The DDG drivers
The drivers of the macrophyte DDGs strongly differed for DDG measures
and richness components. Whereas pairwise correlations detected many
strong relationships across richness components, multiple models
revealed significant variables only for DDG measures of alpha richness.
The Rmax correlates of the different richness
components with a very similar set of abiotic parameters. AllRmax correlates with area. This reflects
species-area relationships (SARs) (Connor and McCoy 1979, Lomolino 2000,
Qian et al. 2007, Patiño et al. 2014) for macrophytes, which is also
shown by high correlation of area with total gamma richness. Looking at
the GAMM results for non-linear responses R(β,max) and R(γ,max) did not show any significant
results, but R(α,max) is exclusively influenced
by the “SiO2 & Conductivity axis” (PC1) (Fig.
3). Area, spectral absorption coefficient and water level
fluctuations (WLF) also have a high contribution to PC1 and thus,area might be the key driving force again. The SAR of macrophytes
was already shown in a several studies (Alahuhta et al. 2020), still a
study comparing SAR of macrophytes with terrestrial plants would be very
informative. Here, it would be interesting to add information
about lake bathymetry, as lake area is just a proxy for the colonisable
area per depth. However, it was not shown yet that the size of lakes
also influences the shape of DDG.
The D(β,max) and D(γ,max) could not be explained with abiotic variables, neither by correlations
nor by a GAMM. Unlike D(α,max) , the gamma, and
consequently beta, values along DDG are more variable, indicating
spatial heterogeneity and possibly unsaturation (Karger et al. 2014).
Still, D(α,max) correlates positively withphosphorus and temperature_sd , and negatively withO2 dissolved and transparency .
Furthermore, looking at non-linear influences, theD(α,max) is affected by all four PCA-axes.
The PC2 (Temperature & Ptot axis ) shows the
highest influence (Fig. 3). This means that in lakes with highphosphorus concentration and/or high temperature the DDG
peaks in shallower waters. Phosphorus is the limiting factor for
phytoplankton growth and phytoplankton reduces the light availability
for macrophytes. In contrast for macrophytes, it is still debatable
whether the phosphorus concentration in the water is a limiting
growth factor (Carr et al. 1997). One important point to consider in
this debate is that rooted submerged macrophytes can also take up
nutrients from the sediments (Lacoul and Freedman 2006). Hence,phosphorus might affect macrophytes by promoting phytoplankton
growth, which then reduces light availability and shifts DDG to
shallower depths. Besides phosphorus , temperature is a
major factor influencing metabolic processes as photosynthesis and
respiration. Additionally in lakes, higher temperature result in
higher nutrient levels due to increased mineralization and internal
fertilization processes (Moss 2012). Internal fertilization processes
occur when higher water temperatures lead to increased layering
stability, prolonged oxygen consumption, anoxia in deep waters,
resulting in anoxic resuspension of phosphorus from the lake
sediments. These resuspended nutrients promote phytoplankton growth,
thus reducing light for macrophytes. Therefore, the PC2
(Temperature & Ptot axis ) describes the
productivity gradient in lakes, caused by lower light availability
leading to a shallower maximum of species richness.
Besides light quantity, light quality also influencesD(α,max) , which is indicated by the influence of
the PC4 (O2diss– SAC axis ). With highO2diss content and low spectral absorption
coefficient at 254nm (SAC, a measure of coloured dissolved organic
matter – CDOM ) we observe richness peaks at deeper waters. On
the one hand, CDOM reduces damaging UV-B radiation. On the other
hand, it reduces light availability. Thus, we see a diametrically
opposed effect of light quantity and light quality which might
contribute to the prevailing pattern of highest species richness at
medium depth level. In general, if light resource represents the main
component of productivity in lakes, the mid-depth DDG might follow the
intermediate productivity hypothesis (VanderMeulen et al. 2001,
Rajaniemi 2003, Huston 2014).
Besides light, also temperature seems to influenceD(α,max) , via surface water temperature and its
influence on light availability (as explained above) and via the lake’s
layering depth. This second mechanism by which temperature layering
affects DDG is demonstrated by that along PC3 (Tempsd – Chloride
axis ) D(α,max) decreases. A hightemperature_sd (shallow epilimnion – the upper temperature
layer in a stratified lake) promotes a shallowD(α,max) , while a lowtemperature_sd (broad epilimnion) allows deeperD(α,max) . Temperature_sdis positively correlated to temperature demonstrating that higher
temperatures can lead to a shallower upper warm layer in water bodies as
the stratification is more stable (Adrian et al. 2009).
The weakest influencing effect (lowest drop contribution) is provided by
PC1 (NH4N – SiO2 & Conductivity axis ). Just at very high values
of PC1 that D(α,max) becomes shallower. Asconductivity is negatively correlated with transparency(cor=-0.71, p <0.001), we speculate that also here
transparency is the actual mechanism that influencesD(α,max) .
In summary, the main influences on D(α,max) seem
to be as expected factors of water quality that influence light quantity
(transparency, influenced by phosphorus and temperature), light quality
(CDOM) and layering depth (temperature). The main influence onR(α,max) is the lake surface area.