Hydrological high flow events, or episodes, are pulses of water that shape the pelagic food webs of lakes. These episodes promote hydrologically‐driven terrestrial inputs of dissolved organic matter (DOM) and nutrients (nitrogen (N) and phosphorus (P)) that influence lake productivity. At northern latitudes, episodes have typically occurred during spring snowmelt events (Andreásson, Bergström, Carlsson, Graham, & Lindström, 2004). However, the timing, frequency, and magnitude of spring flood episodes are changing in response to climate change (e.g., earlier occurrence of events with lesser frequency and magnitude; Andreásson et al., 2004; Creed, Hwang, Lutz, & Way, 2015) and are increasingly surpassed by extreme rainfalls in summer and early autumn (Creed, Hwang, et al., 2015; Min, Zhang, Zwiers, & Hegerl, 2011; Westra et al., 2014). In addition to changes in water delivery rhythm, northern lakes are facing hydrological intensification (e.g., wetter conditions in wet area, dryer conditions in dry areas; Huntington, 2006) which influence the delivery of terrestrial DOM and nutrients, and lake water retention times.

Change in magnitudes of water input (runoff), and its seasonal rhythm (episodes) affect whether lakes act as “shunts” (less resistance, faster flow of water shunted from land to lake, and through the lake, where the residence time and thus degree of processing of constituents in the lake water are reduced) or “chemostats” (more resistance, slower flow of water from land to lake, and through the lake, where the residence time and degree of processing of constituents in the lake water are increased). These terms have been used for streams and rivers (Creed et al., 2015; Raymond, Saiers, & Sobczak, 2016) but can be applicable to lakes as well. The potential consequences of the changing nature in water input may drive lakes from autochthonous (chemostats) to allochthonous (shunts) DOM dominance, or the opposite, with cascading effects throughout the pelagic food web. Here I will illustrate how the “shunt vs. chemostat concept” can be viewed in different time frames (e.g., annually or seasonally) and how this impacts the pelagic food web structure of northern lakes.

The annual turnover of water in lakes is defined by the theoretical water residence time (TWRT; lake volume divided by output water losses). In boreal landscapes, the TWRT is highly related to the catchment to lake area ratio (CA:LA) (Seekell et al., 2014; Kothawala et al., 2014) at least in a given region since CA correlate quite well with lake volume (cf. Sobek, Nisell, & Fölster, 2011). This means that the TWRT increases with decreasing CA:LA (i.e., short TWRT when CA and runoff is high relative to the lake size and its volume,). The CA:LA and the TWRT to a large extent determine lake DOM and nutrient concentrations in boreal lakes (Figure 1). From numerous lake studies in northern Sweden (see references in Figure 1) we know that short TWRT (high CA: LA) promotes shunt lakes where allochthonous DOM and nutrient concentrations are high, light conditions poor, with high bacterial (BP) to phytoplankton production (PP) ratios (BP:PP ratios) and pCO² concentrations, and where the phytoplankton community is dominated by flagellates (usually mixotrophs) capable of ingesting bacteria (Isaksson, Bergström, Blomqvist, & Jansson, 1999). In shunt lakes, the proportions of unselective filtering and bacteria feeding cladocerans (Vrede & Vrede, 2005) are further increased within the zooplankton community. With longer TWRT (lower CA:LA) lakes becomes increasingly more of a chemostat, with lower allochthonous DOM and nutrient concentrations, better light conditions, lower BP:PP ratios and pCO² concentrations, and increasingly higher phytoplankton‐ and autochthonous DOM production. In chemostat lakes, the proportions of larger autotrophs and autotrophic flagellates are further increased within the phytoplankton community, and calanoid copepods feeding selectively on phytoplankton (Sommer & Sommer, 2006) often dominate the zooplankton community. Thus, shifts from shunt to chemostat, which is highly linked to hydrology and TWRT, very much shape the pelagic food web structure in lakes.

The patterns recognized in Figure 1 are based on comparative studies among lakes when CA:LA and TWRT gradually change. However, our knowledge on how the pelagic food web structure in a given system, where CA:LA is constant, is impacted when the water delivery rhythm changes seasonally is much more limited. A question which also arises when comparing lakes is whether the change is gradual (as illustrated in Figure 1) or if there are thresholds in TWRT where lakes more suddenly shift from shunt to chemostat or the reverse. In a previous study we identified two “break‐off points” in TWRT (100 and 200 days) where the external control on lake bacterial communities dropped as a result of longer TWRT and reduced bacterial import (Lindström, Forslund, Algesten, & Bergström, 2006). More research is needed into whether similar and sudden break points also exist for separating shunt from chemostat lakes, and their associated pelagic food web structures, presumably by examining lakes distributed along gradients of increased TWRT and reduced terrestrial DOM and nutrient loading in regions where climate and N deposition is similar. We know that the vast majority of the Swedish lake population has TWRT between 1–100 days (78%), and very few have TWRT of >200 days (ca 12%) (Lindström et al., 2006). Thus, most Swedish lakes should act as shunts and be dominated by allochthonous DOM (cf. Kothawala et al., 2014). Weyhenmeyer, Norman, and Tranvik (2016) also modelled that a majority of Sweden's freshwaters will become browner following increased flushing of DOM from terrestrial systems due to a predicted 32% increase in precipitation to 2030 under the worst case climate scenario, and that the relative change in “browness” would be the most severe in lake with TWRT >6 years. This suggests that few lakes in the Swedish boreal landscape will shift from chemostats to shunts with cascading impacts on pelagic food web structure due to increased precipitation and terrestrial DOM and nutrient loading. Instead, the majority of the Swedish boreal lakes will become increasingly more “shunted.”

It could be that precipitation episodes and the seasonal rhythm of water flux in a given system impact whether lakes function as shunts or chemostats. If northern lakes have a biological memory that is “hardwired” to spring floods, the diminution of these events and more frequent interruptions by extreme summer rainfall episodes might create disruptions between supply and organism demands at different trophic levels in the pelagic food web. We can assess these seasonal changes in hydrology and their potential impacts on pelagic food web structure using data from Lake Örträsket and Lake Övre Björntjärn (see Figure 2 and associated references) which according to their TWRT (0.1 years, high CA:LA) can be classified as shunt lakes. Again, the hydrology will determine the magnitude of DOM and nutrient delivery to the lakes; however, in this case we will assess different rhythms of water, DOM and nutrient delivery (cf. Figure 2).

During a regular spring flood year, DOM and nutrient loading is linked to the size of the spring flood, which also determines the WRT of the epilimnion which in this case increases with number of days after the spring flood event. Over the summer the lake will shift from a shunt to a chemostat, with high BP:PP ratios in early summer caused by delivery of fresh DOM and limiting nutrients, where limiting nutrients in DOM primarily are allocated to bacteria due to their innate superiority in inorganic nutrient uptake compared to phytoplankton (Currie & Kalff, 1984). As the number of days after the spring flood event increases, easily degradable substrates in DOM are depleted and BP declines. Limiting nutrients are now allocated to phytoplankton causing increased PP and declining BP:PP. In late summer, PP will be higher than BP since phytoplankton can produce more carbon per limiting nutrient than bacteria (C:N:P stoichiometry by weight: phytoplankton 41:7:1 (Redfield, 1958) vs. bacteria 20:6:1 (Caron, Porter, & Sanders, 1990)). Total basal production (the sum of BP and PP; cf. Jansson, Karlsson, & Blomqvist, 2003) is therefore highest in late summer when PP dominates during periods of long epilimnion WRT and when the lake starts to function as a chemostat. This sets up special conditions for phytoplankton development, where mixotrophs dominate the phytoplankton community in connection with high BP, which switches to higher proportions of autotrophs later in summer (not illustrated in Figure 2). Zooplankton community composition and their resource use is very much linked to this pattern in basal production and the phytoplankton community composition, with larger proportions of cladocerans of high allochthony in early summer, and greater proportions of selectively feeding calanoids of high autochthony in late summer (Berggren, Ziegler, St‐Gelais, Beisner, & del Giorgio, 2014; Karlsson et al., 2012).

Under a situation of an earlier spring flood of much lower magnitude, DOM and nutrient delivery will be reduced and the epilimnion TWRT will increase, promoting a longer chemostat period over summer. The BP in connection to the earlier spring flood will be lower and of shorter duration. The PP will start earlier and extend for a longer period of time over summer. The paradox of this is that the total basal production might not be much lower compared to the spring flood event, despite lower DOM and nutrient input, since limiting nutrients in this case are allocated to phytoplankton during a longer period of time in summer. These changes are also reflected in the phytoplankton‐ and zooplankton community composition, which will have high proportions of autotrophs and calanoid copepods, respectively, throughout the summer.

Under a situation of a spring flood and a summer rainfall episode, BP will increase again after the rainfall episode due to delivery of fresh DOM and nutrients. The PP will be pushed, occurring after the rainfall episode later in summer over a shorter duration, and with a higher peak as caused by greater DOM and nutrient delivery. Total basal production will be higher with a substantial proportion performed by bacteria. Hence, the lake will be shunted several times, and chemostatic behaviour when PP production dominates will occur late in summer with reduced duration. The phytoplankton and the zooplankton community will return to high proportions of mixotrophs and cladocerans after the summer rainfall episode, and autotrophs and calanoids will become more abundant late in summer in connection to the PP peak.

Under a situation of a spring flood followed by frequent smaller rainfall episodes, such as under a rainy summer, the lake will be shunted throughout the summer, with BP production exceeding PP which will be very low. The phytoplankton‐ and zooplankton communities will be dominated by mixotrophic flagellates and cladocerans, respectively. An important aspect to consider is that spring floods are caused by snow melt which does not affect incoming light, whereas high flow episodes in summer are related to rainfall events when light conditions usually are poor. Hence, protracted summer rainfall events will have a twofold negative impact on PP, by increasing competition for limiting nutrients and reducing the light conditions in lake water. However, warming and consistent trends toward reduced lake ice‐cover (Magnuson et al., 2000) combined with earlier and smaller spring flood events (Andreásson et al., 2004; Creed, Hwang, et al., 2015) may on the other hand improve light conditions in lakes promoting enhanced PP (Weyhenmeyer, Westoo, & Willen, 2008).

Hence, data from these two shunted lakes (Figure 2) suggest close cascading effects between the rhythm of the water, the DOM and nutrient delivery, and the structure of the pelagic food web. As expected small organisms with short generation times (bacteria, flagellates) respond quickly to changes in epilimnetic water retention times and are benefitted relatively to larger organism with longer generation times (larger phytoplankton) when lakes are being shunted. Thus, moving from spring floods to situations of spring floods of low magnitudes, whether or not one or more summer rainfall episodes occur, are not causing “collapses.” Instead, strong changes in timing and magnitudes in BP and PP production occur in summer, linked to DOM and nutrient delivery which zooplankton quickly adapt to. Since shunt lakes usually are humic (Figure 1), with low benthic algal production (Ask et al., 2009), it remains to be evaluated how these changes in the seasonal rhythm of water input and DOM and nutrient delivery, which shape the pelagic food web structure (Figure 2), might impact fish which rely increasingly more on pelagic food resources when lake DOM concentrations increase (Karlsson et al., 2009). Although the quantity of basal resources (BP + PP) might not change much for zooplankton when comparing a summer with a low spring flood vs. a summer with a distinct spring flood event, it will be very different in quality. For example, bacteria lack essential fatty acids (Müller‐Navarra, 2008) and are not fed upon by calanoid copepods (Sommer & Sommer, 2006). A bacterial based pelagic food web is longer and of lower trophic transfer efficiency compared to a phytoplankton based food chain (Jansson, Persson, De Roos, Jones, & Tranvik, 2007). The functionality of the lakes will further be impacted, with transition between CO₂ or biomass producers driven by changes in hydrology. The relative importance of internal nutrient recycling (Levine & Schindler, 1992; Sterner & Hessen, 1994) and external nutrient loading for phytoplankton nutrient limitation will also be impacted (Bergström, Karlsson, Karlsson, & Vrede, 2015; Downing & McCauley, 1992), regardless of whether a lake's hydrological connection with its catchment is enhanced or suppressed.

We need to assess to what extent these patterns and changes in pelagic food web structure recognized for northern Swedish lakes (Figures 1 and 2) are applicable to lakes in other areas within and beyond the boreal zone. Lakes are exposed to different hydrology and seasonal rhythms of water, and supplied with different quantities and qualities of DOM (DOM: nutrient stoichiometry, aromaticity) that will impact pelagic food webs. Annual water retention times have been shown to be negatively related to DOM turnover across inland waters (Catala'n, Marce', Kothawala, & Tranvik, 2016; Zwart, Sebestyen, Solomon, & SE., 2017), and extreme precipitation episodes have been recognized as “hot moments” of carbon turnover that promote heterotrophy in lakes (Zwart et al., 2017) similar to what has been shown for humic Lake Örträsket (Bergström, 2009; Bergström & Jansson, 2000; Drakare et al., 2002; Jansson, Bergström, Blomqvist, Isaksson, & Jonsson, 1999). However, the extent to which TWRT can be used to identify a lake's pelagic food web structure, whether thresholds in TWRT that account for shifts between chemostatic and shunting behaviour are similar between arctic, boreal and temperate regions, and the precise form of such shifts remain unknown.

If we are facing smaller spring floods and more frequent summer rainfall episodes, then we also need to assess the cascading effects on pelagic food web structures on a seasonal basis, and evaluate if changes in “hot moments” in what we would regard as shunt lakes promote similar impacts and shifts in pelagic food web structure when assessing lakes at local, regional, national or continental scales. In northern landscapes, external DOM and nutrient delivery to lakes is highly related to terrestrial primary production and hydrology (Hope, Billet, & Cresser, 1994; Jansson, Hickler, Jonsson, & Karlsson, 2008). However, changes in hydrology (promoted by precipitation increases) in areas with similar terrestrial primary production may also cause dilution of lake DOM concentrations but at the same time increase lake water retention times (Larsen, Andersen, & Hessen, 2011). Much less research has been focused on the specific impact of lake water retention time on pelagic food webs relative to the role of changes in lake DOM. This should be possible to assess in areas with large precipitation gradients and little change in terrestrial landscape cover. When these questions have been resolved, then we may advance even further in applying the “shunt to chemostat” concept to lakes and separate the impacts of hydrology from those associated with the quantity and quality of DOM delivery on pelagic food web structure and lake productivity.

水文高流量事件或情节是水脉冲，它们塑造了湖泊中上层食物网。这些事件促进了水文驱动的溶解有机物（DOM）和营养物质（氮（N）和磷（P））的陆地输入，从而影响了湖泊的生产力。在北纬，春季融雪事件通常会发生（Andreásson，Bergström，Carlsson，Graham和Lindström，2004年）。但是，春季洪水发生的时间，频率和强度会随着气候变化而变化（例如，发生频率和强度较小的事件越早发生；Andreásson等人，2004； Creed，Hwang，Lutz和Way，2015年）），并且在夏季和初秋的极端降雨中越来越多（Creed，Hwang，et al。，2015 ; Min，Zhang，Zwiers，＆Hegerl，2011 ; Westra et al。，2014）。除了供水节奏的变化外，北部湖泊还面临着水文强度的加剧（例如，湿润地区的湿润条件，干旱地区的干燥条件； Huntington，2006年），这会影响陆地DOM和养分的输送以及湖泊保水时间。

入水量的变化（径流）及其季节节律（短片）会影响湖泊是否充当“分流器”（阻力较小，水从陆地流向湖泊，并流经湖泊的水流更快，在这里停留时间较长，因此减少了湖水中成分的处理程度）或“化学稳定剂”（更大的阻力，水从陆地流向湖泊以及流经湖泊的速度较慢，增加了湖水中成分的停留时间和处理程度） ）。这些术语已用于溪流和河流（Creed等人，2015年; Raymond，Saiers和Sobczak，2016年），但也适用于湖泊。水输入性质变化的潜在后果可能将湖泊从土质（化学稳定剂）变成异质（分流器）的DOM优势，或相反，在整个中上层食物网中连锁反应。在这里，我将说明如何在不同的时间范围内（例如，每年或季节性）观察“分流与化学稳定概念”，以及这如何影响北部湖泊的中上层食物网结构。

湖泊中水的年周转量由理论水停留时间（TWRT；湖泊体积除以产出水损失）定义。在北方景观中，TWRT至少与给定区域中的流域与湖泊面积比（CA：LA）高度相关（Seekell等，2014； Kothawala等，2014），因为CA与湖泊量有很好的相关性（请参阅Sobek，Nisell和Fölster，2011年）。这意味着TWRT随着CA：LA的减小而增加（即，当CA和径流相对于湖泊的大小和体积较大时，TWRT较短）。CA：LA和TWRT在很大程度上决定了北部湖泊中的湖泊DOM和养分含量（图1）。从瑞典北部的众多湖泊研究（参见图1中的参考文献）中，我们知道，短TWRT（高CA：LA）会促进分流湖，其中异源DOM和养分浓度高，光照条件差，细菌（BP）到浮游植物产生高（PP）比（BP：PP比）和p CO ²浓度，以及浮游植物群落以能够摄取细菌的鞭毛虫（通常是混合营养菌）为主导的地区（Isaksson，Bergström，Blomqvist和Jansson，1999年））。在分流湖中，浮游动物群落中非选择性过滤和细菌喂食枝角类鱼类的比例进一步增加（Vrede＆Vrede，2005）。随着更长的TWRT（较低的CA：LA），湖泊变得越来越趋于稳定，异源DOM和养分浓度降低，光照条件更好，BP：PP比率和p CO ²浓度降低，浮游植物和自生DOM的产量也越来越高。在chemostat湖泊中，浮游植物群落中较大的自养生物和自养鞭毛虫的比例进一步增加，并且cal足类pe足类以浮游植物为食（Sommer＆Sommer，2006）通常主导浮游动物群落。因此，与水文和TWRT高度相关的从分流向化学稳定的转变，极大地影响了湖泊中上层食物网的结构。

图1中识别的模式是基于CA：LA和TWRT逐渐变化时湖泊之间的比较研究。但是，当给水节奏随季节变化时，我们对CA：LA恒定的给定系统中上层食物网结构的影响的知识更加有限。比较湖泊时，还会出现一个问题，即变化是否是渐进的（如图1所示），或者TWRT中是否存在阈值，在该阈值中，湖泊更突然从分流转变为恒化器或反向。在先前的研究中，我们确定了TWRT（100天和200天）中的两个“突破点”，由于更长的TWRT和减少的细菌进口，对湖泊细菌群落的外部控制下降了（Lindström，Forslund，Algesten和Bergström，2006年）。需要进一步研究是否也存在类似和突然的断裂点，以将分流器与恒化器湖及其相关的中上层食物网结构分开，大概是通过研究沿气候和气候变化地区TWTW增加和陆地DOM和营养负荷减少的梯度分布的湖泊来进行的。 N沉积是相似的。我们知道，绝大多数瑞典湖泊人口的TWRT在1-100天之间（78％），而很少有TW​​RT> 200天（约12％）（Lindström等，2006）。因此，大多数瑞典湖泊应作为分流器，并以异源DOM为主（参见Kothawala等，2014）。Weyhenmeyer，Norman和Tranvik（2016）还模拟，由于在最坏的气候条件下，到2030年预计会有32％的降水增加，因此随着陆地系统对DOM的冲刷增加，瑞典的大部分淡水将变得更棕色，并且“粗度”的相对变化将是TWRT> 6年的湖泊中最严重。这表明，由于降水量和陆地DOM和养分含量的增加，瑞典北方景观中很少有湖泊会从恒化器转变为分流器，从而对远洋食物网结构产生连锁影响。相反，大多数瑞典的北方湖泊将变得越来越“分流”。

给定系统中的降水事件和水通量的季节性变化可能会影响湖泊是作为分流器还是化石器。如果北部湖泊具有与春季洪水“硬连线”的生物记忆，则这些事件的减少以及夏季极端降雨事件造成的更频繁的干扰可能会在中上层食物网中的不同营养水平上造成供应与生物体需求之间的干扰。我们可以使用Örträsket湖和ÖvreBjörntjärn湖的数据（见图2和相关参考资料）评估这些水文学的季节性变化及其对中上层食物网结构的潜在影响，根据它们的TWRT（0.1年，CA：LA高）可以得出。被列为分流湖。同样，水文将决定DOM的数量和向湖泊的营养输送。然而，

在常规的春季洪水年中，DOM和养分含量与春季洪水的大小有关，这也决定了epi虫的WRT，在这种情况下，随春季洪水发生的天数增加。在夏季，由于新鲜DOM和限制养分的输送，夏季初夏时分，湖泊将由分流器转变为恒化器，BP：PP比率较高，其中DOM中的限制性养分主要由于细菌固有的无机优势而主要分配给细菌。与浮游植物相比养分吸收（Currie＆Kalff，1984）。随着春季洪水事件后天数的增加，DOM中易于降解的底物将被消耗，而BP下降。现在将有限的营养物分配给浮游植物，导致增加的PP和降低的BP：PP。在夏季末期，PP会比BP高，因为浮游植物的极限养分比细菌产生的碳更多（细菌的化学计量比为C：N：P：浮游植物41：7：1（Redfield，1958年），细菌比为20：6：1。 （Caron，Porter，＆Sanders，1990）。基础总产量（BP和PP的总和；请参阅Jansson，Karlsson和Blomqvist，2003年因此，在夏季晚些时候，PP占主导地位，并且湖泊开始起化学稳定作用时，PP在夏季末期占主导地位。这为浮游植物的发展创造了特殊条件，其中混合营养生物与高BP一起主导着浮游植物群落，在夏季晚些时候转换为更高比例的自养生物（图2中未显示）。浮游动物群落组成及其资源利用与基础生产和浮游植物群落组成的这种模式密切相关，在夏季初，较高比例的高别拟金鱼的锁骨，在夏季末选择性地喂养高自养率的花色素类动物（Berggren ，齐格勒，圣格莱，贝斯纳和德尔·乔尔乔，2014年；卡尔森等人，2012年）。

在较早的春季洪水的情况下，将减少DOM和养分的输送，并增加epi虫的TWRT，从而延长夏季的化学稳定期。与较早的春季洪水有关的BP将较低，持续时间较短。PP将更早开始，并在夏季延长更长的时间。矛盾的是，尽管DOM和养分输入较低，但总基础产量与春季洪水事件相比可能不会低很多，因为在这种情况下，在夏季较长的时间内将有限的养分分配给浮游植物。这些变化也反映在浮游植物和浮游动物的群落组成中，在整个夏季，它们将分别具有高比例的自养生物和cal类co足类动物。

在春季洪水和夏季降雨的情况下，由于新鲜DOM和养分的输送，BP在降雨之后将再次增加。PP将被推动，发生在夏季晚些时候的降雨之后，持续时间较短，并且由于较高的DOM和养分输送而导致峰值较高。总基础产量将更高，其中大部分由细菌完成。因此，该湖将被分流数次，而当PP生产占主导地位时，其化学趋向性将在夏季末发生，持续时间缩短。在夏季降雨之后，浮游植物和浮游动物群落将恢复到高比例的混合营养生物和枝形螯虾，并且随着PP峰的到来，夏季后期自养生物和类钙质将变得更加丰富。

在春季洪水泛滥而降雨频繁的情况下，例如在一个多雨的夏天，整个夏天湖将被分流，BP的产量超过PP，这将非常低。浮游植物和浮游动物群落将分别由混合营养鞭毛虫和钩角藻为主。需要考虑的一个重要方面是春季洪水是由融雪引起的，不会影响入射光，而夏季的高流量事件与通常在光照条件较差的降雨事件有关。因此，旷日持久的夏季降雨事件将增加限制营养素的竞争并减少湖水的光照条件，从而对PP产生双重负面影响。然而，变暖和减少湖冰覆盖的一致趋势（Magnuson等，2000）加上早期和较小的春季洪水事件（Andreásson等人，2004 ; Creed，Hwang等人，2015）另一方面可能会改善促进PP的湖泊的光照条件（Weyhenmeyer，Westoo，＆Willen，2008） 。

因此，来自这两个分流湖的数据（图2）表明，水的节律，DOM和养分输送与中上层食物网的结构之间具有密切的级联效应。正如预期的那样，具有较短生成时间的小型生物（细菌，鞭毛）对表皮水滞留时间的变化做出快速响应，并且在分流湖泊时，具有较长生成时间的大型生物（较大的浮游植物）相对受益。因此，从春季洪水转向低强度春季洪水的情况，无论是否发生一个或多个夏季降雨事件，都不会造成“崩溃”。取而代之的是，夏季BP和PP生产的时间和幅度发生了很大的变化，这与浮游动物迅速适应的DOM和养分供应有关。由于分流湖通常很腐殖质（图1），2009）。尽管在比较春季洪水较少的夏季与春季洪水明显的夏季时，浮游动物的基础资源量（BP + PP）可能不会有太大变化，但质量会大不相同。例如，细菌缺乏必需脂肪酸（Müller-Navarra，2008年），而没有被类co足类足类动物摄食（Sommer＆Sommer，2006年）。与基于浮游植物的食物链相比，基于细菌的远洋食物网更长且营养传递效率更低（Jansson，Persson，De Roos，Jones和Tranvik，2007年）。随着CO ₂之间的过渡，湖泊的功能将进一步受到影响。或受水文学变化驱动的生物质生产者。内部营养物再循环（Levine＆Schindler，1992； Sterner＆Hessen，1994）和外部营养物负荷对浮游植物营养限制的相对重要性也将受到影响（Bergström，Karlsson，Karlsson和Vrede，2015； Downing和McCauley，1992）。 ，无论湖泊与集水区之间的水文联系是否得到加强或抑制。

我们需要评估瑞典北部湖泊（图1和图2）公认的中上层食物网结构的这些模式和变化在多大程度上适用于北方区域内外的湖泊。湖泊面临着不同的水文学和季节性的水律，并提供了不同数量和质量的DOM（DOM：营养化学计量学，芳香性），这将影响远洋食物网。已证明年度保水时间与内陆水域的DOM转化成负相关（Catala'n，Marce'，Kothawala和Tranvik，2016年; Zwart，Sebestyen，Solomon和SE。，2017年），并且极端降水事件有被认为是促进湖泊异养性的碳转换的“热门时刻”（Zwart等，2017））类似于腐殖酸的厄特勒斯凯特湖（Bergström，2009 ;Bergström＆Jansson，2000 ; Drakare et al。，2002 ; Jansson，Bergström，Blomqvist，Isaksson，＆Jonsson，1999）。但是，TWRT可用于识别湖泊中上层食物网结构的程度，TWRT中解释化学作用和分流行为之间变化的阈值在北极，寒带和温带地区之间是否相似，并且这种变化的精确形式仍然存在未知。

如果我们面对的是春季洪水较少和夏季降雨频繁的情况，那么我们还需要评估季节性对中上层食物网结构的级联效应，并评估所谓“分流湖”中“炎热时刻”的变化是否在评估地方，区域，国家或大陆规模的湖泊时，促进中上层食物网结构的类似影响和变化。在北部景观中，外部DOM和向湖泊的养分输送与陆地初级生产和水文学高度相关（Hope，Billet和Cresser，1994； Jansson，Hickler，Jonsson和Karlsson，2008）。）。但是，在与陆地初级生产相似的地区，水文变化（受降水增加的推动）也可能导致湖泊DOM浓度降低，但同时增加了湖泊保水时间（Larsen，Andersen和Hessen，2011年））。相对于DOM改变的作用，很少有研究集中在湖泊保水时间对中上层食物网的具体影响上。在降水量梯度大且陆地景观覆盖变化很小的地区，应该可以进行评估。解决了这些问题后，我们可以进一步在湖泊中应用“将化粪池分流”的概念，并将水文学的影响与与DOM传递的数量和质量对中上层食物网结构和湖泊生产力的影响分开。