A positive relation with temperature increase was also observed
for nitrogen pollutants (Fig. 4c and d). As Table 4 shows, the
improvement in efficiency was lower compared to OM, with a mean
increase of 3–10% in performance. Significant differences were
found only for a few units (Table SM-2), e.g., for units W2 (planted
with cattails) and W10 (Type III). On the other hand, removal of TKN
loadings was slightly higher at low temperatures, while NH4
+-N
load removal was practically constant (Table 4), with simultaneous
lack of statistically significant differences.
Fig. 4e and f shows that phosphorus removal was favored at
higher temperatures; for OP, the percent removal increased up to
15% at high temperatures (Table 4). A statistically significant difference
was found only for unit W10 and for OP (Table SM-2). It has
to be mentioned that the performance of the unplanted unit W3
was practically not affected by temperature, while units W8 and
W9 showed a slight reduction at high temperatures. These latter
observations were made for both effluent concentrations and ALR
values.
The response of the wetland systems to temperature variations
has to do with various parameters. Microbial reactions, like
OM decomposition, nitrification and denitrification, are favored
at higher temperatures, resulting in lower effluent concentrations
[34]. Sorption of phosphorus is an endothermic reaction [35], which
means that low temperatures decrease the sorption capacity of the
bed, as observed by others [36]. Additionally, the nutrient uptake
by plants for their growth and function during the warm season
(spring–summer) also enhances the removal of nutrients at high
temperatures. Significant differences mainly appeared only for OM
removal with temperature variations. This probably means that the
operation regime of the VFCWs provides enough oxygen amounts
for the various processes throughout the year.
During the data processing, the temperature effect was tested
for each one of the three operation years separately. The analysis
showed statistically significant differences only for the first operation
year for OM and nitrogen and not during the other two. This:
(a) confirms the presence of a start-up effect during the first year;
and (b) indicates that the feeding regime, which includes resting
days, dominates on wetland performance, by ensuring adequate
oxygen transfer for BOD decomposition and ammonia oxidizing,
thus, limiting the temperature effect.