory (Ross, 1982; Ikeda, 1987), this may reflect a failure of wild specimens brought into the laboratory to acclimate to the experimental conditions (i.e. limited space of rearing container, light conditions, handling, etc.). Lesser Ml of laboratory-reared specimens, compared with MI estimated from the growth of field populations, has been observed in the mysid Gnathophausia ingens (Childress and Price, 1978) and the hyperiid amphipod T.japonica (Ikeda, 1990). The parameter p characterizes the R-Trelationship, which may change seasonally within the same habitat (Gilfillan, 1972), but this was not found by Small and Hebard (1967) or in the present study (see Results). The parameter p is a temperature coefficient of metabolism, and is proportional to the Qm [= e23026"l0p, see equation (3) in Ikeda (1985)]. In terms of 01O, P of 0.0347 (95% CI: 0.0296-0.0398) obtained in this study is equivalent to 2.22 (1.98-2.50). The present results of Qw values are compared with those being reported by other workers on E.pacifica (Table III). Despite differences in the methodology for measuring oxygen consumption, in the number of temperature settings, and in habitats of this animal, the d o = 1.98-2.38 reported by other workers falls well within the range obtained in this study. An anomalous Qw = 0.28, obtained by Paranjape (1967), suggests that the temperature of 20°C is out of the natural temperature range for E.pacifica off British Columbia. Swimming activity during the experiments is known as a factor affecting the Qw of oxygen consumption rates of E.pacifica (Torres and Childress, 1983). In this regard, all experiments in Table III were carried out without controlling the activity of E.pacifica during the experiment; hence, the results may be inferred as being close to routine metabolism (cf. Ikeda, 1985). Thus, all three parameters gained in the present study agree well with those obtained by other workers from different habitats of E.pacifica, although both higher and lower ends of the temperature range within which this animal can thrive appear to be dissimilar. On these grounds, the present conclusion of the optimum 1766

Temperature and growth efficiency of Kpacifica TaMe m. Qla values for the oxygen consumption rates of E.padfica Co

Method-

Reference

Off British Columbia, Canada 5,10,15, 15,20 Off Oregon, USA 5,10,15 Off Oregon, USA 5,10,15

2.38 0.28 2.11 1.98"

Water bottle

Paranjape (1967)

Manometric Manometric

Off Washington, USA Toyama Bay, Japan Sea

2.0

Small et al. (1966) Small and Hebard (1967) Ross (1982) This study

Habitat

ExptT

8,12 1,5,10,15,20

Water bottle 2.22 Oj electrode (1.98-2-50)
20°C in summer, but very cold water (