1 2 One notable type of bioturbation in marine soft sediments involves the excavation of large pits and 3 displacement of sediment associated with predator foraging for infaunal benthos. Batoids are among 4 the most powerful excavators, yet their impact on sediment has been poorly studied. For expansive 5 temperate tidal flats, only relatively small proportions of the habitat can be sampled due to physical 6 and logistical constraints. The knowledge of the dynamics of these habitats, including the spatial and 7 temporal distribution of ray bioturbation, thus remains limited. We combined the use of aerial 8 photogrammetry and in situ benthic sampling to quantify stingray feeding pits in Tomioka Bay, 9Amakusa, Japan. Specifically, we mapped newly-formed pits over an 11-ha section of an intertidal 10 sandflat over two consecutive daytime low tides. Pit size and distribution patterns were assumed to 11 scale with fish size and reflect size-specific feeding behaviors, respectively. In situ benthic surveys 12were conducted for sandflat-surface elevation and prey density (callianassid shrimp). The volume 13 versus area relationship was established as a logistic function for pits of varying sizes by 14 photographing and refilling them with sediment. This relationship was applied to the area of every 15 pit detected by air to estimate volume, in which special attention was paid to ray ontogenetic change 16 in space utilization patterns. In total, 18103 new pits were formed per day, with a mean individual 17 area of 1060 cm 2 . The pits were divided into six groups (G1 to G6 in increasing areas), with 18 abundances of G1, G2+G3, and G4−G6 being medium, high, and low, respectively. Statistical 19 analyses using generalized linear models revealed a marked preference for the higher prey-density 20 areas in G1 and the restriction of feeding grounds of G4−G6 to the lower shore, with G2+G3 being 21 generalists for prey density and sandflat elevation. The lower degrees of overall bioturbation by G1 22 and G4−G6 were spatially structured for the eight sub-areas demarcated by prey density and sandflat 23 elevation, while G2+G3 homogenized the state over the sandflat. The newly-formed pits' sub-areal 24 2 mean numerical, excavated-areal, and displaced-sediment-volume densities per day were confined to 25 small ranges: 0.14−0.17 m −2 , 132−223 cm 2 m −2 , and 551−879 cm 3 m −2 (latter two including 119 26 shallow non-pit excavations). These bioturbation rates are positioned at relatively high levels 27 compared with those by rays from other geographic regions. The present procedure is applicable to 28 the assessment of disturbance by any surface-sediment excavators on tidal flats if their pit 29 dimensions are discernible from the air. 30 31

No abstract

No abstract

No abstract

Authors' contributions: AT supervised the study and wrote the manuscript. YS and SO undertook the laboratory experiments. JI undertook the field experiment. YH and SS conducted the larval sampling at sea. ST assisted in analyzing data and interpretation of results.

Running page head: Umezawa et al.: Field diet of callianassid shrimp larvae ABSTRACT: The field diet of meroplanktonic decapod crustacean larvae is poorly known, despite standard use of microzooplankton as food in laboratory culture. Using callianassid shrimp Nihonotrypaea harmandi larvae collected from a 65 m deep inner-shelf location off mid-western Kyushu, Japan, between June and August 2012 and 2013 and mass-reared in the laboratory, phytoplankton-based diet through larval development (zoeae I−VI to decapodid) was demonstrated. When the pure-cultured diatom Chaetoceros gracilis was fed to zoeae, survival rate to decapodids was 3.4 to 3.9% in 26 to 40 d at 22°C, which was comparable to previous rearing results for zoeae fed microzooplankton. Trophic enrichment factors (TEFs) 1 from stable isotope (SI) analysis of zoeal whole-body tissue in the laboratory were 2.0‰ for δ C and 1.9‰ for δ 15 N. In the field water column, diatoms dominated the nano-to micro-sized plankton, accounting for 38 to 81% of the biovolume, followed by heterotrophic protists. The trophic position (TP) estimated from amino acid-specific δ 15 N values for the field-collected zoeae VI was 2.1 (TPGlu/Phe) or 2.7 (TPAla/Phe), suggesting that those zoeae fed on mixtures of phytoplankton and heterotrophs including protists. Bulk SI analyses were performed for particulate organic matter (POM; proxy for phytoplankton), microzooplankton (mainly calanoid copepods), and shrimp zoeae to elucidate the diet of larvae in the water column. A shift in SI from fresh to degraded POM was determined through the incubation of field-collected POM. Based on this shift during degradation and larval TEFs, phytoplankton and their sinking detritus with heterotrophic protists were estimated to be the principal diet for those larvae residing mostly below the chlorophyll maximum layer.

25Species of the free-burrowing amphipod genus, Urothoe, are common macrobenthos on open sandy 26 beaches. On intertidal sandflats, some species are associated with burrows or tubes of large infauna. 27How this link is formed under sheltered settings was examined. On an intertidal sandflat emersed for 28 300 m seaward in mid-western Kyushu, Japan, U. carda co-occurred with the deep burrow-dwelling 29 callianassid shrimp, Nihonotrypaea harmandi. Amphipods resided in the surface 5-cm sediment 30 outside shrimp burrows, as confirmed by sediment coring and burrow casting. In the summertime 31 during 1980 to 1981, the shrimp and amphipod populations were confined to the upper shore at 32 mean densities of 182 and 701 inds m -2 , respectively. In winter to spring, when the sediment surface 33 mixing was caused by seasonal wind-induced waves, only the amphipod extended distribution to the 34 lowest shore. By 1983, the shrimp increased mean density by 2.5 times and distribution range to the 35 lowest shore. In the summers of 1984, 2010, and 2015, the amphipod extended distribution to the 36 lowest shore, with small variations in population size. Three marked changes in substrate properties 37 were associated with the shrimp inhabitation: thicker oxidized layer (proxy for oxygenated layer) in 38 the sediment column; looser surface sediment, as evaluated with vane shear strength; and coarser 39 and better-sorted surface sediment with less mud content. At least the former two changes were 40 attributable to shrimp bioturbation, which could provide the amphipod with more permeable and 41 softer substrates, leading to the formation of facultative commensalism. 42

Many decapod crustaceans in marine intertidal habitats release larvae toward coastal oceans, from which postlarvae (decapodids: settling‐stage larvae) return home. Decapodid settlement processes are poorly understood. Previous studies showed that in Kyushu, Japan, the callianassid shrimp population on an intertidal sandflat of an open bay joining the coastal ocean near a large estuary released eight batches of larvae basically in a semilunar cycle from June through October and that decapodids performed diel vertical migration, occurring in the water column nocturnally. We conducted (a) frequent sampling for population density and size‐composition on the sandflat through one reproductive season, (b) planktonic and benthic sampling for decapodids around the bay mouth, and (c) current meter deployment at three points across the bay mouth for tidal harmonic analysis. On the sandflat, six batches of newly‐settled decapodids (settlers) occurred in a semilunar periodicity until October, with peaks occurring 0–3 days before syzygy dates except for the first one. For larval Batches 1–4, buoyancy‐driven shoreward subsurface currents during July to mid‐October would transport some pre‐decapodid‐stage larvae (zoeae) toward the bay. The absence of expected settler Batches 7–8 would be due to the converse subsurface currents caused by water‐column mixing and seasonal winds after mid‐October, carrying zoeae offshore. Once in the bay, phasing of night and nighttime‐averaged shoreward tidal current explained the settlement pattern for Batches 1–4. For Batches 5–6 occurring in mid‐September to mid‐October, water currents generated by seasonal wind and tidal forcings may have caused peak settlement after the time expected from tidally‐driven decapodid transport.