The Distribution of the Invasive Shrimp Neocaridina Davidi (Decapoda: Caridea: Atyidae) in Relation to Environmental Parameters in a Stream at Kunitachi, Tokyo, Japan
Total Page：16
File Type：pdf, Size：1020Kb
Crustacean Research 2021 Vol.50: 33–39 ©Carcinological Society of Japan. doi: 10.18353/crustacea.50.0_33 The distribution of the invasive shrimp Neocaridina davidi (Decapoda: Caridea: Atyidae) in relation to environmental parameters in a stream at Kunitachi, Tokyo, Japan Keisuke Onuki Abstract.―The distribution of the invasive freshwater shrimp, Neocaridina davidi, in a spring-fed stream at Kunitachi, Tokyo, Japan, was investigated in relation to the stream’s environmental parameters. Compared to previous studies, the population density of N. davidi was found to be remarkably high (up to 315 ind./m2). The in- crease in population density is likely because the water temperature in the spring-fed stream remains 17–20°C in the winter, prolonging the period during which the shrimps can spawn. Generalized linear mixed models showed that adult shrimps were more abundant in areas with high emergent plant coverage. In contrast, juveniles were more abundant in areas with low current velocity. These results provide essential clues to the interspecific relationships between N. davidi and native spring organisms that inhabit the same physical environments. Key words: Atyidae, invasive species, microhabitat, springs, Tokyo ■ Introduction Neocaridina spp. and the native organisms. The recent invasion of N. davidi has been ob- Species of Neocaridina are landlocked fresh- served at Mama-shita springs (＝Mama-shita water shrimps widely distributed in East Asia Yusui) in Kunitachi, Tokyo, Japan, where this (Cai, 1996). Used as fishing bait and ornamen- study was conducted. Mama-shita springs are tal species, they have spread across the world critical components of Tokyo’s aqueous envi- (Englund & Cai, 1999; Niwa, 2010; Klotz et ronment (Ueda et al., 2000; Takamura & al., 2013; Jabłońska et al., 2018; Weiperth et Marui, 2014). The springs are home to several al., 2019). In Japan, several invasive species of rare spring-dependent species, including Japa- Neocaridina, including N. davidi, were found nese eight-barbel loach Lefua echigonia and in western Japan in the early 2000s (Niwa et Amur minnow Rhynchocypris lagowskii stein- al., 2005; Niwa, 2010; Fujita et al., 2011; dachneri (Nishida et al., 2014). However, it is Toyota & Seki, 2014). These species have re- unclear how Neocaridina spp. affect the native cently spread to eastern Japan (Kanazawa, organisms in the unique environment of a 2015; Nishida, 2016; Katayama et al., 2017; spring. For example, the habitat of Neocaridina Mitsugi et al., 2017). The introduction and ex- spp., which can shed light on the interaction pansion of the invasive Neocaridina spp. in Ja- between the invasive and native population, is pan could lead to interbreeding with, and the unknown. competitive exclusion of, native shrimps (Niwa, In this study, I examined the distribution of 2010; Nishida, 2016; Katayama et al., 2017). N. davidi, a representative of the invasive Neo- However, few studies have examined the inter- caridina spp., in the spring-fed stream in the specific relationships between the invasive Mama-shita Springs Park (＝Mama-shita Yusui Received: 4 Dec 2020. Accepted: 26 Feb 2021. Published online: 1 Apr 2021. 33 KEISUKE ONUKI Koen) in Kunitachi, Tokyo, Japan, and investi- Estimating the number of adult and juvenile gated the effect of the stream’s physical envi- shrimp ronmental parameters on the distribution. The collected specimens were identified un- der a stereomicroscope. The adult and juvenile ■ Materials and Methods Neocaridina davidi were identified in each study section and counted. Identification of N. Study site davidi was performed according to Toyota & The study was conducted monthly in a Seki (2014). According to Mitsugi & Suzuki spring-fed stream in the Mama-shita Springs (2018), individuals with a carapace length of Park (35°24′N, 139°25′E) at Kunitachi, Tokyo, more than or less than 3 mm were defined as Japan, from September to November 2019. The adults or juveniles, respectively. stream is approximately 300 m in length and The percentage of adult and juvenile N. davidi joins the Fuchu waterway. In the stream, 12 in each section was studied. One hundred indi- study sections (Sec. 1–Sec. 12), each of a con- viduals of N. davidi were randomly selected stant area of 2 m2 in September and November from the specimens of each study section to and 4 m2 in October, were set up (Fig. 1). have their carapace length measured. For study sections with fewer than 100 sampled individu- Sampling shrimp als, the carapace length of all the individuals Sampling was carried out at each study sec- was measured. tion by settling a small set net with a mesh size The number of adults and juveniles in each of 2 mm in the stream and blocking its flow. study section was calculated using the follow- All samplings were performed by the same per- ing equation; sons by cornering shrimps into the set net. The * Pa* Pj collected specimens were brought back to the PPa＝ all ××,. PPj＝ all PP** laboratory, fixed in 10% neutral formalin, stained all all in Rose Bengal, and stored in 70% ethanol. P is the number of individuals in the entire study section, P* is the number of individuals whose carapace length was measured, and the Fig. 1. Map of the Mama-shita Springs Park (left) and the location of the study sections (right). Arrows indicate water flow direction. 34 Crustacean Research 50 ENVIRONMENTAL PARAMETERS AFFECTING INVASIVE SHRIMP DISTRIBUTION indices all, a, and, j denote all individuals, The ki value is a constant given to the sedi- adults, and juveniles, respectively. ment category i (large gravel: 4, medium grav- el: 3, fine gravel: 2, fine sand: 1) and Ci is the Measuring physical environmental parameters coverage of the sediment category i (Dohi et In September and October, surveys were car- al., 2006). ried out in all sections except for Sec. 4 and 9. In November, all 12 sections were surveyed. Statistical analysis The emergent plant coverage, submerged plant Two generalized linear mixed models coverage, current velocity, water depth, water (GLMM) were used to determine the effect of temperature, and sediment score were exam- the stream’s physical environmental parameters ined in each study section using the following on the population density of the adult and juve- methods. nile N. davidi. R 4.0.2 was used for all analyses The respective coverage of the emergent (R Core Team, 2020). plant, Japanese sweet flag Acorus gramineus, The spatial autocorrelation variables were and submerged plant, Japanese bur-reed Spar- calculated to consider the effect of spatial ganium japonicum, was calculated by dividing proximity on the distribution of N. davidi. Mo- the plant community’s area in the study section ran’s eigenvector maps (MEM) were construct- by the area of the study section. The average ed using the relative neighborhood method, current velocity of each study section was and the scores for each study section were cal- calculated by measuring the time required for a culated using the R package “adespatial” (Dray float to flow for 1 m at the center of the current et al., 2020). MEM1 to MEM4 were used in three times. The average water depth of each the following analysis (Fig. 2). study section was calculated by measuring the A GLMM with a negative binomial distribu- water depth at five randomly-selected points tion was constructed using the R package using a folded tape measure. Water temperature “MASS” to investigate the relationship be- was measured using Quanta (HYDROLAB). tween the density of the juvenile or adult N. Sediment was classified into four categories: davidi and the stream’s physical environmental large gravels (＞64 mm), medium gravels (4 to parameters. The response variable was the 64 mm), fine gravel (2 to 4 mm), and fine sand number of juvenile or adult N. davidi in each (＜2 mm); the area of coverage of each catego- study section. The explanatory variables were ry was calculated. The sediment score S, which the scaled physical environmental parameters indicated the sediment roughness in each study and MEM1 to MEM4, with an offset term ac- section, was calculated by counting for the area of each study section. To avoid multicollinearity, Spearman’s correlation S＝∑ k C . i i coefficients were calculated among all explana- Fig. 2. The spatial distributions of the scores of 1–4th axes of Moran’s eigenvector maps (MEM). Large squares represent high MEM scores, and white and black squares indicate negative and positive scores, respectively. Crustacean Research 50 35 KEISUKE ONUKI tory variables, and water depth and sediment increasing the density of the shrimp. In fact, scores, which have ρ＞0.6 with current velocity, three and two ovigerous females were found in were excluded from the explanatory variables. October and November, respectively, when no The model was selected using a multi-model ovigerous females were found in Mitsugi & inference approach. All the subsets of the mod- Suzuki (2018) (Table 1; Fig. 3(A)). els based on the global model were produced and ranked based on the Akaike’s information Physical environmental parameters affecting criterion (AIC). The best performing models the distribution of adult and juvenile N. davidi with ΔAIC＜2 were used to perform model av- The physical environmental parameters of eraging using the R package “MuMIn” (Bartoń, the stream were measured (Fig. 3(B)). The pa- 2020). rameters such as the emergent plant coverage, submerged plant coverage, water depth, current ■ Results and Discussion velocity, and sediment score differed among the study sections. On the other hand, the water High Neocaridina davidi density temperature was generally comparable in all During the study, the invasive shrimp, N. da- study sections at about 19.7°C (Fig. 3(B)). vidi, and the crayfish, Procambarus clarkii, Although spatial autocorrelation was detect- were collected (Table 1; Fig.