Introduction

Describing and explaining biodiversity gradients have been central goals of biogeography and ecology since the beginning of the respective fields (Gaston 2000). Improving our understanding of the biodiversity gradients and their drivers is still an important requirement to deal with impending species loss. Therefore, many studies have explored environmental gradients as explanatory variables for biodiversity patterns along different geographic scales (Rahbek 2004, Whittaker et al. 2007) like (1) latitude (Stehli et al., 1969; Rohde, 1992; Pontarp et al., 2019; Etienne et al. 2019), (2) elevation (Hutchinson 1953, Rahbek 1995, Lomolino 2001, Nogués-Bravo et al. 2008, Colwell and Rangel 2010, Sanders and Rahbek 2012, Graham et al. 2014, Rahbek et al. 2019), (3) tree height in forests (Petter et al. 2016), (4) depth in soils (Rendoš et al. 2016, Jakšová et al. 2019) or (5) depth in water (Rex and Etter 1998, Smith and Brown 2002, Gong et al. 2015). These geographical gradients share some environmental gradients, which are expected to influence spatial structuring of diversity gradients, e.g. temperature, light or seasonality. However, the shorter spatiotemporal scales are, the less confounding biogeographical contingencies there are, such as the legacy of the glacial cycles on latitudinal gradients, and dispersal/connectivity limitations. Hence, studying gradients expressed at short spatiotemporal extents may provide valuable insights on drivers of biodiversity. On the one hand, the best studied biodiversity gradients are the large spatiotemporal gradients like the decreasing diversity with latitude (Pontarp et al. 2019) and the decreasing or hump-shaped diversity with elevation (Rahbek 1995). On the other hand, the short spatiotemporal gradients, like depth in freshwaters, are often overlooked.
Freshwater ecosystems have a high biodiversity, but also show a high rate of species loss (Strayer and Dudgeon 2010, Gatti 2016, He et al. 2017), exceeding those of terrestrial systems (Dudgeon et al. 2006). Despite this, studies that focus on the diversity gradient in freshwater are surprisingly scarce, although light gradients in freshwater must represent a very strong driver. The few studies seem to show predominantly a general decrease of biodiversity along the depth gradient, e.g. for bacteria (Cantonati et al. 2014, Zhao et al. 2019), Chironomids (Zhao et al. 2019) or diatoms (Kingsbury et al. 2012, Stoof-Leichsenring et al. 2020), or hump-shaped patterns along depth, e.g. for diatoms (Zhao et al. 2019) or submerged macrophytes (Ye et al. 2018). Thus, it is not clear if the patterns are generalizable for different species groups and even across different lakes.
Macrophytes play a pivotal role in lakes by reducing nutrient concentrations (Song et al. 2019), by providing food for a lot of other species (Bakker et al. 2016) and by giving shelter to a large number of other aquatic organisms like zooplankton, juvenile fish and amphibians (Jeppesen et al. 1998). However, there are several knowledge gaps on macroecology of freshwater plants (Alahuhta et al. 2020) and especially the depth pattern of submerged macrophytes is sparsely studied and remains unclear (Fu et al. 2014a, b, Ye et al. 2018). The few studies that have looked at depth distribution of macrophytes in lakes focussed on Lake Erhai in Yunnan Province, China. They report a hump-shaped pattern along the water depth gradient for species richness and community biomass of submerged macrophyte species (Ye et al. 2018). Looking at all functional types including emergent species, Lake Erhai shows a significant decrease in taxonomic and functional diversity along the water depth gradient and its niche differentiation (Fu et al. 2014a, b). Hence, it remains unclear if the described pattern is generalizable and whether it stays robust over time.
The lack of studies is intriguing because the environmental gradients along lake depth represents one of the sharpest found in nature, with strong variation in just few meters. With increasing lake depth, multiple abiotic factors that influence the growth of macrophytes (light, temperature, nutrients, water quality, disturbances/hydrologic variability) drastically change (Bornette and Puijalon 2011). Light is gradually attenuated with increasing depth due to absorption and scattering, resulting in a specific reduction of light quality and quantity depending on depth and on the water turbidity. Water temperature in deep lakes does not decrease gradually, but rather abruptly with depth (Bornette and Puijalon 2011). The formation of thermally stratified lakes results in an abrupt thermocline, especially during growing season. The thermocline influences the within-lake fluid dynamics in each thermal layer, further leading to stratified gradients in nutrients and water quality components during stratification (Bornette and Puijalon 2011). Moreover, mechanical disturbances, like wind or waves, lose their influence gradually with depth (Van Zuidam and Peeters 2015). Also the probability that water level fluctuations result in drying up the soil (Evtimova and Donohue 2016) is reduced and browsing pressure by water fowl decreases with depth (Bakker et al. 2016). How these different environmental gradients influence the species richness of macrophytes stays unclear, although knowing the processes shaping species diversity might help to predict how global change will affect biodiversity and how management strategies might mitigate potential negative diversity responses.
This study aims to describe the depth distribution of macrophyte diversity and assessing the relative importance of its drivers. Specifically, we ask the following questions:
  1. 1.1. What is the general shape of the depth diversity gradient (DDG) of submerged macrophytes in deep lakes? 1.2. Are there differences between lakes and diversity components (alpha-, beta, gamma richness)?
  2. What are the drivers for macrophyte DDG?
  3. 3.1. Has the DDG being stable over recent years? 3.2. Are temporal trends general or lake-specific?
To address these questions, we use the macrophytes occurrence data of 274 transects taken along 13 years across 28 natural deep lakes in Bavaria, that were mapped for the monitoring of the European Water Framework Directive. We expect a hump-shaped DDG (question 1.1) corresponding to previous punctual empirical evidence and following the typical patterns found along elevation. We assume no broad differences between lakes and diversity components as the pattern is supposed to be generalisable (question 1.2). To tackle question (2), we test whether the shape of the DDG can be explained by geographic and physical-chemical characteristics of the lakes. We expect water quality to have a high degree of influence, since water quality influences resource availability (light, temperature). Finally, we assess for question 3.1 -3.2 whether there have been detectable temporal changes in the DDG. We suppose that the DDG is a quite stable pattern over time as macrophytes react slowly to changes (Bakker et al. 2013). However, due to the overall warming in annual average water temperature during the last decades we expect that species richness increases, as invasive species are expected, and warm-adapted species might expand. Our results provide the most refined and extensive assessment of macrophyte biodiversity patterns in freshwater lakes up to date, giving insights for the development of long-term conservation strategies for freshwater systems in general.