Larval Performance of Amphidromous and Landlocked Atyid Shrimp Species in the Genus Paratya under Different Feeding Conditions

© 2020 Academia Sinica, Taiwan Open Access Larval Performance of Amphidromous and Landlocked Atyid Shrimp Species in the Genus Paratya under Different Feeding Conditions Katsuyuki Hamasaki*, Masakazu Okada, Sota Nishimoto, and Shigeki Dan Department of Marine Biosciences, Tokyo University of Marine Science and Technology, Konan, Minato, Tokyo 108-8477, Japan. *Correspondence: E-mail: [email protected] (Hamasaki) E-mail: hazike1[email protected] (Okada); [email protected] (Nishimoto); [email protected] (Dan) Received 13 October 2020 / Accepted 1 December 2020 / Published 15 December 2020 Communicated by Benny K.K. Chan Freshwater shrimps in the family Atyidae exhibit two life history traits: 1) amphidromy, with many small embryos hatching into planktotrophic larvae that develop in the sea, and 2) landlocked, with few large embryos hatching into non-feeding lecithotrophic larvae, or facultative lecithotrophic larvae that develop in freshwater. The lecithotrophy of larvae is considered an adaptation to limited food conditions in the freshwater environment. Furthermore, faster development and settlement behaviours that large larvae exhibit are considered adaptations that allow the larvae to stay in or near parental habitats in fast-flowing streams. We therefore hypothesized that the facultative lecithotrophic larvae of landlocked shrimps might better adapt to limited food conditions as their large body size develops, an adaptation to maintain habitat position in flowing streams, than do planktotrophic larvae of companion amphidromous species developing in the sea. To test this hypothesis, we compared the larval feeding habits and size of two closely related species in the genus Paratya: the amphidromous P. compressa, with planktotrophic larvae, and the landlocked P. improvisa, with facultative lecithotrophic larvae. Larvae were reared by being fed commercially preserved or cultured phytoplankton (Tetraselmis sp.) and cultured zooplankton rotifers. Paratya compressa larvae did not survive, but P. improvisa larvae did develop into the juvenile stage under poor feeding conditions with preserved Tetraselmis alone, supporting our hypothesis of the effects of different larval feeding habits in amphidromous and landlocked atyid shrimp species. Hatchlings were larger and larval duration was shorter in P. improvisa than in P. compressa. Paratya improvisa larvae exhibited settlement behaviour beginning with the early zoeal stage. Our results also highlighted the retention strategy by which landlocked P. improvisa larvae stay in or near parental habitats. Key words: Larval development, Larval survival, Feeding habit, Lecithotrophy, Retention strategy. Citation: Hamasaki K, Okada M, Nishimoto S, Dan S. 2020. Larval performance of amphidromous and landlocked atyid shrimp species in the genus Paratya under different feeding conditions. Zool Stud 59:70. doi:10.6620/ZS.2020.59-70. BACKGROUND Freshwater shrimps of the family Atyidae are distributed worldwide, except in Antarctica. They are a highly species-rich group of decapod crustaceans (Decapoda: Caridea) that inhabit various freshwater bodies from torrential mountain streams down to sluggish, oligohaline waters and rivers or streams inside karst caves (De Grave et al. 2008; De Grave and Fransen 2011; Cai and Ng 2018). Atyid shrimps play important roles in stream food webs and ecosystems, functioning as primary consumers and food sources for predators (Pringle et al. 1993; Covich and McDowell 1996; Covich et al. 1999; Crowl et al. 2001; Oeding et al. 2020). Atyid shrimps exhibit one of two life history traits: amphidromous or landlocked (Shokita 1979; Hayashi and Hamano 1984; Bauer 2013). Amphidromous adults Zoological Studies 59:70 (2020) doi:10.6620/ZS.2020.59-70 1 © 2020 Academia Sinica, Taiwan inhabit and reproduce in freshwater environments, but their larvae require saline waters for successful development (Shokita 1979; Hayashi and Hamano 1984; Nakahara et al. 2005). Amphidromous species spawn many relatively small eggs (Shokita 1979; Hayashi and Hamano 1984; Jalihal et al. 1993; Bauer 2013), and newly hatched larvae (stage 1 zoeae) passively drift from the freshwater environments into the sea (Ideguchi et al. 2000 2007; Hamano et al. 2005). Larvae develop through longer periods of complex zoeal stages in the brackish water of estuaries and coastal bays or in the open sea (Hayashi and Hamamo 1984; Jalihal et al. 1993; Ideguchi et al. 2007; Yatsuya et al. 2013). After recruitment to the mouth of a coastal river or stream, juveniles migrate up into the adult freshwater habitat (Hamano and Hayashi 1992; Hamano et al. 2005; Bauer 2013; Yatsuya et al. 2013). On the contrary, landlocked species complete their entire lifecycle within freshwater habitats and exhibit intraand interspecific variabilities in their reproductive traits (Shokita 1979; Hancock 1998; Hancock et al. 1998; Bauer 2013). In a few landlocked species with small eggs, extended larval development—as in the case of amphidromous shrimps—occurs in freshwater habitats like large lakes (Bauer 2013). In general, however, landlocked shrimps spawn a relatively small number of large eggs with considerable amounts of yolk. These produce morphologically advanced hatchlings, resulting in a shortened larval development period with few successive zoeal stages (abbreviated larval development) before moulting into the juvenile stage (sensu Møller et al. 2020), or directly producing juveniles (direct larval development) (Shokita 1979; Hayashi and Hamano 1984; Jalihal et al. 1993; Bauer 2013). Amphidromy is considered a plesiomorphic (ancestral) life history trait in freshwater shrimps (Shokita 1979), and current evidence from physiological, developmental, and phylogenetic considerations supports the hypothesis of an evolutionary process from marine ancestral species through freshwater amphidromous to landlocked species (Bauer 2013). It has been suggested that insufficient, unpredictable, or seasonally short production of food for larvae in freshwater environments (Shokita 1979; Anger 2001 2006; Bauer 2013) are the major driving forces behind the landlocked life history trait associated with few and large eggs. Larger embryos produce non-feeding lecithotrophic larvae, which thoroughly rely on yolk left over from the embryo, or facultative lecithotrophic larvae, which rely on endogenous yolk and exogenous feeding (Jalihal et al. 1993; Anger 2001; Kawamura and Akiyama 2010). Besides larval nutritional conditions, faster development and settlement behaviours that large larvae exhibit are considered adaptations that allow the larvae to stay in or near parental habitats in fastflowing streams (Shokita 1979; Mashiko 1982 1990; Dudgeon 1985; Wong 1989; Hancock 1998; Hancock et al. 1998). We therefore hypothesized that the facultative lecithotrophic larvae of landlocked shrimps adapt better to limited food conditions while developing their large body size—an adaptation to maintain habitat position in flowing streams—than do planktotrophic larvae of companion amphidromous species developing in the sea. However, to the best of our knowledge, no studies have compared the larval feeding habits and size between closely related amphidromous and landlocked shrimp species. In the present study, we tested the hypothesis that feeding habits and size affect the larval performance of two species in the genus Paratya Miers 1882: the amphidromous P. compressa (De Haan 1844) and the landlocked P. improvisa Kemp 1917. We conducted larval culture experiments under different feeding conditions. Our results highlight the different larval feeding habits and sizes, as well as the adaptive value of these traits in two related species, one with amphidromous and the other with landlocked atyid larvae. MATERIALS AND METHODS Test species Paratya compressa and P. improvisa are closely related species of atyid shrimps (Ikeda 1999; Page et al. 2005). Paratya compressa is an amphidromous species with many small eggs (Kawamura and Akiyama 2010). It inhabits lower and middle reaches with boulders in south-western Japan (Shokita 1979; Hamano and Hayashi 1992; Suzuki et al. 1993; Ikeda 1999; Usami et al. 2008; Saito et al. 2012). Paratya improvisa is landlocked with few, relatively large eggs (Kawamura and Akiyama 2010). It inhabits lentic systems of the lower to upper reaches in north-eastern Japan (Ikeda 1999; Imai 2006; Yatsuya et al. 2012). Kawamura and Akiyama (2010) reported that both P. compressa and P. improvisa larvae undergo eight zoeal stages before moulting into the juvenile stage. Paratya compressa larvae are shown to be planktotrophic, whereas P. improvisa larvae can moult to stage 3 zoeae without food (i.e., facultative lecithotrophy) and have a shorter larval developmental duration than do P. compressa larvae. page 2 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan Larval source Culture experiments were conducted in a laboratory at the Tokyo University of Marine Science and Technology, Tokyo, Japan in 2017 and 2018. Wild mature female and male test shrimps were captured from April to May using scoop nets at the following locations in Chiba Prefecture, Japan: P. compressa, Banda River (34°58'N, 139°46'E); and P. improvisa, Yoro River (35°21'N, 140°08'E). Females were cultured with males in aerated aquaria (23 litres) under natural photoperiod conditions (20–30 shrimps per tank, unknown sex ratio). Aquarium temperature was maintained at approximately 23°C using a heater and an air conditioner in the laboratory. This falls within the range of summer temperatures during the reproductive season of atyid shrimps in their natural habitat (Imai 2006; Saito et al. 2012; Yatsuya et al. 2012). Artificial diets for ornamental shrimps or fish were provided daily. Half of the rearing water was replaced with fresh water every three days. Ovigerous females were transferred and individually contained in beakers (1 litre). Ovigerous females were fed an artificial diet each day. 200 ml of rearing water was exchanged for fresh water daily. Newly hatched larvae (stage 1 zoeae) from two brood females of each species were used for the larval culture experiments. The larval hatching date of each species-brood was as follows: P. compressa brood 1, August 31, 2017; P. compressa brood 2, May 12, 2018; P. improvisa brood 1, July 14, 2017; and P. improvisa brood 2, July 23, 2017. The stage 1 zoeae from each brood were sampled, fixed with 5% neutral formalin for one day, and then preserved in 70% ethanol. The carapace length of 10 specimens of each species-brood was measured from the posterior margin of sessile eyes to the posterior end of the carapace (Nakahara et al. 2007) using a microscope equipped with a digital camera and image analysing system (Nikon Digital Sight and NIS-Elements software, Nikon Corp., Tokyo, Japan). Feeding conditions We previously conducted experiments to examine the dietary effects of euryhaline phytoplankton Tetraselmis sp. (cell diameter, ~10–12 μm) and euryhaline zooplankton, rotifers in the Brachionus plicatilis species complex (small-morphotype; body size, ~0.1–0.2 mm in lorica length) on the survival, duration, and growth of larvae of four amphidromous atyid shrimp species in the genus Caridina H. MilneEdwards 1837. These were C. leucosticta Stimpson 1860, C. multidentata Stimpson 1860, C. serratirostris De Man 1892, and C. typus H. Milne-Edwards 1837 (Hamasaki et al. 2020). In these experiments, we used a commercial, condensed paste of Tetraselmis sp. stored at 4°C (preserved Tetraselmis) (Tetraselmis 3600®, Reed Mariculture Inc., Campbell, CA, USA) and live Tetraselmis sp. cultured in our laboratory (cultured Tetraselmis). Tetraselmis sp. is a green unicellular alga with four distinct flagella. Algal cells of cultured Tetraselmis can move around freely in seawater, whereas those of preserved Tetraselmis were intact but nonviable. In the present study, larvae of Paratya species were reared under five different feeding conditions with phytoplankton, rotifers, and their combinations as follows: 1) preserved Tetraselmis (PT), 2) cultured Tetraselmis (CT), 3) rotifers (R), 4) preserved Tetraselmis and rotifers (PTR), and 5) cultured Tetraselmis and rotifers (CTR). Our previous experiments revealed that cultured Tetraselmis and rotifers successfully sustain the larval development into the juvenile stage in Caridina species under CT and CTR conditions, but the nutritional conditions of their larvae were considered poor when they were fed preserved Tetraselmis and/or rotifers under the PT, R, and PTR conditions (Hamasaki et al. 2020). Tetraselmis sp. were cultured in glass containers with seawater (salinity, 34 ppt; 24–26°C) supplemented with fertilizers for algal culture (KW21, Daiichi Seimo Co. Ltd., Kumamoto, Japan) under constant illumination. The rotifers were cultured in plastic containers with seawater (salinity, 24 ppt; 23–24°C) supplied with commercial, condensed freshwater phytoplankton, Chlorella vulgaris (Super Chlorella V12, Chlorella Industry, Tokyo, Japan). Larval culture Thirty newly hatched larvae (stage 1 zoeae) were used for each feeding condition in each species brood. Larvae were housed individually in the wells of sixwell cell culture plates, each well of which contained 8 ml of artificial saline water. The feeding density and larval rearing environments followed our previous study of cultured larvae of Caridina species (Hamasaki et al. 2020): Tetraselmis sp., 1 × 105 cells ml-1; rotifers, 20 individuals ml-1; salinity, 17 ppt; and temperature, 26°C. Although P. improvisa is a landlocked shrimp, its larvae can develop in saline water and live longer in saline water of 8.5 ppt and 17 ppt than in water of other salinity levels (0, 25.5, and 34 ppt) under starvation conditions (Kawamura and Akiyama 2010). Larval rearing temperature (26°C) is in the range of the summer sea surface temperatures along the Pacific coasts of Japan (Japan Meteorological Agency 2020) or in a natural pond (Imai 2006), where larvae of P. page 3 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan compressa and P. improvisa may develop, respectively. The salinity of larval rearing water was regulated using artificial seawater salt (Sealife, Marinetech Co. Ltd., Tokyo, Japan). Larval culture plates were set in temperature-controlled incubation chambers under a photoperiod cycle of 14 h light and 10 h dark. Each morning, glass pipettes were used to transfer larvae to clean culture wells containing fresh saline water and food, and the number of live and dead larvae were recorded under a stereomicroscope. Each larva was also checked for the presence of an exuvia. Larval rearing was terminated when all surviving larvae had moulted into the juvenile stage. We observed later stage larvae under a stereomicroscope and determined whether they moulted into the juvenile stage based on the morphologies of their first and second antennae, as described by Kawamura and Akiyama (2010). The final survival rate of larvae was defined for each feeding condition in each species-brood as: (number of larvae that moulted into the juvenile stage)/(number of initial larvae) × 100. Juvenile survivors of each species-brood were fixed and preserved in the same way as stage 1 zoea specimens, and the carapace length was measured from the posterior margin of the orbit to the posterior end of the carapace (Nakahara et al. 2007). Statistical analysis Statistical analyses were performed using R statistical software (R4.0.2; R Core Team 2020) with a 5% significance level. Survival curves were compared among different feeding conditions in each species with a log-rank test stratified by the brood identity using the survdiff function implemented in the survival package (Therneau 2020). The effects of different feeding conditions (categorical explanatory variable) on the number of days required to moult into the juvenile stage (i.e., larval duration) (response variable) was evaluated in each species using a generalized linear mixed-effects model (GLMM) with a Poisson distribution, taking into account the inter-brood variability (Zuur et al. 2009). A linear mixed-effects model (LMM) was also employed to compare the larval growth (i.e., the carapace length of juveniles) (response variable) among different feeding conditions (categorical explanatory variable) in each species. In these analyses, brood identity was included as the random intercept effect. The GLMM and LMM analyses were performed using the glmer function (log link) and lmer function (identical link) in the lme4 package (Bate et al. 2015), respectively. The statistical significance of the explanatory variable was evaluated with the type II Wald chi-square test and Type II Wald F test with Kenward-Roger d.f. in the GLMM and LMM analyses, respectively, using the Anova function in the car package (Fox and Weisberg 2011). Differences in larval duration and carapace length between feeding conditions in each species were tested with Tukey’s method using the glht function in the multcomp package (Hothorn et al. 2008). The difference between the numbers of moults during the larval stage among different feeding conditions was not statistically tested because some exuviae of larvae appeared to be missed during the culture operations (see the RESULTS section). RESULTS The carapace length of stage 1 zoeae (mean ± standard deviation) of P. compressa broods 1 and 2 was 0.277 ± 0.030 mm and 0.321 ± 0.030 mm, respectively. In P. compressa, we observed a significant difference in larval survival to the juvenile stage between different feeding conditions (χ2 = 185.6, d.f. = 4, P < 0.0001) (Fig. 1A, B). When larvae were supplied with preserved Tetraselmis alone, brood 1 larvae never moulted and 23.3% of brood 2 larvae moulted once. All larvae from both broods died within 16 days of hatching (Fig. 1A, B; see supplementary Table S1 for survival and moulting data on larvae that had died prior to moulting into the juvenile stage). Larvae survived to the juvenile stage when they were fed cultured Tetraselmis and/or rotifers, and preserved Tetraselmis with rotifers under the CT, R, PTR, and CTR conditions. Larval survival rates began to vary among feeding conditions between 6 and 13 days after hatching (Fig. 1A, B), and the final survival rate was higher in the CTR group (90.0%) than in other groups (40.0–60.0%) in brood 1. In brood 2, the final survival rate was higher in the CT and CTR groups (60.0–66.7%) than in other groups (3.3–10%) (Fig. 2A; see supplementary Table S2 for survival and development data on larvae that survived to the juvenile stage). Feeding conditions significantly influenced the larval phase duration (Fig. 2B) (χ2 = 104.4, d.f. = 3, P < 0.0001) and carapace length of juveniles (Fig. 2C) (F3, 110.5 = 3.70, P = 0.0140). Larval development and growth tended to be enhanced in the CT and CTR groups. The carapace length of stage 1 zoeae of P. improvisa broods 1 and 2 was 0.626 ± 0.050 mm and 0.603 ± 0.023 mm, respectively. Larvae of P. improvisa tended to stay on the bottom of the rearing container beginning with the early zoeal stage. In P. improvisa, survival curves were significantly different between feeding conditions (χ2 = 134.8, d.f. = 4, P < 0.0001) (Fig. 1C, D). Larvae survived to the juvenile stage in all feeding groups except for brood 2 larvae in the PT page 4 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan group, which moulted 1–4 times but died within 13 days of hatching (Fig. 1C, D; Table S1). Larval survival was substantially improved in the CT, PTR, and CTR groups, as shown by the higher final survival rates (93.3–100%) compared to other groups (16.7–53.3%) (Fig. 2A; Table S2). Significant differences were found in larval duration (Fig. 2B) (χ2 = 95.5, d.f. = 3, P < 0.0001) and the carapace length of juveniles (Fig. 2C) (F4, 196.4 = 152.6, P < 0.0001) for different feeding conditions. Larval development and growth tended to be improved in groups with higher survival rates, particularly the CT and CTR groups. The number of times that larvae moulted before reaching the juvenile stage varied from 3–13 in P. compressa and 2–6 in P. improvisa (Table S2), and the minimum moulting frequencies were lower than the number of zoeal stages (eight) reported for these species (Kawamura and Akiyama 2010), suggesting that some exuviae were missed during the culture operations. DISCUSSION Our larval culture experiments demonstrated that feeding conditions significantly affected larval survival, duration, and growth for the two Paratya species with different life history traits, namely amphidromous P. compressa and landlocked P. improvisa (Figs. 1 and 2). The larval development and growth tended to be enhanced under the feeding conditions that resulted in higher survival rates in both species. Although larval survival among the different feeding conditions varied between the two broods of P. compressa, overall, the final survival rates were high when larvae were fed with cultured Tetraselmis and cultured Tetraselmis with rotifers. Larvae did not survive to reach the juvenile stage when fed preserved Tetraselmis alone (Table S1). In P. improvisa, the final larval survival rates were high when larvae were fed cultured Tetraselmis, preserved Tetraselmis with rotifers, and cultured Tetraselmis with rotifers compared to those of larvae fed solely on rotifers. A small proportion of larvae (brood 1) survived and developed into the juvenile stage, even when they were fed preserved Tetraselmis alone. Thus, the survival and development of P. compressa and P. improvisa larvae differed among the various feeding conditions. In our larval culture experiments, early larvae of both P. compressa and P. improvisa appeared to ingest both preserved and cultured Tetraselmis because green microalgal substances were observed in the digestive tracts of the larvae under a stereomicroscope. Nevertheless, when larvae were fed preserved Tetraselmis alone, the only larvae that survived long enough to reach the juvenile stage was the brood 1 larvae of P. improvisa. We established that larval development and survival depend on diet. As such, it is interesting to note that different nutritional components may play key roles. For instance, the fatty acid composition and amount of fatty acid in live food significantly affect larval survival, growth, and development in aquatic Fig. 1. Changes in the survival rate (circle symbols) and cumulative moulting rate (triangle symbols) into the juvenile stage in relation to days after hatching of the larvae hatched from two different brooding females in two atyid shrimp species, Paratya compressa (A, B) and P. improvisa (C, D). Larvae were reared on five different diet combinations of phytoplankton (Tetraselmis sp.) and zooplankton rotifers: preserved Tetraselmis (PT), cultured Tetraselmis (CT), rotifers (R), preserved Tetraselmis with rotifers (PTR), and cultured Tetraselmis with rotifers (CTR). Larval rearing was terminated when all surviving larvae had moulted into the juvenile stage. 0 20 40 60 80 100 0 5 10 15 20 25 30 35 40 45 50 55 0 5 10 15 20 25 30 35 40 45 50 55 A) P. compressa (brood 1) 0 20 40 60 80 100 B) P. compressa (brood 2) Survival rate and moulting rate (%) PT CT R PTR CTR Days after hatching C) P. improvisa (brood 1) D) P. improvisa (brood 2) 0 20 40 60 80 100 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 0 20 40 60 80 100 page 5 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan organisms, including decapod crustaceans (Montaini et al. 1995; Anger 2001; Tocher 2010). The preserved Tetraselmis suecica has been shown to maintain its fatty acid profile when stored at 4°C (Montaini et al. 1995). Feeding behaviour is another factor that may affect larval nutritional intake. Atyid shrimp larvae ingest phytoplankton via suspension feeding (Nakahara et al. 2007). Consequently, the voluntary movement of live Fig. 2. Survival rate from hatching to the juvenile stage (A), number of days required to moult into the juvenile stage (larval duration) (B), and carapace length of juveniles (C) in the larvae of two atyid shrimp species, Paratya compressa and P. improvisa. Larvae were reared on five different diet combinations of phytoplankton (Tetraselmis sp.) and zooplankton rotifers: preserved Tetraselmis (PT), cultured Tetraselmis (CT), rotifers (R), preserved Tetraselmis with rotifers (PTR), and cultured Tetraselmis with rotifers (CTR). Larval rearing was terminated when all surviving larvae had moulted into the juvenile stage. Bars and vertical lines in panels (B) and (C) indicate mean and standard deviation values, respectively. Numbers of individuals are shown above the bars for each species brood in panel (B). Differences in larval duration and carapace length between feeding conditions in each species (P < 0.05) are indicated by different lowercase letters in the tables following the feeding conditions in panels (B) and (C), respectively. 0 10 20 30 40 50 1 2 1 2 Larval duration (days) B) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1 2 1 2 Carapace length (mm) C) P. compressa P. improvisa P. compressa P. improvisa Species and brood number 111 2 11 3 2 2 1 2 5 33 5 2 3 0 20 40 60 80 100 1 2 1 2 Survival rate (%) A) PT CT R PTR CTR P. compressa P. improvisa CT R PTR CTR b b b a CT R PTR CTR b ab a b PT R PTR CTR a c b a c PT R PTR CTR c a b d ab CT CT page 6 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan algal cells of cultured Tetraselmis might enhance the efficiency of suspension feeding, whereas the nonviable cells of preserved Tetraselmis might substantially lower the larval suspension feeding efficiency, resulting in larval nutritional deficiency. Kawamura and Akiyama (2010) examined the starvation tolerance of newly hatched larvae of P. compressa and P. improvisa under different temperature (19, 21, 23, 25, 27, and 29°C) and salinity (0, 8.5, 17.0, 25.5, and 34.0 ppt) combinations. They reported that, under starvation conditions at 25–27°C and salinity of 17 ppt, which are equivalent to the larval rearing conditions in the present study, mean survival time was ~7–9 days in P. compressa and ~6 days in P. improvisa. They also showed that P. compressa larvae never moulted, whereas 100% and 84–92% of P. improvisa larvae moulted once and twice, respectively. In our larval culture experiments, under the feeding condition with preserved Tetraselmis alone, the mean survival time of larvae that died was ~9–11 days in P. compressa and ~11–13 days in P. improvisa. In P. compressa, brood 1 larvae never moulted, and a small proportion of brood 2 larvae moulted once. Paratya improvisa larvae, however, moulted 1–3 times (mean, 2.0 times) in brood 1 and 1–4 times (mean, 2.0 times) in brood 2 (Table S1). Consequently, the relative survival time of fed larvae to that of unfed larvae was more prolonged in P. improvisa than in P. compressa under the limited nutritional condition with preserved Tetraselmis alone. This is likely due to the facultative lecithotrophy of P. improvisa larvae. Larval survival rates of P. compressa linearly decreased from about 6–11 days after hatching and larvae did not survive to the juvenile stage under feeding with preserved Tetraselmis alone. Supplementing rotifers with preserved Tetraselmis, however, sustained larval survival and development into the juvenile stage (Fig. 1). Larval survival curves were similar under the feeding conditions with rotifers alone and with preserved Tetraselmis and rotifers, particularly in brood 2 (Fig. 1). These results suggest that P. compressa larvae might be able to ingest rotifers from an early stage. Additionally, feeding cultured Tetraselmis with rotifers improved larval survival and developmental rate, particularly from about 13 days after hatching (brood 1), compared to feeding cultured Tetraselmis alone (Fig. 1). Our results suggest that ingesting larger rotifers (~0.1– 0.2 mm) coupled with cultured Tetraselmis (~10–12 μm) might enhance the larval feeding efficiency and thereby improve the nutritional condition of larvae from the middle stage, resulting in a higher survival rate and faster development of larvae (Fig. 2). Given that feeding rotifers alone sustained larval survival into the juvenile stage but preserved Tetraselmis alone resulted in a decrease in survival from about 6 days after hatching (Fig. 1), we deduced that P. improvisa larvae are likely able to digest rotifers from an early stage. The dietary effects of preserved Tetraselmis and rotifers on larval performance were significantly improved when they were given to the larvae in combination rather than alone (Fig. 2). Atyid shrimp larvae can ingest detritus (Hayashi and Hamano 1984; Hancock 1998) and rotifers can be cultured and reproduced when fed Tetraselmis sp. (Hagiwara et al. 2007). Detritus (i.e., faeces) generated by rotifers that were fed preserved Tetraselmis may make a better food source for P. improvisa larvae. Taken together, this suggests that P. improvisa larvae could survive into the juvenile stage under the poor feeding conditions limited to preserved Tetraselmis and detritus derived from rotifers fed on preserved Tetraselmis. This supports our hypothesis that facultative lecithotrophic larvae of landlocked shrimps might better adapt to limited food conditions than equivalent larvae of companion amphidromous species. Hamasaki et al. (2020) conducted larval culture experiments to determine the larval ecology of four amphidromous atyid shrimps, C. leucosticta, C. multidentata, C. serratirostris, and C. typus. They reared larvae under the same five feeding conditions employed in the present study. Larvae of C. serratirostris died because they were trapped by the surface tension of the water during moulting. In the three remaining species, feeding with preserved Tetraselmis, rotifers, and a combination of the two did not sustain the larval survival or development into the juvenile stage. Larvae could, however, survive to become juveniles when they were fed cultured Tetraselmis alone and cultured Tetraselmis and rotifers. Hamasaki et al. (2020) suggested that C. leucosticta larvae may have adapted to develop in eutrophic brackish environments with substantial phytoplankton biomass because they moulted into juveniles with high survival rates by being fed only cultured Tetraselmis. Caridina species thus behaved differently from the P. compressa larvae used in this study, as evidenced by the P. compressa larvae’s ability to develop into the juvenile stage under the poor condition of being fed rotifers alone and being fed preserved Tetraselmis and rotifers (Fig. 1). Paratya compressa larvae are able to develop in oligotrophic environments in coastal and offshore sea, thus enabling wide sea dispersion. Phylogeographic studies are required to determine the extent of P. compressa larvae’s aquatic distribution. Direct larval development or abbreviated larval development into the juvenile stage of landlocked freshwater shrimps have been considered adaptations to stay in or near parental habitats in fast-flowing page 7 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan streams (Shokita 1979; Mashiko 1982 1990; Dudgeon 1985; Wong 1989). Morphologically advanced larvae may have the ability to grasp trailing vegetation and roots and thereby avoid being swept away by fastflowing water (Dudgeon 1985). Hancock (1998) examined the relationship between egg size and larval development of Paratya australiensis Kemp 1917 collected from headwater streams of the Conondale Rage, Queensland, Australia. Shrimps with large eggs and small eggs were obtained from upper and lower altitude sites, respectively. He reported that larval size depends on egg size. Large larvae from large eggs developed faster than did those from small eggs, but large and small larvae had similar larval developmental pathways into the juvenile stage. Hancock (1998) and Hancock et al. (1998) suggested that the production of larger, more rapidly developing larvae at upper altitude sites may reduce the chance of larvae being swept into downstream sites, from where it would be difficult to repopulate the source population. The body size (carapace length) of stage 1 zoeae was larger in P. improvisa (mean, 0.603–0.626 mm) than in P. compressa (mean, 0.277–0.321 mm), but juvenile body size was similar in P. improvisa (mean, 0.847–1.112 mm) and P. compressa (mean, 0.868–1.048 mm), and larval duration was shorter in P. improvisa (mean, 8–18 days) than in P. compressa (mean, 21–40 days) (Fig. 2; Table S2). Larval developmental stages are not shortened in P. improvisa because they have eight zoeal stages, similar to amphidromous P. compressa (Kawamura and Akiyama 2010). Larvae of P. improvisa tended to exhibit settlement behaviour in the rearing container beginning with the early zoeal stage. Therefore, faster development and behavioural characteristics of P. improvisa larvae are considered to be adaptations that increase the chance of maintaining the larvae’s position in or near parental habitats in lower to upper reaches of various lentic systems (Ikeda 1999; Imai 2006; Yatsuya et al. 2012). CONCLUSIONS Our larval culture experiments support the hypothesis that facultative lecithotrophic larvae of landlocked shrimps might adapt better to limited food conditions than do planktotrophic larvae of companion amphidromous species. Facultative lecithotrophic larvae of landlocked P. improvisa could survive and develop into the juvenile stage under poor feeding conditions that are not suitable for the survival and development of planktotrophic larvae of amphidromous P. compressa. The present study thus provides empirical data that suggest that limited food conditions are a major selective force influencing landlocked life history traits with few and large eggs (i.e., larvae). Such limitations include insufficient, unpredictable, or seasonally short production of food in freshwater habitats (Shokita 1979; Anger 2001 2006; Bauer 2013). Our results and observations further highlight strategies that landlocked P. improvisa larvae use to stay in or near parental habitats—i.e., fast development and behavioural characteristics. Acknowledgments: We thank the members of our laboratory for helping with laboratory work. We are grateful to the anonymous reviewer and the editor for their valuable comments and suggestions, which have improved the manuscript. Thanks also to Noah Last of Third Draft Editing for his English language editing. This study was supported by JSPS KAKENHI grant number JP17K00578 to KH. Authors’ contributions: KH designed the study. MO and SN performed the laboratory experiments. KH and SD analysed the data and prepared the manuscript. All authors approved the final manuscript. Competing interests: The authors declare that they have no conflicts of interest. Availability of data and materials: All data are provided within the manuscript and supplementary materials. Consent for publication: Not applicable. Ethics approval consent to participate: The present study complied with current Japanese laws. REFERENCES Anger K. 2001. The biology of decapod crustacean larvae. Crustacean Issues 14. A,A. Balkema, Lisse, The Netherland. Anger K. 2006. Contribution of larval biology to crustacean research: a review. Invertebr Reprod Dev 49:175–205. doi:10.1080/07924 259.2006.9652207. Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting linear mixed-effects models using lme4. J Stat Softw 67(1):1–48. doi:10.18637/jss.v067.i01. Bauer RT. 2013. Amphidromy in shrimps: a life cycle between rivers and the sea. Lat Am J Aquat Res 41:633−650. doi:103856/vol41issue4-fulltext-2. Cai Y, Ng PKL. 2018. Freshwater shrimps from karst caves of southern China, with descriptions of seven new species and the identity of Typhlocaridina linyunensis Li and Luo, 2001 (Crustacea: Decapoda: Caridea). Zool Stud 57:27. doi:10.6620/ ZS.2018.57-27. Covich AP, McDowell WH. 1996. The stream community. In: Reagan DP, Waide RB (eds) The food web of a tropical rain forest. page 8 of 10Zoological Studies 59:70 (2020) © 2020 Academia Sinica, Taiwan University of Chicago Press, Chicago, IL, USA. Covich AP, Palmer MA, Crowl TA. 1999. The role of benthic invertebrate species in freshwater ecosystems. BioScience 49:119-127. doi:10.2307/1313537. Crowl TA, McDowell WH, Covich AP, Johnson SL. 2001. Freshwater shrimp effects on detrital processing and nutrients in a tropical headwater stream. Ecology 82:775–783. doi:10.1890/00129658(2001)082[0775:FSEODP]2.0.CO;2. De Grave S, Cai Y, Anker A. 2008. Global diversity of shrimps (Crustacea: Decapoda: Caridea) in freshwater. Hydrobiologia 595:287–293. doi:10.1007/s10750-007-9024-2. De Grave S, Fransen CHJM. 2011. Carideorum catalogus: the recent species of the dendrobranchiate, stenopodidean, procarididean and caridean shrimps (Crustacea: Decapoda). Zool Meded 85:195–589. Dudgeon D. 1985. The population dynamics of some freshwater carideans (Crustacea: Decapoda) in Hong Kong, with special reference to Neocaridina serrata (Atyidae). Hydrobiologia 120:141–149. doi:10.1007/BF00032135. Fox J, Weisberg S. 2011. An R companion to applied regression, 2nd edn. Sage Publications, Thousand Oaks, CA, USA. Hagiwara A, Suga K, Akazawa A, Kotani T, Sakakura Y. 2007. Development of rotifer strains with useful traits for rearing fish larvae. Aquaculture 268:44–52. doi:10.1016/ j.aquaculture.2007.04.029. Hamano T, Hayashi K. 1992. Ecology of an atyid shrimp Caridina japonica (De Man, 1892) migrating to upstream habitats in the Shiwagi Rivulet, Tokushima Prefecture. Crust Res 21:1−13. doi:10.18353/rcustacea.21.0_1. (in Japanese with English abstract) Hamano T, Ideguchi K, Nakata K. 2005. Larval drift and juvenile recruitment of amphidromous freshwater shrimps (Decapoda: Caridea) in the Nishida River, western Japan. Aquacult Sci 53:439–446. doi:10.11233/aquaculturesci1953.53.439. (in Japanese with English abstract) Hamasaki K, Nishimoto S, Okada M, Kimura A, Otsubo K, Dan S. 2020. Dietary effects of phytoplankton and zooplankton on larval survival, duration and growth of four Caridina species (Decapoda: Caridea: Atyidae) under laboratory conditions. Crust Res 49:225-236. doi:10.18353/crustacea.49.0_225. Hancock MA. 1998. The relationship between egg size and embryonic and larval development in the freshwater shrimp Paratya australiensis Kemp (Decapoda: Atyidae). Fresh Biol 39:715– 723. doi:10.1046/j.1365-2427.1998.00323.x. Hancock MA, Hughes JM, Bunn SE. 1998. Influence of genetic and environmental factors on egg and clutch sizes among populations of Paratya australiensis Kemp (Decapoda: Atyidae) in upland rainforest streams, south-east Queensland. Oecologia 115:483– 491. doi:10.1007/s004420050545. Hayashi K, Hamano T. 1984. The complete larval development of Caridina japonica De Man (Decapoda, Caridea, Atyidae) reared in the laboratory. Zool Sci 1:571−589. Hothorn T, Bentz F, Westfall P. 2008. Simultaneous inference in general parametric models. Biom J 50:346–363. doi:10.1002/ bimj.200810425. Ideguchi K, Hamano T, Nakata K. 2007. Timing of egg hatch of amphidromous freshwater shrimps in a small river (the Nishida River), western Japan. Fish Sci 73:961−963. doi:10.1111/j.14442906.2007.01420.x. Ideguchi K, Shibata Y, Kikkawa T. 2000. Larval distribution of amphidromous atyid shrimps (Decapoda: Caridea: Atyidae) in the estuary of the Yukinoura River, Japan. Trans Nagasaki Biol Soc 51:5−13. Ikeda M. 1999. Genetic analyses for the problem of species Paratya compressa (Decapoda: Atyidae). Aquabiology 21:299–307. (in Japanese with English abstract) Imai T. 2006. Reproduction and growth of the freshwater shrimp, Paratya compressa improvisa in the Tsurugata Marsh, Akita Prefecture. Aquacult Sic 54:171–178. doi:10.11233/ aquaculturesci1953.54.171. (in Japanese with English abstract) Jalihal DR, Sankolli KN, Shenoy S. 1993. Evolution of larval developmental patterns and the progress of freshwaterization in the prawn genus Macrobrachium Bate, 1868 (Decapoda, Palaemonidae). Crustaceana 65:365–376. doi:10.1163/156854093X00793. Japan Meteorological Agency. 2020. 10-day mean sea surface temperatures. https://www.data.jma.go.jp/gmd/kaiyou/data/db/ kaikyo/jun/sst_HQ.html. Accessed 20 Sep. 2020. Kawamura T, Akiyama N. 2010. Larval development and salinity response of two species of fresh water shrimps Paratya compressa, P. improvisa, living in Shizuoka Prefecture. Aquacult Sci 58:127–133. doi:10.11233/aquaculturesci.58.127. (in Japanese with English abstract) Mashiko K. 1982. Differences in both the egg size and the clutch size of the freshwater prawn Palaemon paucidens De Hann in the Sagami River. Japan J Ecol 32:445–451. doi:10.18960/ seitai.32.4_445. Mashiko K. 1990. Diversified egg and clutch sizes among local populations of the freshwater prawn Macrobrachium nipponense (De Haan). J Crust Biol 10:306–314. doi:10.1163/193724090X00113. Møller OS, Anger K, Guerao G. 2020. Patterns of larval development. In: Anger K, Harzsch S, Thiel M (eds) Developmental biology and larval ecology, The Natural History of the Crustacea, Vol 7. Oxford University Press, New York, NY, USA, pp. 165–194. Montaini E, Zittelli C, Tredici MR, Grima EM, Sevilla JMF, Pérez JAS. 1995. Long-term preservation of Tetraselmis suecica: influence of storage on viability and fatty acid profile. Aquaculture 134:81–90. doi:10.1016/0044-8486(95)00034-Y. Nakahara Y, Hagiwara A, Miya Y, Hirayama K. 2005. Larval rearing of three amphidromous shrimp species (Atyidae) under different feeding and salinity conditions. Aquacult Sci 53:305–310. doi:10.11233/aquaculturesci1953.53.305. (in Japanese with English abstract) Nakahara Y, Hagiwara A, Miya Y, Hirayama K. 2007. Larval development of three amphidromous shrimp species (Atyidae). Bull Fac Fish Nagasaki Univ 88:43−59. (in Japanese with English abstract) Oeding S, Taffs KH, Reichelt-Brushett A, Oakes JM. 2020. Carbon and nitrogen stable isotope analyses indicate the influence of land use on allochthonous versus autochthonous trophic pathways for a freshwater atyid shrimp. Hydrobiologia 847:2377–2392. doi:10.1007/s10750-020-04209-x. Page TJ, Baker AM, Cook BD, Hughes JM. 2005. Historical transoceanic dispersal of a freshwater shrimp: the colonization of the South Pacific by the genus Paratya (Atyidae). J Biogeogr 32:581–593. doi:10.1111/j.1365-2699.2004.01226.x. Pringle CM, Blake GA, Covich AP, Buzby KM, Finley A. 1993. Effects of omnivorous shrimp in a montane tropical stream: sediment removal, disturbance of sessile invertebrates and enhancement of understory algal biomass. Oecologia 93:1–11. doi:10.1007/BF00321183. R Core Team. 2020. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org/. Accessed 22 June 2020. Saito M, Yamashiro T, Hamano T, Nakata K. 2012. Factors affecting distribution of freshwater shrimps and prawns in the Hiwasa River, southern central Japan. Crust Res 41:27–46. doi:10.18353/ crustacea.41.0_27. Shokita S. 1979. The distribution and speciation of the inland water page 9 of 10Zoological Studies 59:70 (2020)