Connectivity of nekton assemblages along articial reefs and adjacent waters in Haizhou Bay

: 3 The successful construction of marine protected areas (MPAs) in temperate waters largely 4 depends on our understanding of the distribution and coexistence of organisms with varying 5 habitat preferences, which helps us to better understand the community patterns mediated by 6 connectivity in coastal areas. This study was conducted to examine the connectivity of nekton 7 assemblages in artificial reefs and adjacent waters, which included five habitats: the artificial 8 reef area (AR), aquaculture area (AA), natural area (NA), estuary area (EA) and comprehensive 9 effect area (CEA), in Haizhou Bay in October 2020. Analysis of variance (ANOVA) showed 10 that there were significant differences in the characteristics and abundances of nekton in each 11 habitat ( P <0.05). Approximately 38.2% of the individuals were found in at least three habitats, 12 and very few species were present in only a single habitat. Several highly abundant nekton 13 species were selected according to the kernel density estimates (KDEs), and their body lengths 14 varied gradationally among habitats, potentially indicating migration and diffusion during their 15 life history. The results showed that artificial reefs and adjacent waters in Haizhou Bay are 16 related by similar nekton assemblages and ontogenetic variation. Finally, this study has 17 implications for the conservation and monitoring of nekton assemblages in artificial reefs and 18 adjacent waters, highlighting that the principle of connectivity should be taken into 19 consideration in the design of MPAs and MPA networks that can be applied in different stages 20 of implementation and in different combinations of scenarios. The respectively. These results indicated that C. lucidus has a strong ability to shift among habitats and a wide distribution in different habitats. The body length of C. lucidus was larger in and than in the other possibly because the cross- habitat shifts of C. lucidus led to some potential connectivity between RA and EA. The maximum sizes of O. oratoria and P. trituberculatus were 15.6 cm and 18.6 cm in CEA, and the minimum sizes were 4.8 cm and 3.5 cm in AR and NA. The maximum and minimum sizes of T. curvirostris were 11.4 cm and 4.7 cm in CEA. These results could indicate some possible individual shifts between habitats.


23
The global trends of continuous processes associated with human activities and 24 coastal development, such as overfishing, habitat destruction and marine environmental 25 pollution, have led to the general degradation of the entire coastal ecosystem, which 26 has imposed tremendous pressure on estuaries, harbors, gulfs and nearshore regions 27 (Walker et al., 2014;Dance et al., 2015;Cordova et al., 2018;Reis et al., 2019). Over 28 the years, artificial reefs have been constructed as human-made structures to increase 29 environmental quality and species abundance in marine ecosystems for coastal 30 ecological restoration (Seaman and Sprague 1991;Anne et al. 2015;Folpp et al. 2020) 31 by creating suitable habitats and places for many marine organisms to grow, reproduce, 32 forage and hide (Sherman et al., 2002) . At present, ecological principles combining the 33 planning, design and operation of artificial reefs have been extensively investigated in 34 many coastal areas (Whitmarsh et al. 2008;Walker et al. 2014;Anne et al. 2015;Folpp et al. 2020). Given that artificial reefs aim to retard marine habitat degradation, protect 36 endangered species and restore biodiversity, artificial reefs are regarded as conservation 37 and enhancement tools for marine environments and habitat recovery (Dafforn et al., 38 2015;Becker et al., 2017). 39 Habitat connectivity provides an important perspective for further study of coastal 40 ecosystems (Dance et al., 2015;Diana et al., 2018). It includes two mechanisms:  (Wang, 1993). Because of 51 numerous human activities, such as overfishing, port construction and waterway 52 transportation, since the 1980s, the habitat environment and fishery resources in 53 Haizhou Bay have been vastly and adversely affected, resulting in the fragmentation of 54 habitats and the destruction of ecosystem structure (Zhang et al., 2006;Zhang et al., 55 2013). Since 2002, the local government has begun to build marine protected areas 56 (MPAs) dominated by artificial reefs for ecological restoration and resource 57 conservation in Haizhou Bay (Zhang et al., 2006;Wu et al., 2012). The construction of 58 artificial reefs affects aquatic biodiversity and food web ecology by affecting the flow 59 of water, sediments and organisms (Clark et al. 1999;Sherman et al. 2002). These To support the sustainable socioeconomic development of MPAs, therefore, it is 66 necessary to thoroughly explore the relationship between biodiversity and connectivity 67 in artificial reefs and adjacent waters.

68
In this study, we aim to (1) analyze the nekton assemblages and connectivity in According to the available geographical coordinate data sets in the study area 89 (34°49.20-34°55.00N, 119°16.167-119°59.50E), five major investigation areas ranging 90 from the Linhong Estuary to the artificial reef area were set, including the estuary area 91 (EA), the most polluted area and featuring a coarse sand substrate; the aquaculture area 92 (AA), an area mainly used for culturing shellfish and algae; the artificial reef area (AR), 93 a protected area consisting of a series of reefs on the sea bottom; the natural area (NA), an area that has not been overly impacted by humans; and the comprehensive effect 95 area (CEA), an area in which several habitats co-occur. The distribution of the sampling 96 sites is shown in Figure 1.

97
To determine how the nekton utilized the habitat, the species were divided into the 98 following nine habitat groups: (1) AR species, which were present only in the artificial

178
The results of one-way ANOVA showed that the mean numbers of individuals of 179 fish, shrimp, crab, shellfish, and cephalopods were significantly different between NA 180 and AR and between CEA and AA (P<0.05). The mean number of individuals in AR,

181
CEA and AA was significantly higher than that in NA (P<0.05).

195
Univariate PERMANOVA results showed that the species abundances of the five 196 marine habitats were significantly different (P<0.05) ( Table 1). Analysis of each habitat 197 separately revealed that there were significant differences in species abundance 198 between the two groups of five habitats (P<0.05) ( Table 2). The community structure 199 of the five habitats was inconsistent not only at the overall level but also among each 200 of the habitats. The results of clustering analysis were visualized by nonmetric multidimensional 208 scale analysis ( Figure 5). According to NMDS, habitat arrangement in the multivariate 209 space clearly revealed a separation along the NMDS 1 axis, which clearly showed that  In the present study, strong links were also identified between the size-related swimming ability, also showed little variation in this study. The reason is that P.

306
trituberculatus is active in coastal waters but not in offshore regions, such as AR and 307 NA (Song et al. 1989). Moreover, EA is also the discharge area of P. trituberculatus in 308 Haizhou Bay, which verifies the view in this study. In total, the body length of each 309 nekton gradually increases from EA to RA, not only indicating that EA is a breeding 310 ground for most fish, shrimp and crabs but also reflecting the ontogenesis of some   Study area and sampling site 110 AR, CEA, AA, EA and NA represent the arti cial reef area, comprehensive effect area, aquaculture 111 area, estuary area and natural area, respectively. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.