胃内容情報を用いたマエソ属の摂食戦略についての2・3の試論

1) Introduction : Stomach contents are a first hand clue to the feeding activities of fishes. However, due to the complexity of biotic phenomena it is very difficult to reach a clear understanding of their relationship to feeding strategies. Here, some trails were made using various perspectives based on the stomach contents of the Lizard fish.

2) Material and Method : Prey are often found in varying degrees of digestion within the stomach. In order to estimate the food intake of the fish, it is thus necessary to reconstruct to the original the weight of each prey item. Thus, ration size(r), the summation of the weights of individual prey found in the stomach, restored to their original size based on length/weight and/or vertebral column segment size/body length relationships5)7), was deemed a more accurate indicator of feeding than the unaltered stomach content weight, (f). The Lizard fish (Saurida elongata and S. undosquamis) were selected for analysis from 12 systematic experimental trawls conducted throughout the eastern Seto Inland Sea of Japan, Tab.2-1. See Notation & Abbr.,Tab. 2-2.

3) On Ration, Prey Size and Satiation in the Field : For the predator, the most favorable circumstances naturally occur when prey resources are abundant during a feeding period (Cruise 8). The quantity of prey ingested by the Lizard fishes ranged widely though composed mainly of one to several Anchovies, Fig. 3-2, while the number of prey ingested varied according to the size of individual prey. That is, an inverse relationship existed between the number of prey ingested and prey size with fewer prey being ingested when prey were larger, Fig.3-3. Also, the interval between feedings for predators grouped according to the number of prey items found in the stomach indicates that the feeding intesvals varied with prey number, Fig.3-4. In addition Fig.3-5 the size of the last ration was estimated from the relative stomach contents just prior to (f) and the amounts of food ingested during the last feeding bout (F). Combined with data on the physiological maximum of these species, it was found that the Lizard fish, while posessing the capacity to feed at levels up to 40% of its body weight, does not normally do so in the field, Fig.3-1, and once satiated, subsequent prey were not ingested until the stomach contents are reduced by digestion to roughly half of the satiation level. Thus, there appears to be a cessation of feeding by fishes in the field when ecological satiation is achieved.

4) On the Probability of Prey Ingestion in the Natural Environment : In the course of time after the ingestion of a prey item the stomach contents change in the following manner : P (Ingested prey identifable) - A (Amorphous contents) - E (Empty stornach), Fig. 4-1. From this, one schematical model of feeding activities at the group level, using P, A and E as criteria, was devised, Fig. 4-3A. From intital analysis based on the premise that prey are distributed at random in space, possible compositions of P, A and E were found located on the lines defined by the points a-b-c-d-e, Fig. 4-3B. These points correspond to those of Fig.4-3A. Looking at the data from serveral cruises, one can see that this assumption is very seldom supported by actual field data, Fig.4-2. Subsequently, the more probable assumption that prey distribute in patches in spaces implies a negitive binomimal distribution: (q-p)¯-k After simulations using variable parameters of m and P, smaller k values, which imply that prey are distributed with increasing patchiness, are located nearer to E between the A-E axis for any level of P. The level of P in turn, implies the abundance of prey resoures, Fig.4-4, Fig.4-5. Thus, we can read from the informations concerning P, A and E that the chances of ingestion of prey by any group of predators in the field comes not only from the abundance of prey resources but also from the usually patchy spatial distribution of the prey.

5) On the Cost of Prey Ingestion in the Field

5-1) Quantity of ingested prey and the occurrence of empty stomachs : The quantity of prey ingested by predator groups fluctuated widely, Fig.5-1A, while there was a clearer correlation between average prey size and predator size, Fig.5-1B. Also, for predator groups, there was an inverse relationship between the size and number of prey ingested, Fig.5-1C. No significant change in the average prey size in relation to the per cent of predators with empty or near empty stomaches (%E') was found, Fig.5-2B. However, a clear inverse relationship exists between the quantity of ingested prey and %E', Fig.5-2A. As the occurrence of empty stomachs indicates the failure of predators to ingest prey, at least within the previous 24hrs., these results indicate that the chance of encounting prey strongly influences the amount of food ingested by predators at the group level, i.e. the gains. Thus, the predator must often expend some period of time before it encounters ally prey item.

5-2) Cost of handling prey by taxa and size : Lizard fish prey ordinarily on several prey taxa over a wide range of prey sizes(w/W), as shown in Fig.5-3 (for instance Cruise 5). The main prey taxa found year round in the diet of Lizard fish differ by number and weight, Fig.5-4A; in profile, Fig.5-4B ; and in relative size, 1/L, w/W, Fig. 5-4C. Although short-stout prey Fig.5-4B(1)-(6) appear more beneficial by weight, they actually occupy less than 30% of the diet (Fig.5-4A). On the contrary, slender prey, including anchovies which were dominate, occupy almost 50% of the diet. The posture in the stomach of ingested prey, i.e. the per cent of prey ingested from the head, differed in relation to the size (l/L) of the prey, Fig.5-5. These results may suggest that short-stout prey with spines anct larger prey require more of less greater "handling" efforts on the part of the predator. Exceptions would include Dragnets Fig.5-4B (6) which, having a distinct projectiou on the opercle, would be a more difficult prey to handle. The posture of the anchovy in the stomach, in relation to size, differed from other prey taxa, suggesting a difference in the "approaching" and "handling" of the pelagic school-forming prey as compared with other demersal prey taxa.

6) On thc Cost-Benefit Relationships of Food Intake for Predators in the Field: from "searching" to "ingestion"

6-1) The feeding behavior circuit : The feeding behavior of the Lizard fish is described by the following sequence : Searching - Handling - Digestion - Appetite - Searching, Fig.6-1. This circuit would close when the volume of ingested food is reduced to, or below, the ecological satiation level (1). Within this behavioral scheme, an abbreviated circuit becomes manifested when a hungry predator successively encounters small prey at a rate at which searching time becomes inconsequential (2).

6-2) Cost-Benefit Model - Its application to field data: The following model of the Cost/Benefit relationships is proposed, Fig.6-2. Cost = t_s + t_h where t_s = time spent searclzing, expressed by E'% = (E% + γA%), and t_h = handling expressed by the product of (prey size) and (no. of prey ingested). Benefit = r (the quantity of ingested prey). r values follow the curve defined by the equation, r. = R (1 - ℓ¯αt), having an upper asymptot at R. The efficiency of Benefit/Cost tan θ = r/(t_s + t_h) increases to a maximum, as shown by the tangent line in Fig.6-2, and then decreases as r approaches R.'23) Simulations of the model were carried out graphically using data of cruises from May to August, Fig.6-3. Various sets of the parameters R, r/R, t and α, derived from the simulations, are shown, R and r, Fig.6-4; r/R and t_s,Fig.6-5A and r/R and α, Fig.6-5B. As shown schematically in these three figures, r/R values are rather low when R (prey abundance) is high, Fig.6-4, and when t_s (searching cost) and/or α(1/α = the easy of handling of the prey) are small, Fig. 6-5A and Fig.6-5B, respectively. These results indicate that when prey resources are generally abundant, predators halt exploitation of a food resource at a relatively low level to begin searching for a new patch, as there is a great probability of finding one after a short search time. In contrast, under ordinary circumstances, i.e. daily conditions, Lizard fish behave as efficienct predators based on energy savings, though for such a sit-and-wait predator "searching costs (t_s)" may be considerable by time scale.

7) DISCUSION: ON THE FEEDING STRATEGIES OF PREDATORS

7-1) Cost-Benefit strategy in variable food resoures environments : Results from the analysis of Costs (searching and handling), Benefits (ingested food) and prey occurrence (total number of individuals for each prey taxa found in the stomachs of predators) are shown in Tab.7-1. The data have been grouped (A-F) by the parameters R, rR, t_s, α, Fig.7-1. For instant, general features within some of these groups include; Croup A: These predators ingested many prey indivicluals(∑n/P=2.0), dominated by unknownlarval fish(T.L.=20 - 30mm). These predators received low benefit (r), after relatively higher exploitation of a poor food resources environment (R = 3). In constrast, Group F predators ingested medium numbers of prey (∑n/P = 1.2 - 1.6), dominated by ancliovies. Benefit was good (r = 8-11 ) after relatively lower exploitation (r/R = 30 - 60%) of a plentiful food resources environment (R = 13 - 20). Further more, Group D predators ingested a medium number of prey (∑n/P =: 1.3 - 1.5), composed of many species of demersal fishes, such as gobies, codlets, sand launce, dragnets, and shrimps, etc. with a relatively lower occurrence of anchovies. Benefit was poor (r = 4 - 6), after relatively medium exploitation (r/R= 70 - 78%) of a poor resources environment. The alternation of feeding behavior in response to changes in the prey environment infers that Lizard fish can behave as if efficient Cost-Benefit decision makers71), that is, adopting the most cost-effective feeding beliavior in response to its immediate prey environment. This adaptability, in turn, naturally assunxes the ability to control its feeding beliavior aecording to its surroundings.

7-2) Satiation, physiological food storage capacity and discontinueus feeding in the field : Among fishes, the storage capacity of the digestive systems ranges widely, as for example from continuously feeding stomachless fishes, e.g. Goldfish, to voracious sit-and-wait feeders, e.g. Goose fish.8)37) Discontinous feeding behavior is often associated with the morphology and physiology of the digestive stern,20) and in addition, with the ability to ingest a wide range of prey sizes and types. Furthermore, there is a relationship between feeding strategies and prey resources, that is, in terms of the abundance and spatial distribution of the prey. In general, predators which eat large but less abundant prey often have large stomachs and vice versa. Thus, taken together, there appears to be an interrelatioiiship anlong the morphology of the digestive system, feeding strategies and prey resources .32)33) Feeding schools of Mackerels and Skipjack migrate over a wide area and, after searching, encounter prey patches at intervals. However, in general, they do not satiate to their physiological capacity but reform into schools and renew searching behavior after some sub-maximal degree of satiation. Thus, a trade-off between two separate behaviors, i.e. feeding and schooling, appears to be opperating.12)13) In this regard, changes in behavior within the feeding sequence of the pelagic, migratory predators can be analysed as if based on efficiency criteria. Generally speaking, fusiform shaped fishes, e,g. Anchovies, Mackerels, Skipjacks, Tuna, etc., including such divergent sit-and-wait fishes, such as Lizard fish and Goose fish, feed on large, beneficial prey whenever possible. However, when prey resources are poor, these fishes will also feed on small, less profitable prey. For Lizard fish, there is similarly, a change in the feeding behavior between sit-and-wait feeding on benthic prey and mobile searching behavior for pelagic prey based upon the prey resource environment.

7-3) Prey size, breadth of diet and feeding modes : The relative size of the prey ingested (w/W) ranges widely in the field and this has also been confirmed for fishes held in aquaria, Fig.7 - 2. 24)25) In addition, the range of prey taxa exploited in the field by any predator is, in general, also wide, except for those true specialist such as Plecoglossus altivelis. A further phenomenon found in nature is the altesation of the feeding mode of the predator, as is the case for the anchovy, whose feeding behavior changes from continuous filter feeding within minute plankton swarms to biting when some profitably exploitable quantity of large but ingestable prey is encountered. 41)42)43) Similarly behavioral diversity related to feeding is not as uncommon as generally conceived ,44)45)46)47) even for such apparently specialized chichlids as Peterotilapia. 48)50) Discontinuous feeding type predators, in general, have a substantial breadth of diet49), that is, they have developed the abilities to exploit a more or less wide range of prey taxa and prey size. This phenomenon is reflexed in the diversity of feeding modes exhibted by these predators depending on the characteristics of the prey. In this regard, choice of a suitable prey itern by taxa or size, connected with the choice of feeding mode, if necessary, can be viewed as if based on Cost-Benefit feeding principles.

7-4) "Persisting with" or "Shifting from" the prey environment : The short-term retainment of information pertaing to the prey environment can be expected to have some influence on whether a predatory fish decides to 'persist with' or 'shift from' a particular prey patch or location, i. e. "habitat". In the situation where the main prey items are spatially well-dispersed demersal prey, and thus searching costs become a sizeable component of total costs.the predator would naturally avoid lengthy migrations and try to exploit this poor prey environment as thoroughly as possible;intially attacking nearby/easily handled prey and progressively increase searching time and/or attach less easily handled prey. Thus, the efficiency of feeding activities in such a situation would gradually decrease. On the contrary, where patches of super-abundant pelagic prey occur, a predator would be expected to exploit at first weaker, smaller and/or easily captured prey around the periphery of thc patch. Over time, the efficiency of feeding at a given patch would gradually decrease as these preferred prey items diminish. At some point, the predator would begin searching for a new prey patch as the superabundance of prey in the overall environment would influence the expectations of encountering such a patch within a relatively short time. As a consequence, prey resources which move and/or migrate would eventually be expected to induce similiar behavioral responses in the predator, in the latter case. Theoretically, a predator 'should persist with a given resource until the gain rate in that resource is reduced to become equal to the maximum mean rate obtainable from the total environment'. 23)77) In conclusion, tlie feeding behavior of the Lizard fish can be visualized as a set of decisions based on Cost-Benefit principles, i.e. persisting at or shifting to a new habitat in search of prey, depending on the prey environment.

7-5) Notes : In short, each predator has its own specific features concerning feeding. These include morphological and physiological adaptations, as well as, behavioral abilities to adapt to variable ecological circumstances. Both of these aspects are simultaneously critical to its survival. Thus, a study of feeding strategy, which means to clarify the basic principles of feeding behavior, should be designed to embrace both of these aspects concurrently. For an investigation of this type, well designed experiments in aquaira and/or field observations will be powerful weapons, if suggested by the previous analysis of actual field data. In particular, careful attention should be paid to the difficulties of designing such experiments which aim to clarify the processes envolved in "searching" for a discontinuous feeder and also, to those studies based on stomach content analysis.