Introduction

Many species of fish undergo periods of starvation due to various factors, including temperature declines associated with seasonal changes, spawning migration, or regional and seasonal decreases in food supply (Mustafa and Mittal, 1982; Weatherley and Gill, 1987; Lee et al., 1999). To survive periods of starvation, fish utilize biochemical, physiological, and behavioral strategies in addition to endogenous reserves of energy derived from metabolic processes (Mustafa and Mittal, 1982; Weatherley and Gill, 1987; Lee et al., 1999; Park et al., 2002; Hur et al., 2006a, 2006b; Park, 2006). However, the latter can leave little energy for other biological functions such as somatic growth. As a result, growth in body size can slow considerably during starvation.

The efficiency of oxygen consumption in fish is directly related to their metabolic processes through which they can produce physical energy. In turn, this ultimately determines their population density, food supply, and fish yield (Dalla et al., 1998). Mehner and Wieser (1994) have reported the relationship between depletion of energy reserves and changes in oxygen consumption in perch (Perca fluviatilis). Starvation can decrease oxygen uptake efficiency in perch larvae and traira (Hoplias malabaricus) subjected to extended periods of starvation (Mehner and Wieser, 1994; Rios et al., 2002).

Stresses can induce the release of catecholamine and cortisol in fish, causing rapid metabolism of high-energy storage compounds (Barton and Iwama, 1991). These catabolic processes have harmful biochemical effects on fish health. They also impair growth (Specker and Schreck, 1980). Fish exhibit primary, secondary, and tertiary responses to stress (Barton and Iwama, 1991). Their primary response involves rapid changes in plasma levels of catecholamine and corticosteroid. When these responses to stressful conditions exceed normal levels, harmful secondary and tertiary responses occur. Therefore, stress can induce changes in energetic metabolic processes, reduce growth rate, interfere with reproduction, and lead to rapid changes in flesh quality following death (Barton and Iwama, 1991).

Blood chemistry can serve as an indicator of an animal's physiological state. Many factors (age, sex, nutrition, season, and circadian rhythm) may affect blood chemistry. Information is available for plasma glucose changes during starvation in Atlantic cod (Gadus morhua L.), European eel (Anguilla Anguilla), pike (Esox lucius), toadfish (Opsanus tau), goldfish (Carassius auratus), and American eel (A. rostrata) (Tashima and Cahill, 1968; Chavin and Young, 1970; Larsson and Lewander, 1973; Ince and Thorpe, 1976; Moon, 1983). Studies of plasma-free fatty acids during starvation stress have also been performed in European eel, American eel, rainbow trout (Oncorhynchus mykiss), pike, and toadfish (Tashima and Cahill, 1968; Larsson and Lewander, 1973; Ince and Thorpe, 1976; Moon, 1983; Sumpter et al., 1991). Rios et al. (2002) have noted that erythrocyte senescence and hematological changes are induced by starvation in Hoplias malabaricus.

The cobitid loach (Misgurnus anguillicaudatus) is a freshwater species in the loach family of Cobitidae (Nelson, 2006). This species is native to east Asia. As a popular aquarium fish, it has been introduced to other places in Asia, Europe, and North America. The cobitid loach inhabits mud, ponds, and rice fields that are subjected to periodic drying, resulting in starvation. As the domestic market has expanded rapidly in recent years, cobitid loach has become a commercially important freshwater species in Korea. The objective of this study was to investigate effects of starvation on physiological and metabolic profiles of cobitid loach.

Results

The starved group of cobitid loach (Misgurnus anguillicaudatus) rapidly lost vitality. Therefore, the experiment was terminated. The cumulative survival was 87 ± 1.2% in the fed group and 40 ± 2.4% in the starved group in each of triplicate tanks (Fig. 1). Starvation resulted in growth retardation which was evident based on differences in body length and depth between the fed and starved groups, although they continued to grow and remained in good condition. Results of two-way ANOVA and measurements of the oxygen consumption rate between the fed and starved groups are shown in Tables 3, 4 and Fig. 2. For the starved group, the dissolved oxygen concentration decreased to 1 ppm over 48 h. In the fed group, it decreased to 1 ppm over 28 h (Fig. 2). Table 3 shows pH values and respiratory frequency (gill cover movement) for fed and starved groups. Respiratory frequency also gradually decreased over 48 h (Table 3). The pH decreased markedly over 12 h. It then decreased gradually up to 48 h (Table 3). The respiratory frequency of the starved group was significantly higher than in the fed group. It decreased with time during the experiment (p < 0.05). Oxygen uptake per unit of respiratory movement was less for starved group than that for the fed group. CO2 and NH4+ concentrations increased markedly for 6 h. They then increased gradually up to 48 h (Table 4). Their concentrations in the fed group were significantly higher than those in the starved group. They increased during the experiment (p < 0.05).

Dose-related mortality occurred in the fed fish exposed to 120 mg/l nitrite and in the starved fish exposed to 80 mg/l nitrite (Fig. 3). Most fish survived 96 h of exposure to 0 and 70 mg/l of nitrite whereas 75% of fish survived 96 h of exposure to 110 and 120 mg/l of nitrite in the fed group or 70 and 80 mg/l of nitrite in the starved group. LC50, 95% confidence ranges, and experimental conditions for various exposure groups are summarized in Table 5 and Fig. 4. LC50 value at 96 h was 132.45 mg/l for the fed group and 93.89 mg/l for the starved group. The LC50 value at 24 h was 351.32 mg/l for the fed group and 274.91 mg/l for the starved group. The LC50 for each group increased with shorter exposure time (p < 0.05).

Figures 5~7 show variations in stress response during 48 h of exposure to sublethal nitrite concentrations (110 ppm for the fed group and 70 ppm for the starved group). In the fed group, plasma cortisol concentration increased significantly (p < 0.05) from 1.7 ± 0.2 μg/dl to 22.9 ± 3.4 μg/dl over 12 h, but decreased significantly (p < 0.05) to 8.9 ± 2.9 μg/dl at 24 h and 4.7 ± 0.4 μg/ dl at 48 h (Fig. 5). In the starved group, plasma cortisol concentration increased significantly (p < 0.05) over 12 h from 1.7 ± 0.2 μg/dl to 30.5 ± 1.5 μg/dl, but decreased significantly (p < 0.05) at 24 h (10.6 ± 1.1 μg/dl) and 48 h (4.9 ± 1.0 μg/dl) (Fig. 5). Plasma cortisol levels in fed or starved groups differed significantly between 1 and 24 h (p < 0.05). Plasma glucose concentrations in each group increased significantly (p < 0.05) over 12 h, from 29.7 ± 0.7 mg/dl to 132 ± 4.4 (in the fed group) or 240 ± 3.9 mg/ dl (in the starved group) (Fig. 6). Plasma glucose concentrations decreased significantly from 12 to 48 h. Plasma glucose levels in fed or starved group differed significantly between 1 and 48 h (p < 0.05). At 12 h, plasma cortisol and plasma glucose levels in fed and starved groups differed greatly (p < 0.05). Lactic acid content of both groups increased over 48 h (p < 0.05). For this parameter, there was no significant difference between the two groups (p > 0.05; Fig. 7).

Discussion

Larsson and Lewander (1973) have noted that many fishes undergo natural periods of starvation during the year. They consequently gain ability to withstand prolonged food shortages. Such periods may amount to weeks, months, or even years. They may cause extensive loss of energy stores in the body as fish consumes its own tissues to remain alive (Weatherley and Gill, 1987). In this study, cobitid loach (Misgurnus anguillicaudatus) were starved over 30 days. This demonstrates that this species can tolerate long-term food shortages.

When a fish is exposed to chronic stress, its metabolic reactions are altered by changes in the hypothalamic-pituitary-adrenocortical axis (HPA axis), with hypothalamic secretion of corticotropin-releasing hormone (CRH) occurring through the limbic system and the reticular formation. CRH can stimulate the release of adreno- corticotropic hormone (ACTH) from the pituitary gland which then stimulates the release of cortisol from the target organ (the interrenal gland), promoting a metabolic stress reaction. Thus, cortisol is an important index of stress reaction (Specker and Schreck, 1980). Severe starvation can result in coma and catabolic disease of the gastrointestinal system. Basic metabolic reactions maintain energy levels and body tissues (Jung et al., 2003). Guyton (1991) has reported three stages of physiological change during starvation. In the first stage, glycogen stored for immediate use is hydrolyzed, releasing glucose. In the second stage, acetyl-CoA is oversupplied relative to oxaloacetate (because of lipid use), leading to acidosis. In the third stage, fish are compromised by protein exhaustion. In the process of starvation, sugars, lipids, proteins, and other essential nutritional elements decrease rapidly, ultimately leading to the collapse of immune, circulatory, and endocrine systems that result in death (Guyton, 1991).

In our study, dissolved oxygen concentration decreased over 48 h in the starved group and over 28 h in the fed group. The respiratory frequency also gradually decreased over 48 h, indicating that respiratory function was decreased in all experimental groups during the experimental period. The pH and concentrations of CO2 and NH4+ showed different trends to those of dissolved oxygen concentration. However, decreasing rates of pH and concentrations of CO2 and NH4+ were similar to the decreasing rate of oxygen consumption. That is, the metabolic rate of fish in all experimental groups decreased during the experimental period.

Toxicity and effects of nitrite vary among fish species. They depend on test conditions, including fish size, water ionic composition, and temperature (Doblander and Lackner, 1997). Chloride and other anions in water can provide protection against nitrite during active branchial uptake (Williams and Eddy, 1986). Thus, small amount of Cl- (e.g. 1 mM) is likely to afford protection against high nitrite levels (Eddy et al., 1983). The effect of nitrite on fish is greater in Cl--poor water. Thus, nitrite is more toxic to fresh- water organisms than to organisms living in seawater (Grosell and Jensen, 1999).

Nitrite accumulation in the plasma probably causes methemoglobinemia and malfunction of hemopoietic activity, both of are effects of nitrite intoxication (Costa et al., 2004). Although methemoglobinemia is not directly related to high mortality in fish exposed to nitrite (Costa et al., 2004), the passage of nitrite into the blood stream may increase blood cell lysis (Knudsen and Jensen, 1997), changes in plasma electrolyte balance (Huertas et al., 2002), and efflux of K+ from red blood cells (RBCs; Martinez and Souza, 2002) which is evident in an increase in the number of shrunken RBCs. Dysfunctional erythrocytes may be removed from blood circulation because of oxygen shortage, causing a reduction in total erythrocyte count (Park et al., 2007). Stress response may explain the significant difference between fed and starved groups in this study.

Plasma cortisol and glucose levels are useful indicators of stress in fish (Park et al., 2008). Plasma levels of cortisol and glucose are elevated in red drum (Sciaenops ocellatus) simultaneously exposed to stressor (Massee et al., 1995). Barton and Iwama (1991) stated that "Usually, the phenomenon that plasma cortisol concentration of fishes rises by stress is the first order reaction and the phenomenon that plasma glucose concentration rises is the result of second-order reaction by hormone rise reaction caused by stress". A similar trend has been reported in gray mullet (Mugil cephalus) and kelp grouper (Epinephelus bruneus) (Park et al., 2008). In the present study, starvation in cobitid loach caused greater stress response to nitrite. Our results may be useful as a guide in the regulation and scheduling of feeding as an indicator of sodium nitrite stress and in developing an index to determine nutritional status of cobitid loach.