基于最大熵模型的东洞庭湖典型沉水植物生境评价

Maxent model-based evaluation of habitat of typical submerged plants in East Dongting Lake

为阐明东洞庭湖沉水植物生长的关键生境因子和适宜生境分布特征,以东洞庭湖典型沉水植物苦草属为指示物种,结合苦草属地理分布、气候和水文环境变量,基于最大熵模型建立东洞庭湖苦草属生境适宜性模型并对苦草属的生境进行评价。结果表明:训练数据的受试者工作特征曲线下面积值为0.893,最大熵模型预测效果好;年均温变化范围、最湿月降水量、水深、最暖季度降水量、温度季节性变化标准差和平均日较差是影响东洞庭湖苦草属生境的关键因子,以最大训练敏感性及特异性阈值0.503和平衡训练遗漏率、预测面积及阈值0.063作为适宜生境和次适宜生境的划分阈值,将东洞庭湖苦草属生境划分成适宜、次适宜和不适宜3类,其中适宜生境面积为227.62 km^(2),占湖区湿地总面积的17.14%。研究成果揭示了影响东洞庭湖典型沉水植物苦草属生长的生境因子和适生区特征,以期为湖区生态修复提供参考依据。

To clarify the key habitat factors affecting the growth and suitable habitats of submerged plants in East Dongting Lake,a habitat suitability model of Vallisneria in East Dongting Lake was established based on the maximum entropy model,taking the typical submerged plant Vallisneria as the indicator species and combining the geographical distribution,climate and hydrological environmental variables of Vallisneria.The habitat of Vallisneria was evaluated.The results showed that the AUC of the training data was 0.893,and the maximum entropy model had good prediction effect.The range of annual mean temperature,precipitation of the wettest month,water depth,precipitation of warmest quarter,temperature seasonality and mean diurnal range were the key habitat factors of Vallisneria in East Dongting Lake.Taking the threshold of maximum training sensitivity and specificity of 0.503 and the threshold of balanced training omission rate,predicted area and threshold of 0.063 as the dividing thresholds for suitable and sub-suitable habitats,the habitat of Vallisneria in East Dongting Lake could be divided into three categories,namely suitable,sub-suitable,and unsuitable.The suitable habitat area was 227.62 km^(2),accounting for 17.14%of the total wetland area in the lake.The results reveal the habitat factors and the characteristics of suitable areas that affect the growth of submerged plant Vallisneria in East Dongting Lake,providing reference for ecological restoration of the lake.