Towards deconfounding hydrological and seasonal temperature variability in the determination of selected limnological variables of a temperate floodplain ecosystem

Abstract

This paper presents a methodological approach that can aid the assessment of variation in limnological variables in response to hydrological phases and seasonal temperature variation. The investigation was conducted in the Danube River floodplain, Kopački Rit, Croatia. The proposed approach facilitates a more accurate examination of the effects of first factor, while maintaining second factor less variable, and vice versa. Samples were collected in the floodplain channel and lake along a temporal connectivity gradient, and were grouped according to three hydrological phases (isolation, flow pulse and flood pulse) and according to three seasons (spring, summer and autumn). A permutational multivariate analysis of variance was used to analyse the hydrological effects within seasons, as well as seasonal effects within hydrological phases on selected limnological variables: pH, conductivity, dissolved oxygen, nitrites, nitrates, ammonium, total Kjeldahl nitrogen, total phosphorus, chlorophyll a concentrations and ratio of copiotrophic to oligotrophic bacteria. A multivariate data cloud was structured by hydrological phases during all investigated seasons, but not by all seasons during different hydrological phases. The strongest impact of hydrological phases on limnological variables occurred during summer. Seasonal effects were significant during the isolation and the flood pulse phases. The flow pulse phase somehow obscured the effects of seasonal change, resulting in temporal floodplain homogenization. One possible explanation is that water temperature determines potential limnological conditions, whereas hydrology determines actual conditions. Such results must be taken into account if the accurate effects of hydrology and seasonal change are to be assessed in temperate floodplain ecosystems. Copyright © 2014 John Wiley & Sons, Ltd.

Effect of depth on food‐web interactions in a thermally stratified floodplain lake following inundation

“…By ignoring the interaction, it could be easily and inaccurately concluded that the hydrological effects were statistically significant at both sample layers, although they were, in reality, significant only in the surface‐water layer. The interactive effects of hydrology and water temperature on environmental variables have already been noted on a broader spatial and temporal scale in Kopački Rit floodplain where the strongest hydrological impact on ecological variables occurred during the warmest part of the year (Palijan, ). Throughout our study, notable hydrological influence appeared in the warmer water layer.…”

“…10–15 °C) occurring from April to September. Thermal stratification usually lasts for 2 to 4 months (Palijan, ). A smaller difference (∆t ca.…”

“…5 °C) persists during the rest of the warmer part of the year. Minor variability in temperature between the surface‐ and bottom‐water layers (∆t up to 1 °C), or the absence of it, occurs during the colder part of the year, from late autumn to early spring (Palijan, ). Throughout stratification, the decrease in water temperature is continuous towards the lake bottom.…”

“…The hydrological phases are as follows: Danube water level at the Apatin gauge was (a) <1.67 m = isolation; (b) 1.67–3 m = flow pulse; and (c) >3 m = flood pulse. On average, each hydrological phase cumulatively lasts for one‐third of the year (Palijan, ), with the highest water levels reached during spring and early summer and the lowest during autumn. The flood pulse in Kopački Rit usually reaches 2 m above the flooding threshold water level, and only in extreme cases, 5 m above it.…”

“…The number of samples from the surface and bottom from each station was 11, of which five was collected during flowing (flow pulse n = 5) and six during flooding conditions (flood pulse n = 6). The investigated period was chosen based on previous results in the occurrence of thermal stratification of the lake, which regularly occurs during late spring and summer (Palijan, ). The sampling ended after the bottom‐water layer temperature had reached the surface‐water values from the beginning of the investigation.…”