Jap. J. Limnol. 41, 4, 185-202, 1980 . Geographical Variations in Body Size, Brood Size and Egg Size of a Freshwater Shrimp, Palaemon paucidens DE HAAN, with Some Discussion on Brood Habit*,** Machiko NISHINO Abstract The variations in body size, egg size and brood size of berried females among populations have not been known in Palaemon paucidens. These were examined among 28 local populations from various inland waters over its range of geographical distribution. Body size differs among the populations. Egg size is almost constant regardless of body size within a population, but differs among the popu- lations, and the largest from Lake Akan is about seven times as large as the smallest from Lake Biwa. Brood size increases with body size, however, relative brood size is almost similar within a population. Among the populations, mean brood size and mean relative brood size vary, but fall within a limited range, except for the mean relative brood size in Lake Biwa, which is exceptionally large. Relative brood weight is almost similar in a population, as well as mean relative brood weight among the populations. Thus, mean relative brood size is almost inversely proportional to mean egg size. A geographical trend is noticeable in the variations of these two values (i. e., larger mean egg size and smaller mean relative brood size in colder waters), except for the populations in Lake Biwa and two rivers of Boso. The small body size of the population in Lake Biwa can be related to its life history, in which a longer period is spent in colder conditions, and the high relative brood size can be regarded as its adaptive strategy of allocation. described by YOKOYA (1931) from a pond 1. Introduction in Tokyo, besides the apparently smaller Palaemon paucidens DE HAAN is a sizes of their eggs in comparison with that common freshwater shrimp in Japan, dis- from other localities. tributed widely from Yaku Is. to Hokkai- It has been already recognized by KUBO do, as well as in Sakhalin, Etorofu Is., (1942) that there are two groups of species and Korean Peninsula to the north (RATH- of the genus Leander (=Palaemon) in BUN, 1902; KUBO, 1942; SHOKITA, 1975). Japan and adjacent areas that differ in egg It inhabits various kinds of inland waters, size, and Palaemon paucidens belongs to the group of large egg size. KAMITA from small ponds to large deep lakes and from the mouth to the upper course of (1961) extends KUBO's finding over the rivers. During the course of the study on Japanese freshwater shrimps through his the life history of this species in Lake extensive survey, and indicates that the Biwa, I noticed that the newly hatched eggs of the present species are larger than zoeae released by the females from this those of any other palaemonid and atyid shrimps in Japan, although this place has population were much smaller than that been taken over by Macrobrachium shoki- * Contributions from the Otsu Hydrobiological Station, Kyoto University, No.264 (Foreign Language Series). tai (SHOKITA, 1973a). KAMITA regards •* This study was partly supported by the Science this attribute of Palaemon paucidens as Research Fund from the Ministry of Education, "Paleolimnology of Lake Biwa" . that which made itself possible to be dis- 186 Geographical Variations of a Freshwater Shrimp tributed further north. He also states that the number of eggs carried by a female is closely related to the size of the mother shrimp, but he denies the geographical variation in egg size or in egg number for Palaemon paucidens. These parts of his statement are, however, confusingly cited by SHOKITA (1975) and TATSUKAWA (1977) as that the females of northern populations tend to bear larger eggs than southern ones in Palaemon paucidens. On the con- trary, KAMITA, in the same paper, refers to the apparent variation in size of the eggs of Paratya compressa. Nevertheless, little attention has been paid to the varia- tions in the size and number of eggs within a population or among local populations Fig. 1. Map showing the localities of the mate- rials examined. The numbers of of freshwater shrimps and prawns. the localities are identical to those of Table In the present study, the results of ex- 1., as well as in other text-figures. No. amination of the size and number of the 27 is Syananuma pond and No. 28 is eggs carried by a female among different Lake Tibesan. Symbols mean oligo- trophic lakes (○), eutrophic or bog populations of Palaemon paucidens are lakes (◎), ponds (●), middle or upper presented and analysed. The geographical courses of rivers (＋), and river variations in these features and their mouth (-). meaning are discussed from the viewpoint of adaptive significance for sustaining the Ten eggs, which were apparently not population. artificially injured, shrunk or bursting on 2. Materials and Methods hatching, were removed from each female, and their lengths and widths were meas- Berried females were collected from vari- ured under a stereoscopic microscope to the ous inland waters in Japan to cover the nearest O.01 mm. The body length of each range of the geographical distribution of mother shrimp was also measured to the the species, as well as the range of its nearest O.1 mm (from the base of eye to habitat, during their breeding seasons from the tip of telson) , and the wet weight with 1974 to 1978. All specimens caught were and without eggs were obtained to the preserved in 10 % formalin, except some nearest O.001 g. The total weight of eggs of those from a pond in Aomori Prefecture, carried by a shrimp was calculated from which were preserved in 70 % alcohol. the latter two values. At the same time, The specimens collected by Prof. D. the number of eggs carried by a female MIYADI in Lake Tibesan of Sakhalin and was counted. Measurements were made of Syananuma pond of Etorofu Is. in 1934, all of the above characters for all berried and retained at the Otsu Hydrobiological females when the specimens were rela- Station, were also examined. The localities tively few, or for 30-40 individuals; body and dates of collection of the specimens length, however, was measured on all examined are given in Fig. 1 and Table 1. berried females only when a large number NISHINO 187 of specimens were available. Only the body (No. 22), in which the distribution is bi- length and number of eggs can be consid- modal and the larger portion consists of 2/3 ered reliable in the following treatments of the females. The berried females from for specimens from Lake Tibesan and Hokkaido, except for the small females in Syananuma pond. On the contrary, the Lake Ohnuma (No. 22*), are larger, and egg size is considered reliable for these those of Lake Biwa (No. 1), Mandai-ike localities (No. 2, 5, 6, 18, 20 and 23), since (No. 4), and a pond of Aomori (No. 16), very few specimens were available. The are significantly smaller than the others. measurements were mostly done from The body length of the berried females three months to a year after collection. mostly exceeds 3 cm. The smallest berried As is generally admitted, it was quite female recorded during the present study difficult to properly remove excess water (2.42 cm) is from Lake Biwa (No. 1) ; it from the individual berried shrimps to is much smaller than the 3.1 cm reported insure equal wetness at weighing. Conse- by KAMITA from a moat in Nara or the quently, it was almost meaningless to 2.9 cm cited by KAMBARA et al. (1967) weigh them with an accuracy greater than from a pond in Kagawa. O.001 g, and inevitably the calculated A definite relationship exists between weight of an egg does not suffice for criti- weight including eggs and length on log cal comparison. On the contrary, the scales in each population, as is indicated length and width can be measured with by the highly significant correlation coef- greater precision for each egg. Thus, the ficient (P<0.01), the slope being nearly calculated egg size permits closer analysis, three in most of the populations, indicating and at the same time enables clarification similar body shape within each population of the variations in egg size with the same (Fig. 4). Covariance analysis shows that mother shrimp. Therefore, the calculated the regression equations are not signifi- egg size (i. e., [length (mm)] x [width cantly different in slope between any paired (mm)]2, and called egg volume for conven- populations whose regression lines are toler- ience' sake) is adopted in the following able for covariance analysis. On the con- discussion on egg size. trary, they differ significantly in elevation 3. Results (P<0.05) among most of these popula- tions, suggesting the difference in body 3-1. Body Size of Berried Female shape that has been noticed and is easily The size of the berried female varies discernible by the layman's eye. with the population in terms of different 3-2. Egg Size localities (Figs. 2, 4, 5). Although the ber- As has been pointed out by KAMITA ried females are larger early in the breeding (1961), the sizes of eggs in the advanced season (SAITO and YAMADA, 1954; OKUBO, developmental stages are larger than those 1964; KAMBARA et al., 1967), most of the of the earlier stages. The average volumes specimens examined in the present study of eyed eggs, and similarly their average were collected in the peak period of the weights, are 1.2 to 1.8 times as large as breeding season, namely, from late May those of pre-eyed eggs in all of the popu- to July, and the seasonal variation in size lations studied.