INTRODUCTION
At large spatial scales, a stream’s geomorphic and ecological form and behavior can be predicted based on longitudinal trends (e.g., Downing et al., 2012). From a geomorphic perspective, at the catchment scale, the concept of downstream hydraulic geometry (DHG) shows that several channel form parameters, including width and depth, increase as a power function of bankfull discharge (Leopold & Maddock, 1953), which in turn increases with drainage area. Due to predicted decreases in slope and thus shear stress, bed material grain size should also decrease (Gasparini et al., 1999). Within stream ecology, Vannote et al., 1980 built on the idea of downstream hydraulic geometry by extending the idea of longitudinal patterns or adjustments from the “physical system of geomorphologists into a biological analog”(Vannote et al., 1980), developing the river continuum concept (RCC). The RCC predicts longitudinal patterns in multiple stream ecosystem factors, including the type of macroinvertebrates present, the ratio of photosynthesis to respiration and the relative amount of coarse and fine particulate organic matter, all of which are strongly coupled to the predicted increase in channel width. Furthermore, biodiversity (species richness) often shows a predictable downstream increasing trend towards the middle of the catchment due to dispersal effects and more favorable growing conditions (e.g., Nilsson et al., 1989, Dunn et al., 2011, Kuglerová et al., 2015).
Since the establishment of the DHG and RCC concepts, exceptions to unidirectional
downstream trends have been described, including the geomorphic process domain concept (Montgomery, 1999), the serial discontinuity concept (Ward & Stanford, 1995), patch dynamics (Townsend, 1989; Poole, 2002), the river wave concept (Humphries et al., 2014), and the mighty headwater hypothesis (Finn et al., 2011). However, these concepts of downstream trends, implying longitudinal connectivity, still form the basis for multiple assumptions and concepts within river science (e.g., Surian, 2002; Frings, 2008; Dunne & Jerolmack, 2020). For example, the amount of allochthonous material input into a stream tends to decrease with increased channel width due to the greater percentage of the channel covered by canopy compared to larger channels (Vannote et al., 1980, Doi, 2009). To study these changes, rather than measuring channel width and canopy cover, studies commonly rely on drainage area as a proxy (e.g., Jonsson et al., 2018). If these basic longitudinal gradients do not exist in a certain catchment, then we must re-examine which geomorphic and ecological processes drive longitudinal stream patterns. The serial discontinuity concept, predicts local diversity to increase with increasing distance to dams (Ward & Stanford, 1995), and Green et al. (n.d.) found the same pattern for invertebrates but along a lake-stream system with lakes disrupting the continuity. However, geomorphic and ecological trends are often not examined simultaneously to explain organization of a stream network, even though they are intertwined with each other. Hence, the objective of this study was to examine whether geomorphic and ecological longitudinal trends, as defined by channel width and riparian vegetation richness and community composition, exist in a naturally disconnected catchment consisting of lakes and streams. By understanding the geomorphic and biotic longitudinal organization of a stream network, we can set expectations for recovery after stream restoration.