*2.1. Study Area*

The Yundang Lagoon was selected as the study site, a subtropical urban water body in Xiamen City, Southeast China, which connects to the Western Sea of Xiamen Island through a small canal controlled by a sluice (Figure 1). The Yundang Lagoon used to be a natural bay called Yundang Harbor. The harbor covered an area of about 10 km<sup>2</sup> (6.3 km long and 1.6 km wide). A grea<sup>t</sup> number of land reclamation projects were carried out in the early 1970s, and the Yundang Harbor gradually became a dead lagoon unable to exchange water with the sea. The lagoon area was reduced from the original 10 km<sup>2</sup> to only 2.2 km<sup>2</sup> (1988 data, including 1.0 km<sup>2</sup> of marshes). At present, the total water area of the Yundang Lagoon is approximately 1.5 km2, with a maximum depth of around 5 m and a drainage area of 37 km2, making up 30% of Xiamen Island. The lagoon can be subdivided into di fferent sectors, including a diversion canal, inner and outer sectors, and an innermost canal (Figure 1).

**Figure 1.** Location of the study area (Xiamen Island, Southeast China) and sampling sites (sites A–F) in the Yundang Lagoon. A schematic representation of water circulation in the Yundang Lagoon is given in [34].

### *2.2. Experimental Design and Sampling Activities*

Samples were collected on November 13–14, 2012 at six di fferent sites (A–F) hundreds of meters apart and covering di fferent sectors of the lagoon (Figure 1). Site selection was based on our extensive knowledge of the lagoon's general features [32,34,35]. All sites were characterized by unvegetated soft-bottom sediments with a depth varying from about 1.5–3 meters. At each site, three stations tens of meters apart were randomly chosen and three replicate sediment samples meters apart were collected at each station using a Van Veen grab (30 × 20 cm, penetration depth 18 cm, approximate volume of 7 L), for a total of 54 samples. Each sediment sample was sieved on a mesh size of 0.5 mm and the residue was fixed in a 5% formalin bu ffered solution for macrozoobenthos determination. Prior to sieving, a subsample of the top 1 cm of the sediment was collected from each grab to determine water content and grain-size and chemical (total organic carbon and total nitrogen) analysis. These subsamples were kept refrigerated in a cooler-box until further treatment in the laboratory.

At each site and station, water salinity, temperature, and dissolved oxygen (DO) concentrations were measured using a portable CTD cast (YSI 6600) and near-bottom water samples for chemical analysis (nutrients, biological oxygen demand (BOD5), chemical oxygen demand (CODMn), and chlorophyll-*a* (Chl-*a*)) and suspended solid (SS) determination were collected using a two-liter Niskin bottle, for a total of 18 samples.

### *2.3. Sample Treatment and Analysis*

In the laboratory, water samples were filtered using a 0.45 μm filter. Nutrients (ammonia, nitrate, nitrite, and reactive phosphorus (RP)), BOD5, CODMn, and SS were measured by applying standard analytical methods (Table 1) according to the national standard GB17378.4 [36]. Chl-*a* was measured by fluorometric analysis using a Turner Designs Fluorometer (Mode 10-AU) (San Jose, USA).

**Table 1.** Methods used for seawater analysis in Yundang Lagoon. All methods are according to the national standard GB17378.4 [36].


For the measurements of sediment particle size, diluted hydrochloric acid and hydrogen peroxide were added to the evenly mixed sample to remove carbonates and organic matter. After being washed to remove the acid to attain neutrality, Na-hexametaphosphate 0.6% solution was added to avoid particle flocculation, and the samples were allowed to rest for 24 h. Subsequently, the median particle size of sediments (Md) was measured with a Malvern Mastersizer 2000 laser particle size analyzer (Malvern, United Kingdom), and the measurement data were outputted at 1/4 Φ intervals. The moment method was used to calculate the grain-size parameters of the sediments [37]. The water content (Wc) of the sediment was obtained after drying a sediment subsample at 70 ◦C for 24 h. Analysis of total organic carbon (TOC) and total nitrogen (TN) content of sediments was accomplished by freeze-drying and powdering the sediment sample. Carbonate was removed from the sample with 2 N HCl, at which point it was vacuum-dried. The TOC and TN content of the dried sediment sample was measured using an Elemental vario EL-III element analyzer. Replicate analyses of standards of acetanilide yielded a mean precision of about 0.3% for organic carbon and nitrogen.

The macrozoobenthos from each grab sample were sorted, identified to the species level, when possible, counted under a stereo-microscope and preserved in 75% ethanol. After counting, the wet weight was taken for individuals of the same species in each sample.
