我国是贝类产量大国，贝类以筏式养殖为主。筏式养殖具有双重环境效应，一方面会导致水流变缓、生物沉积加重，另一方面具有净化水体的效果。如何优化养殖规模发挥其最大生态效益，是管理部门急需解决的问题。因此，建立有效的模型综合评估养殖规模及其环境效应尤为重要，可为海洋管理提供科学依据。深圳湾是深圳和香港的跨境海湾，香港侧牡蛎筏式养殖约占总水域面积的20%。长历时的水质监测结果分析表明，深圳湾为磷限制富营养化海湾，初级生产力水平为70.3-568.8 mg C/(m2·d)；养殖区年平均叶绿素a与无机氮、正磷酸盐均有较高相关性，养殖区和非养殖区叶绿素和营养盐分布的差异表明牡蛎养殖具有显著的藻削减和营养盐调控效应。通过二维和三维的CFD模型研究了小尺度悬浮筏式养殖单元的局部流阻损失，在此基础上提出了考虑单元间距和淹没长度影响的无量纲阻力模型。该模型将筏式养殖排布分解为两个正交方向，并在两个方向上估算阻力效应，得到的阻力系数与CFD模型计算结果吻合。并将阻力模型以动量损失的形式耦合到浅水模型动量方程中，建立筏式养殖阻水参数化模型。最后，该模型应用于深圳湾，模型计算结果表明，流速验证平均误差下降了10.4%，由于筏式养殖的阻水效应，中湾养殖区年平均流速下降了23.22%，非养殖区上升15.65%。基于筏式养殖参数化模型，提出了水质-底泥-贝类动态能量收支耦合的水质生态模型，并应用于深圳湾。模型计算结果和观测值比较吻合，模型动态评估深圳湾牡蛎养殖容量为35 ind/m2，与牡蛎摄食实验38 ind/m2接近。计算结果表明，2016年1-6月半年内，牡蛎摄食藻获得22.59 t氮和2.82 t磷，排泄到水体中的氮和磷的含量约14.41 t和2.08 t，并形成约5.67 t氮和0.65 t磷的生物沉积，牡蛎组织体固定了约0.90 t氮和0.13 t磷。筏式养殖对深圳湾内湾藻的峰值削减不明显，但可有效降低海湾的平均藻浓度（下降了19.8%），其中养殖区最明显（下降48.65%），其机制为养殖区的阻水效应延长了水体停留时间以及周期性的潮流增加了牡蛎的摄食影响范围，形成了水动力协同牡蛎摄食的藻削减机制。而深圳湾中部非养殖区流速增大，加快了氮磷运移输出。深圳湾中湾，养殖区的叶绿素平均浓度约为非养殖区的70%左右，由于牡蛎排泄作用，养殖区的无机磷和氨氮均高于湾中部非养殖区。

China is ranked highest production of shellfish, and raft culture is the dominant type of offshore shellfish aquaculture. Raft farming has dual environmental effects. On the one hand, it can slow down water flow and increase biodeposition. On the other hand, it can purify water. How to optimize the scale of breeding to maximize its ecological benefits is an urgent problem to be solved by the management department. Therefore, it is particularly important to establish an effective model to comprehensively evaluate the scale of aquaculture and its environmental effects, which can provide a scientific basis for Marine management.Shenzhen Bay is the cross-border bay between Shenzhen and Hong Kong. Raft oyster farming in Hong Kong waters accounts for about 20 % of the total water area. The analysis of long-term water quality monitoring data showed that Shenzhen Bay was classified as phosphorus restricted eutrophication, and the primary productivity level was 70.3-568.8 mg C/(m2·d). The annual average chlorophyll a was highly correlated with inorganic nitrogen and orthophosphate in the aquaculture area. The difference of chlorophyll a and nutrient distribution between cultured and non-cultured areas indicated that oyster culture had significant algal reduction and nutrient control effects. The local flow resistance loss of small-scale suspended raft aquaculture units was investigated using 2 dimensional and 3 dimensional CFD model, and a dimensionless resistance model considering the effects of unit spacing and submerged length was proposed based on the flow resistance loss. In this model, the raft culture arrangement was decomposed into two orthogonal directions, and the drag effect was estimated in both directions. The drag coefficient obtained by the model was in good agreement with the CFD model calculation results. A parametrized water resistance model of raft culture was established by coupling the resistance model to the momentum equation of shallow water model in the form of momentum loss. Finally, the model was applied to Shenzhen Bay, and the results showed that the average error of velocity verification decreased by 10.4%. Due to the water-blocking effect of raft farming, the annual mean velocity decreased by 23.22% in the aquaculture area and increased by 15.65% in the non-aquaculture area in Middle Bay.Based on the raft aquaculture parametrized model, a water quality ecological model based on the coupling of water quality, sediment and shellfish dynamic energy budget was proposed and applied to Shenzhen Bay. The calculated results of the model agreed well with the observed values. The dynamic estimation of oyster cultured capacity in Shenzhen Bay by the model was 35 ind/m2, which was close to 38 ind/m2 in oyster feeding experiment. The calculation results showed that oyster ingested 22.59 t N and 2.82 t of P, excreted about 14.41 t N and 2.08 t P into water, and formed about 5.67 t N and 0.65 t P by biodeposition. About 0.90 t N and 0.13 t P were fixed in oyster tissues during the six months culture from January to June 2016. Raft culture did not significantly reduce the peak value of algae in the inner bay of Shenzhen Bay, but it could effectively reduce the average algal concentration in the bay (decreased by 19.8%), especially in the aquaculture area (decreased by 48.65%). The mechanism was that the water retention time was prolonged by the water-blocking effect in the aquaculture area and the periodic tidal current increased the feeding range of oysters, which formed a algal reduction mechanism of hydrodynamic cooperation with oyster feeding. However, the increase of flow velocity in the non-aquaculture area in the middle of Shenzhen Bay accelerated the migration and output of nitrogen and phosphorus. In the middle Bay, the average concentration of chlorophyll a in the cultured area was about 70% of that in the non-cultured area. Due to the excretion effect of oysters, the inorganic phosphorus and ammonia nitrogen in the cultured area were both higher than those in the non-cultured area of the middle Bay.