基于two-stage GAM的万泉河口爪哇拟虾虎鱼仔鱼对环境因子的响应OA
Response of Pseudogobius javanicus Larvae to Environmental Factors in the Wanquan River Estuary Based on a Two-Stage GAM
万泉河口作为典型的热带河口生态系统,为多种鱼类提供了关键的繁殖和育幼栖息地,其复杂生境条件和水文水动力条件对鱼类早期资源的分布格局影响显著.爪哇拟虾虎鱼(Pseudogobius javanicus)是该区域仔鱼群落的优势种,对环境变化高度敏感,是评估栖息地质量的重要指示物种.为探明爪哇拟虾虎鱼仔鱼动态的关键驱动因子及响应关系,本研究于2021年5月至2022年4月在万泉河口的7个站点4个生境类型开展周年监测,采集爪哇拟虾虎鱼仔鱼样本及9项环境因子数据,并构建两阶段广义加性模型(two-stage GAM)评估仔鱼对环境因子的响应关系.结果显示,爪哇拟虾虎鱼仔鱼丰度呈现明显的季节性和空间异质性,表现为雨季(6-8月)显著高于旱季,红树林边缘和沉水植物区最高,是仔鱼育幼核心区.GAM1显示,温度、氨氮、溶解氧及叶绿素a(Chl-a)为影响出现概率的关键因子(累计偏差解释率为77.6%,P<0.05).GAM2模型中盐度、pH、温度及Chl-a共同调控仔鱼丰度(累计偏差解释率为77.3%,P<0.05).Two-stage GAM识别的最适"环境窗口"为:温度 26~30 ℃、盐度 4~6 和 20~25、pH 7.2~7.6、氨氮<0.2 mg/L、Chl-a<10 μg/L、DO 6~7mg/L.本研究证实,two-stage GAM能有效克服仔鱼资源调查中零膨胀数据的干扰,可准确揭示仔鱼与环境因子的响应关系.基于研究结果,建议优先保护红树林边缘与沉水植物区,实施生态修复与扰动控制;强化雨季期间氨氮与Chl-a阈值监管,并设立6-8月季节性禁渔区,以提升栖息地质量、促进万泉河口鱼类早期资源的持续恢复和保护.
As a typical tropical estuarine ecosystem,the Wanquan River Estuary provides essential spawning and nursery habitats for various fish species.The complex habitat conditions and hydrological and hydrodynamic environment in this area have significantly influenced the distribution patterns of early fish resources.Among these,the Javanese goby(Pseudogobius javanicus)stands out as the dominant species in the larval fish community of the region.This species is particularly sensitive to environmental changes,making it a crucial bioindicator for assessing habitat quality.To elucidate the key driving factors behind the dynamics of Javanese goby larvae and understand their responses to environmental factors,a year-round monitoring study was conducted from May 2021 to April 2022.The study was conducted at seven sampling sites within the Wanquan River Estuary,which collectively represent four distinct habitat types.During the monitoring period,larval samples of P.javanicus were collected systematically,and nine environmental parameters were measured.The objective of this study was to comprehensively investigate both the temporal and spatial patterns of larval abundance,and to reveal the underlying environmental factors influencing these patterns.A two-stage generalized additive model(GAM)was constructed to accurately model the response of larvae to environmental conditions.This approach is particularly effective for dealing with zero-inflated data common in larval fish surveys,where many observations may record zero abundance due to absence of sampling points or times.The two-stage GAM partitions the modeling process into occurrence probability and conditional abundance stages,allowing for a more nuanced understanding of how environmental factors affect presence and quantity separately.The results of the present study demonstrate that the abundance of Javanese goby larvae exhibits pronounced seasonality and spatial heterogeneity.Specifically,larval abundance was significantly higher during the rainy season(June-August)than during the dry season.Furthermore,the highest larval densities were consistently observed along mangrove fringes and submerged macrophyte zones,indicating that these habitats serve as critical nursery grounds for the species.The spatial preference for these habitats underscores their ecological importance,as they provide essential shelter and feeding opportunities for early life stages.The first component of the two-stage GAM(GAM1),which analyzed the presence versus absence data,identified temperature,ammonia nitrogen concentration,dissolved oxygen levels,and chlorophyll-a concentration as the primary environmental factors influencing the probability of occurrence of Javanese goby larvae.This model explained a cumulative deviance of 77.6%with statistically significant results(P<0.05),highlighting the strong effect of these variables on larval habitat suitability.High ammonia nitrogen levels and low dissolved oxygen typically indicate poor water quality and were inversely related to the probability of larval presence,whereas optimal temperature ranges and moderate chlorophyll-a concentrations favor its occurrence.The second component(GAM2),focusing on larval abundance conditional on presence,revealed that salinity,pH,temperature,and chlorophyll-a collectively governed the variation in larval density,with a cumulative deviance explanation of 77.3%(P<0.05).Salinity emerged as a critical factor,with two optimal ranges identified:10-15 and 20-25,suggesting some flexibility in larval tolerance or site-dependent habitat conditions.The pH window of 7.2 to 7.6 further delineated favorable acidic-alkaline conditions for larval growth and survival.The optimal temperature fell within the 26-30 ℃ range,which is consistent with the species' tropical affinity.Chlorophyll-a levels below 10 μg/L corresponded with better larval abundance,likely reflecting balanced primary productivity that supports appropriate food web dynamics without leading to eutrophication stress.Collectively,the two-stage GAM approach effectively overcame the challenges posed by zero-inflated larval count data,enabling accurate and detailed elucidation of the complex relationships between larval fish and their environmental context.Identifying specific"environmental windows"for key factors provides actionable thresholds that can guide management and conservation efforts.Based on these findings,the study recommends prioritizing the protection of mangrove edge zones and submerged vegetation areas within the Wanquan River Estuary,given their role as core nursery habitats for the Javanese goby and potentially other estuarine fish species.Active ecological restoration initiatives and disturbance mitigation measures should be implemented in these sensitive habitats to maintain or enhance their quality.Moreover,heightened monitoring and regulation of ammonia nitrogen and chlorophyll-a concentrations,especially during the rainy season from June to August when larval abundance peaks,should be enforced to prevent deterioration in water quality.Additionally,the establishment of a seasonal fishing moratorium for the rainy season is recommended.Designating no-fishing zones during this critical period can reduce anthropogenic pressure on the early life stages of fish,thereby facilitating recruitment success and long-term population sustainability.
王先宁;陈枳锐;陈涛;罗亦斯;王智豪;程飞;谢松光;宋一清
海南大学海洋生物与水产学院 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228海南大学生态学院 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228||海南省热带水生生物技术重点实验室 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228||海南省热带水生生物技术重点实验室 海南 海口 570228||海南大学南海海洋资源利用国家重点实验室 海南 海口 570228海南大学海洋生物与水产学院 海南 海口 570228||海南大学南海海洋资源利用国家重点实验室 海南 海口 570228
农业科技
两阶段广义加性模型爪哇拟虾虎鱼时空分布万泉河口环境因子
Two-stage GAMPseudogobius javanicusSpatiotemporal distributionWanquan River EstuaryEnvironmental factors
《渔业科学进展》 2026 (3)
29-39,11
海南省科技专项项目(ZDYF2023SHFZ101ZDYF2024SHFZ070)、国家自然科学基金(323609174237611442166004)、海南大学科研启动基金(KYQD(ZR)21130)共同资助.
评论