Abstract The Mediterranean Sea exhibits a strong basin and regional scale phytoplankton variability correlated to its semi-enclosed nature, complex orography and the variety of physical and chemical processes that regulate its productivity. Herein, using 17 years of ocean-color composites, we investigate differences in the regional patterns of interannual variability in satellite-derived chlorophyll (Chl), a proxy for phytoplankton biomass. A neural network classification, based on the Self-Organizing Maps (SOM) analysis in the time domain, is used to reveal regions of similar temporal variability of Chl in the Mediterranean Sea. Characteristic temporal patterns extracted by the SOM analysis show different scales of variation that can be related to already identified oceanographic features and biogeochemical variability in the Mediterranean Sea. Clear differences are noticed between regions located in the Western basin and Adriatic Sea, where rivers, winter mixing and winds are known to drive variations in primary production at regional scale and regions located in the Eastern basin, represented by a large and rather homogeneous region. Using the SOM-defined characteristic temporal patterns of Chl, we analyzed the regional influence of the North Atlantic Oscillation (NAO) and El Nino Southern Oscillation (ENSO) in the long-term (>1 year) Chl variability. Our results indicate that NAO has more influence in the Chl variations occurring in regions located in the Western basin whereas ENSO exhibits higher impact on the central Mediterranean and Eastern basin during its positive phase. Both NAO and ENSO show non-stationary coherence with Mediterranean Chl. The analysis also reveals a sharp regime shift occurring in 2004–2007, when NAO changed from positive to negative values. This shift particularly affected the winter phytoplankton biomass and it is indicative of climate driven ecosystem-level changes in the Mediterranean Sea. Our results stablish a regional connection between interannual phytoplankton variability exhibited in different regions of the Mediterranean Sea and climate variations.

中文翻译：

---

地中海生物地理区域叶绿素年际变异模式

摘要 地中海表现出强烈的盆地和区域尺度浮游植物变异性，与其半封闭性质、复杂的地形以及调节其生产力的各种物理和化学过程有关。在此，我们使用 17 年的海洋颜色复合材料，研究了卫星衍生的叶绿素 (Chl)（浮游植物生物量的代表）的年际变化区域模式的差异。基于时域自组织图 (SOM) 分析的神经网络分类用于揭示地中海中 Chl 的类似时间变异性区域。SOM 分析提取的特征时间模式显示了不同的变化尺度，这些变化可能与地中海已经确定的海洋学特征和生物地球化学变化有关。位于西部盆地和亚得里亚海的区域之间存在明显差异，已知河流、冬季混合和风会驱动区域范围内初级生产的变化，而东部盆地的区域则以大且相当均质的区域为代表。使用 SOM 定义的 Chl 特征时间模式，我们分析了北大西洋涛动 (NAO) 和厄尔尼诺南方涛动 (ENSO) 在长期（> 1 年）Chl 变化中的区域影响。我们的结果表明 NAO 对发生在西部盆地地区的 Chl 变化的影响更大，而 ENSO 在其正阶段对地中海中部和东部盆地的影响更大。NAO 和 ENSO 都显示出与地中海 Chl 的非平稳相干性。分析还揭示了 2004-2007 年发生的急剧政权转变，当时 NAO 从正值变为负值。这种转变特别影响了冬季浮游植物生物量，它表明地中海气候驱动的生态系统水平变化。我们的结果确定了地中海不同地区浮游植物年际变化与气候变化之间的区域联系。