While the developmental dynamics of ecosystems have been studied in many natural systems, it is unclear if these patterns are to be expected in engineered ecosystems where components have not co-developed over time. Green roofs often begin with a nutrient rich substrate that is paired with slow-growing plants adapted to nutrient poor conditions - perhaps resulting in different developmental dynamics than natural ecosystems. We evaluated changes over time in green roof nutrient content and vegetation communities using a chronosequence in southern Sweden of similarly designed but different aged green roofs spanning between 2 and 22 years. Substrate depth, substrate nitrogen (N), and total N pool sizes varied positively with roof age. These dynamics suggest an accumulation of 2.9 ± 1.1 g N/m²/yr with no indication of leveling off after 22 years. Plant N content (%) positively varied with roof age but plant biomass, plant nutrient pools, and plant diversity did not vary with age. These dynamics indicate a novel developmental scenario where the ecosystem begins with near-stable plant biomass but still accumulates N in the substrate at rates on par with many secondary successional systems. The apparent accumulation of N could not be accounted for by N deposition rates for the region, suggesting substantial N-fixation. The novel developmental dynamics outlined in this study point to the need for a new or expanded ecosystem developmental paradigm that better suits green roofs and perhaps other emerging engineered ecosystems.

虽然已经在许多自然系统中研究了生态系统的发展动态，但尚不清楚在组件未随时间共同发展的工程生态系统中是否期望这些模式。绿屋顶通常以营养丰富的基质开始，再与适应营养不良条件的生长缓慢的植物配对-可能导致与自然生态系统不同的发育动态。我们使用在瑞典南部设计类似但不同的老化绿色屋顶（跨越2到22年）的时间序列，评估了绿色屋顶养分含量和植被群落随时间的变化。底物深度，底物氮（N）和总氮池大小随屋顶使用年限呈正相关。这些动力学表明积累为2.9±1.1 g N / m ²/ yr，22年后没有平稳的迹象。植株氮含量（％）随着屋顶年龄的增加而呈正相关，但植物生物量，植物养分库和植物多样性均不随年龄变化。这些动态表明一种新的发展情景，即生态系统以接近稳定的植物生物量开始，但仍以与许多次生演替系统同等的速率在底物中积累氮。N的表观积累不能由该区域的N沉积速率解释，这表明存在大量的N固着。这项研究概述的新颖的发展动力表明，需要一种新的或扩展的生态系统发展范式，以更好地适应屋顶绿化以及也许其他新兴工程生态系统。