- Email: [email protected]
The dynamics of freshwater phytoplankton stability in the Naroch Lakes (Belarus)
Ecological Indicators 81 (2017) 481–490
Contents lists available at ScienceDirect
Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind
Original Articles
The dynamics of freshwater phytoplankton stability in the Naroch Lakes (Belarus)
MARK
⁎
T.M. Mikheyevaa, A. Parparovb, , B.V. Adamovicha, G. Galb, E.V. Lukyanovaa a b
Belarusian State University, Biological Department, 220030, Nezavisimosti Ave., 4, Minsk, Belarus Kinneret Limnological Laboratory, Israel Oceanographic and Limnological Research, POB 447, Migdal, 14950, Israel
A R T I C L E I N F O
A B S T R A C T
Keywords: Naroch lakes Lake Kinneret Phytoplankton community Ecological stability Trophic status Principle of emergence
The phytoplankton community of the Belarus Lakes Naroch, Myastro and Batorino, which have a Trophic State Index of 42.3, 60.7 and 66.8, respectively, underwent drastic changes to their structure during the period between 1968 and 2012. Thanks to an extensive monitoring program, these changes were well-documented and were qualitatively interpreted as signs of the community destabilization. The main objective of this study was the quantification of the ecological stability of the phytoplankton community in the Naroch Lakes. The approach to the quantification of ecological stability was based on defining the stability index as an inverse of the Euclidean Distance between the current and the reference states of the algal community (EuD-approach). The stability of the phytoplankton community was characterized by two indices: a “combined” index (SI[Comb]), and a “total community” index (SI[TotB]). SI[Comb] was calculated based on the individual taxonomic group biomasses and thus characterizes the stability of a community structure. SI[TotB] was calculated based on the values of the total algal biomass. Analyses of the results of this study extended the plausibility of the EuD-approach for the quantification of lake phytoplankton stability and allowed us to identify the dynamics of the stability of the Naroch Lakes phytoplankton. For the Naroch Lakes, we observed relatively larger SI[TotB] values in comparison with the SI[Comb] values. The results enabled us to examine the relationship between the lake trophic status and the stability of the phytoplankton community.
1. Introduction The use of the concept of “stability” to solve ecological problems (“ecological stability”) has a long history. Usually, drastic changes in ecological unit structure are qualitatively interpreted as a destabilization of these units. Examples of such hydroecological disruptions are shifts in the proportion of different types of primary producers (phytoplankton and macrophytes) in shallow lakes (Scheffer et al., 1993), changes to the structure of the phyto- and zooplankton communities in Lake Kinneret (Gal and Anderson, 2010; Gal and Hambright, 2014; Zohary et al., 2014b) and/or drastic changes in the variability and regularity of the dynamics of primary producers and zooplankton in Lake Sevan (Parparov, 1990; Simonyan, 1991). There is no universal definition of ecological stability, and depending on the objectives of a concrete study, there are hundreds of different definitions of this term (Rykiel, 1985; Grimm and Wissel, 1997; Reynolds, 2006). Different stability properties (e.g., resistance and/or resilience) are progressively becoming the object of natural resources management (Holling, 1996; Ludwig et al., 1997; Walker ⁎
et al., 2002; Groffman et al., 2006; Parparov and Gal, 2012). Therefore, the need to quantify ecological stability (i.e., express it in a mathematical form) and its relationships with various driving forces is growing. The concept of “stability” was first defined in the mathematical form by Lyapunov (sensu Justus, 2008). Apparently, the most known implementations of Lyapunov’s stability are the “predator – prey” equations developed by A. J. Lotka and V. Volterra (sensu Odum, 1971). For the objectives of this study, we note that Lyapunov’s concept does not provide a quantitative estimation of stability (“index of stability”). In this study, we used the definition of the stability of an ecological unit as an inverse of the Euclidean Distance between the unit current state and some predefined “reference” state (Parparov et al., 2015). Parparov et al. developed and employed the simple statistical approach to the quantification of the stability of the Lake Kinneret phytoplankton community (hereafter phytoplankton), which was then also applied to the entire lake ecosystem (Parparov and Gal, 2016). This approach (hereafter the “EuD-approach”) consisted of a set of sequential steps (the “ecological checklist”) defined by Grimm and Wissel (1997). For the Lake Kinneret ecosystem, the ecological stability was estimated
Corresponding author. E-mail addresses: [email protected] (T.M. Mikheyeva), [email protected] (A. Parparov), [email protected] (B.V. Adamovich), [email protected] (G. Gal), [email protected] (E.V. Lukyanova). http://dx.doi.org/10.1016/j.ecolind.2017.05.054 Received 1 December 2016; Received in revised form 13 May 2017; Accepted 22 May 2017 1470-160X/ © 2017 Elsevier Ltd. All rights reserved.
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
using a set of stability indices calculated for the individual taxonomic groups of phyto- and zooplankton, for their aggregated (“combined”) value associated with the community structure and for the total abundance (“total community” stability). Parparov et al. (2015) interpreted the exceedance of the total community stability over the combined stability index values as an appearance of the emergence principle (Webster, 1979; Jørgensen and Nielsen, 2013). Or, rephrased, a higher stability of the total community in regard to the aggregated stability of its components (i.e. the combined stability of the populations of the individual taxonomic groups). Using the EuD-approach provided a means for performing a quantitative comparison between the degrees of stability of different ecological units and for identifying stable and unstable periods. The results of the study by Parparov et al. (2015) corresponded well with the qualitative estimates of the stability status of both the plankton communities and the entire Lake Kinneret ecosystem (Zohary et al., 2014a). The implementation of the EuD-approach in other biotic communities and other lake ecosystems is limited due to a lack of available databases. Stability quantification contributes to long-term ecological monitoring; however, respective case studies are relatively scarce and the existing published materials are not sufficient estimation of the stability (e.g., Goldman, 2008; Rudstam et al., 2016). The Naroch Lakes (Belarus) represent a relatively rare example of lake ecosystems with different trophic status, where researchers were able to follow the remarkable changes that occurred in the aquatic ecosystems structure and functioning over the period 1968–2012 due to a long-term monitoring database (Winberg, 1985; Ostapenya, 2014). The main objective of this study was to quantify the phytoplankton stability of the Naroch Lakes and test the applicability of the EuD-approach to the quantification of phytoplankton stability in these lakes. In this paper, we estimated the relationships between the stability of the phytoplankton community and the lake trophic status. We further tested how common is the phenomenon of a higher degree of stability of the total phytoplankton community in relation to the stability of the phytoplankton structure.
Table 1 The major limnological characteristics of the Naroch Lakes during 1968–2014. Average and Min–Max in brackets. Indices
Batorino
Myastro
Naroch
Surface area, km2 Volume, mln. m3 Average depth, m Water retention time, yr Secchi depth, m Total P, μg L−1 Total N, mg L−1 Chlorophyll a, μg L−1 Trophic State Index (avg. 1968–1977)
6.3 18.7 2.4 1.0 1.1 (0.6–1.6) 51 (22–107) 1.2 (0.6–2.3) 30.1 (2.8–164) 66.8 (60.5–72.6)
13.1 70.1 5.4 2.5 3.2 (1.4–4.7) 39 (23–71) 0.9 (0.4–1.5) 12.4 (0.5–63.3) 60.7 (54.6–66.6)
79.6 710.0 8.9 10–11 6.0 (3.6–8.2) 18 (11–44) 0.7 (0.3–1.2) 3.6 (0.3–12.8) 42.3 (39.6–44.6)
spectrum of uses (in comparison with lakes Batorino and Myastro), and therefore this lake has a significant esthetical and recreational potential. Much of the nutrient and pollution load from the watershed is retained in the upper lakes of the system: Batorino and Miastro. Some limnological parameters of the Naroch Lakes are presented in Table 1. The Naroch Lakes were objects of intense hydroecological studies that were based on an extensive monitoring program initiated in the 1960’s (Winberg, 1985; Ostapenya, 1989). The trophic state of the Naroch Lakes ranged from oligo-mesotrophic (Lake Naroch) to eutrophic (Lake Batorino), and underwent remarkable changes during the studied period (1968–2012). The unusual phenomenon of deeutrophication (oligotrophication) in these lake ecosystems was attributed to drastic changes in the economic activities in their watershed, which between 1980 and 1990 resulted in a 45% decrease of the external phosphorus loading and in an invasion of the mollusk Dreissena polymorpha between 1992 and 2005 (Mikheyeva and Lukyanova, 2006; Ostapenya, 2014). The main features of the oligotrophication shift in the Naroch Lakes were as follows: a drop in the nitrogen and phosphorus concentrations, an increase in the water transparency (as Secchi Depth), a decrease in chlorophyll concentration and an increase of the relative significance of macrophytes in the primary production (Ostapenya et al., 2011). Ostapenya (2014) summarized the changes in the Naroch Lakes ecosystem using the term “benthification”, meaning “… an increase in the importance of benthic processes following increased water clarity promoted by nutrient reduction and Dreissena introduction”.
2. Material and methods 2.1. The Naroch Lakes (Belarus) This system of interconnected lakes includes Lakes Batorino, Myastro and Naroch (Fig. 1). These polymictic lakes are located in the north-west part of Belarus, in the glacial landscape. Lake Batorino, the first lake in the system, is a shallow eutrophic waterbody that connects to the mesotrophic Lake Myastro through a relatively narrow (5–6 m) channel. Meso-oligotrophic Lake Naroch is the largest waterbody of Belarus. The water quality of Lake Naroch is high enough for a wide
2.2. The phytoplankton community of the Naroch Lakes The long-term dynamics and structure of the Naroch Lakes phytoplankton have been described in detail (Mikheyeva and Lukyanova, 2006; Mikheyeva and Zhukova, 2015). During the studied period
Fig. 1. The Naroch Lakes and their location in Belarus.
482
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Chrysophyta biomass, from zero values in 1968–1977 to 14–17% of the total biomass, was a clear sign of the destabilization of the phytoplankton community. The major patterns of the evolution of the Naroch Lakes phytoplankton throughout different periods are described in detail in the “Results” section.
Table 2 The steps in the quantification of Naroch Lakes phytoplankton stability. Ecological unit
Phytoplankton community of Lakes Naroch, Myastro and Batorino Biomasses (B ) of the individual taxonomic groups: • Bacillariophyta, Cyanophyta, Chlorophyta, Cryptophyta,
State variables (SV)
i
Reference period Temporal and spatial scales Ecological distance
3. Methodology
Chrysophyta The sum of their biomass (TotB) 1968–1977 Ice free seasonal biomass mean average for the 0–5 m in Lake Batorino, 0–9 m in Lake Myastro and 0–16 m for Lake Naroch
•
EuDi = ABS ⎜⎛ ⎝
In this study, we followed the approach for quantification of stability (EuD-approach) suggested by Parparov et al. (2015). The data used for the current study (state variables) included the mean biomass (Bi), of each of the five main taxonomic groups, for the vegetative (ice-free) period (May–October), as well as their sum (TotB) documented in the Naroch Lakes during the period 1968–2012. There was no regular data for Lake Batorino during the period of 1984–1990. The water samples for the algae counts were collected monthly at specific monitoring points during from six depths (0.5, 3, 6, 8, 12 and 16 m) in Lake Naroch, four depths (0.5, 4, 7 and 9 m) in Lake Myastro and three depths (0.5, 3 and 5 m) in Lake Batorino. The water samples were then mixed in such a way that the water volume of each level (depth) in the mixed sample was proportional to the fraction of the level in the total water volume according to the lake bathymetry (Mikheyeva and Lukyanova, 2006). The state of the algal community was defined as a set of the individual biomass values at time t. The phytoplankton stability was considered to be identical to the constancy of the community state, i.e., the ability to sustain its structure in relation to the respective reference state, where biomasses of individual groups and their sum (Bi and TotB) were used as their reference states (RefBi and RefTotB; Grimm and Wissel, 1997; Reynolds, 2006). The main elements of this approach are summarized in Table 2 (modified from Parparov et al., 2015; Parparov and Gal, 2016). The changes in the state of the phytoplankton were studied based on the temporal variations in the biomass of each taxonomic group as well as the total phytoplankton biomass. The time step used for the stability calculations was a year. However, as mentioned above, the annual mean values were calculated for the vegetation (ice free) period (May–October). We assumed that the phytoplankton biomass variability observed during the reference period corresponded to the “natural variability”. The lower and upper variability limits of the individual groups and the total phytoplankton biomass during the reference period were defined as the 5th and 95th percentiles of the respective biomass time series during the reference period. The stability index values were defined (Eqs. (3) and (5) in Table 2) as an inverse function of the standardized Euclidean Distance (EuDi) of
Bi − Ref (B )i ⎞ ⎟ STD (B )i ⎠
Standardized ecological distance (1) EuD [TotB] = ABS
AggrEuD =
∑
TotB − Ref (TotB ) STD (TotB ) (EuDi)2
(1a) (2)
AggrEuD = (EuDi )2 Where i represents the biomass of the individual phytoplankton groups Stability Indices (SI)i 0 < SI < 1
(SI ) i =
1 1 + EuDi
Combined stability index
SI [Comb] = Total community stability index
(3)
1 1 + AggrEuD
(4)
1 1 + EuD [TotB] (5) The LSL value was defined as the 5th percentile of the stability index value time series during the reference period. If (SI/LSL) > 1, defines a stable state of the ecological unit; If (SI/LSL) < 1, defines a non-stable state of the ecological unit
SI [TotB] = Lower Stability Level (LSL) LSL-normalized stability index (nSI = SI/ LSL)
(1968–2012), the phytoplankton of the Naroch Lakes consisted of five major taxonomic groups: Bacillariophyta (diatoms, the dominating group in terms of algal biomass in Lakes Naroch and Myastro), Cyanophyta (cyanobacteria, the dominating group in terms of algal biomass in Lake Batorino), Chlorophyta (green algae), Cryptophyta and Chrysophyta (golden algae). On average, these five taxonomic groups comprised about 95% of the total algal biomass. The drastic increase in
Table 3 The dynamics of the decadal average biomass of the individual taxonomic groups (mg L−1) of the Naroch Lakes’ phytoplankton during different periods: I – 1968–1977 (the reference period), II – 1978–1987, III – 1993–2002, IV – 2003–2012. The variability limits: the 5th and 95th percentiles are shown in brackets. Cyanophyta
Cryptophyta
Bacillariophyta
Chrysophyta
Chlorophyta
TotB
Naroch I II III IV
0.11 0.18 0.13 0.35
(0.06–0.19) (0.09–0.33) (0.03–0.29) (0.14–0.66)
0.25(0.14–0.38) 0.13(0.03–0.26) 0.14(0.06–0.24) 0.36(0.18–0.58)
0.39(0.22–0.72) 0.66(0.36–1.27) 0.20(0.04–0.35) 0.24(0.10–0.47)
0.02(0.00–0.11) 0.20(0.04–0.34) 0.12(0.03–0.24) 0.18(0.05–0.42)
0.21(0.01–0.76) 0.06(0.02–0.12) 0.01(0.00–0.04) 0.04(0.00–0.13)
1.04(0.67–1.75) 1.36(0.73–2.22) 0.62(0.32–0.84) 1.26(0.85–1.57)
Myastro I II III IV
2.54 1.56 0.20 0.52
(0.51–7.77) (0.67–3.12) (0.02–0.59) (0.05–1.13)
1.78(0.32–3.55) 0.58(0.24–1.08) 0.24(0.12–0.49) 0.49(0.17–0.74)
2.44(1.33–5.15) 4.82(2.43–8.81) 0.90(0.22–2.25) 1.78(0.17–5.33)
0.00(0.00–0.00) 0.45(0.06–1.14) 0.21(0.03–0.49) 0.44(0.04–1.56)
0.55(0.34–0.86) 0.69(0.41–0.93) 0.02(0.00–0.03) 0.32(0.01–1.22)
7.44(3.75–13.0) 10.18(5.70–15.68) 1.62(0.82–3.32) 3.68(2.11–6.92)
12.45 (5.87–19.90)
0.75(0.10–1.59)
4.91(1.76–10.80)
0.00(0.00–0.00)
2.83(1.13–5.40)
21.93(9.90–32.88)
7.09 (3.39–10.41) 2.74 (0.39–6.31)
0.14(0.06–0.25) 0.41(0.22–0.69)
2.39(0.59–4.80) 3.19(1.52–5.73)
0.56(0.19–1.70) 1.24(0.48–3.15)
0.69(0.20–1.41) 0.62(0.26–1.33)
11.07(4.64–17.09) 8.81(3.85–14.27)
Batorino I IIa III IV a
No data.
483
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
the use of different percentile values: 1st, 5th, 10th, 15th and 20th percentile. The LSL-normalized values of the stability indices that were larger than unity indicated that the population or community was stable, while the values smaller than unity indicated a non-stable status of the phytoplankton (Parparov et al., 2015). In this article, further considerations were based mostly on the LSL-normalized (hereafter “normalized”) stability index values (nSI). Preliminary estimates performed for different time scales showed that the decadal scale indicates the main features and trends concerning stability quantification. Therefore, in order to test the EuD-approach, we examined the main features of the temporal dynamics of the 10-yr mean of the state variables and the normalized stability indices during four decadal periods: 1968–1977 (reference period), 1978–1987, 1993–2002 and 2003–2012. The Student t-test (Zar, 1998) was used to test the statistical hypotheses indicating the significance of destabilization events, for instance, the hypothesis that the normalized SI decadal averages are smaller than unity (n = 10, P < 0.05). The lake trophic state was quantified by the lake Trophic State Index (TSI; Carlson, 1977; Adamovich et al., 2016): TSI = (TSISD + TSITP + TSIChl)/3
(6)
where TSISD,TSITP and TSIChl are individual trophic state indices calculated for the Secchi Depth (SD), total phosphorus (TP) and chlorophyll a (Chl) concentrations as follows: TSISD = −14.39·ln(SD) + 59.91
(6a)
TSITP = 14.43·ln(TP) + 4.15
(6b)
TSIChl = 9.76·ln(Chl) + 30.91
(6c)
The TSI natural variability limits were estimated by the 5th and 95th percentiles calculated for the reference period time series. The relationship between the stability indices of the total phytoplankton community and structure was estimated by the ratio of SI [TotB]/SI[Comb]. The trophic status effects on the phytoplankton stability were examined by regressing SI[Comb] and SI[TotB] versus TSI. We also regressed the changes in the combined SIs versus the changes in the TSI. The changes in the aggregated SIs (SI[Comb], SI[TotB]) and TSI were calculated based on their reference values, as follows: Fig. 2. The dynamics of the 10-yr annual average values of the Naroch Lakes phytoplankton biomass. The numbers above the stacked columns represent the values of the total phytoplankton biomass. The dotted horizontal lines represent the limits of the natural variability (the 5th and 95th percentiles calculated for the reference period) of the total phytoplankton biomass.
ΔTSI = (TSI)ref − (TSI)i
(7a)
ΔSI = (SI)ref − (SI)j
(7b)
where indices “ref” correspond to the reference period, while indices “i” and ‘j’ span from 2 to 4, and indicate the decadal periods, i.e., 2–1978–1987, 3–1993–2002, 4–2003–2012. Similar to the Lake Kinneret case study (Parparov and Gal, 2016), we used the qualitative estimates of the stability of the Naroch Lakes phytoplankton (Mikheyeva and Lukyanova, 2006; Ostapenya et al., 2012; Ostapenya, 2014) as an important criterion for the validation of the EuD-approach. The respective data obtained for Lake Kinneret (Parparov et al., 2015; Parparov and Gal, 2016) were used for the analysis of the study results.
the state variable (i.e., the biomass of the individual algae group or total phytoplankton biomass) from the respective reference state (RefBi) at a given point in time, t (Kindt and Coe, 2005; Greenacre, 2008). The stability of the phytoplankton was quantified by comparing the stability indices using two different aggregating schemes. The first scheme used the combined stability index (SI[Comb]), which was calculated based on the stability indices of the individual phytoplankton (SIi) taxonomic groups (Eq. (4) in Table 2). The combined stability index characterizes the stability associated with the changes to the structure of the community. The second scheme used a total community stability index, calculated based on the total abundance of phytoplankton (SI[TotB], Eq. (5) in Table 2). In order to determine the stability status of an ecological unit and to allow a direct comparison between the stability index values of different ecological units, the SI values were normalized to a lower stability limit (LSL). The LSL value was defined as the 5th percentile of the stability index values calculated for the reference period. In order to estimate the effect of the selected percentile value, for SI[Comb] and SI [TotB], we calculated the sensitivity (relative to the 5th percentile) to
4. Results 4.1. Establishing the reference period Defining the reference period is not straightforward, as there is no clearly formulated procedure for its delineation. Therefore, the establishment of the reference period is essentially subjective, depending on the study subject and the researcher’s interest. In this study, we selected the reference period that corresponded to the state of the community that existed in the past and was considered minimally disturbed 484
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 3. The dynamics of the normalized stability indices (nSI) of the individual phytoplankton group in Naroch Lakes (the 10-yr annual average and STD values). The horizontal dashed lines indicate the nSI value equal to unity, and thus separates the stable and non-stable states of the ecological unit. The asterisks above the column indicate values significantly smaller than unity (at P < 0.05).
4.2. Changes to the biomass, species composition and temporal dynamics of the Naroch Lakes phytoplankton
(Grimm and Wissel, 1997; Donohue et al., 2013). We used the decadal dynamics of the phytoplankton biomass and the Naroch Lakes trophic state as an argument for determining the period to be defined as the reference period. The quantification of phytoplankton stability in three different lakes indirectly assumes that the stability of all three lakes will be estimated according to the same reference period; otherwise, the established stability estimates will be incomparable. When selecting the reference period we had to balance the requirements according to the relative constancy of the phytoplankton structure as well as the constancy of the lake trophic state. Therefore, we selected the period of 1968–1977 as the “reference period” for the phytoplankton of the Naroch Lakes based on the following reasons:
Over a period of more than 40 years (from 1968 until 2012), the average biomass of all five phytoplankton taxonomic groups in the Naroch Lakes underwent notable changes. The following major features of the changes to the biomass of the entire phytoplankton as well as the individual taxonomic groups (Mikheyeva and Lukyanova, 2006; Mikheyeva and Zhukova, 2015) could affect the stability of the Naroch Lakes phytoplankton (Fig. 2 and Table 3):
• the most common feature is the considerable increase in the biomass and relative contribution of Chrysophyta; • the increase in the relative contribution of Cyanophyta in Lake
• Major qualitative conclusions regarding changes to the stability of • •
the Naroch Lakes phytoplankton were based on comparisons with the period of 1968–1977 (Mikheyeva and Lukyanova, 2006; Ostapenya, 2014). Immediately after that period, there were drastic changes in the phytoplankton communities of all three lakes (Table 3). This is the period for which a more complete database exists (note the data deficiency for Lake Batorino for the period of 1984–1990).
• • • 485
Naroch was accompanied by a decrease in this algae group biomass in Lakes Myastro and Batorino; irregular though considerable changes in the total biomass of the phytoplankton in all three studied lakes. The following changes to the decadal average algal biomass should be noted as well: the decrease in the Cyanophyta biomass in Lakes Myastro and Batorino after 1982–1984, below the lower “natural” variability (the 5th percentile) level. In contrast, the dynamics of the
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 5. The scatter plot and best fit regression lines of the relationships between the normalized annual average values of the combined (nSI[Comb]) and total community (nSI[TotB]) stability indices for the Naroch Lakes (A) and for Lake Kinneret (C) phytoplankton. B – the same as in A for the decadal average values in the Naroch Lakes.
Fig. 4. The decadal average ( ± std) values of the normalized SI[Comb] and SI[TotB] values (nSI[Comb] and nSI[TotB]) of the Naroch Lakes phytoplankton. The horizontal dashed lines indicate the value of the nSI values equal to unity, and thus separates the stable and non-stable states of the ecological unit. The asterisks above the column indicate values significantly smaller than unity (at P < 0.05).
•
Table 4 Values of the lower stability limits calculated as smaller percentiles for SI[Comb] and SI [TotB]. In brackets: the error (relative to the 5th percentile value). Percentiles
1% 5% 10% 15% 20%
Lake Naroch
Lake Myastro
Lake Batorino
SI [Comb]
SI [TotB]
SI[Comb]
SI[TotB]
SI[Comb]
SI [TotB]
0.211 (18.2) 0.257 (0.0) 0.316 (22.7) 0.341 (32.6) 0.357 (38.8)
0.423 (5.3) 0.447 (0.0) 0.476 (6.6) 0.499 (11.8) 0.521 (16.5)
0.322 (3.1)
0.496 (1.6)
0.301 (1.9)
0.332 (0.0)
0.505 (0.0)
0.307 (0.0)
0.345 (3.9)
0.515 (2.0)
0.315 (2.4)
0.354 (6.4)
0.521 (3.3)
0.320 (4.3)
0.361 (8.6)
0.527 (4.4)
0.325 (5.9)
0.398 (9.3) 0.438 (0.0) 0.489 (11.6) 0.503 (14.7) 0.506 (15.4)
•
•
Cyanophyta biomass in Lake Naroch showed a weak trend of increase in the 1980′s and after 2000, above the 95th percentile of the respective reference period; the periods of increase of the Bacillariophyta biomass (above the level of 95% in Lakes Naroch and Myastro) were followed by a biomass drop below the 5th percentile level (especially in Lake Myastro, Fig. 2); during the reference period, the Chrysophyta biomass in all the Naroch Lakes was close to zero (Table 3 and Fig. 2). Since the late 1970s, the biomass of Chrysophyta underwent a drastic increase, up to well above the upper limit of the “natural” variability, indicating a drop of stability of this taxonomic group. A particularly impressive change in the Chrysophyta biomass was detected in Lake Myastro: within a year (between 1978 and 1979) the biomass of this algal group increased from zero values in 1968–1977 to 0.33 mg L−1 (on average, for 1978–2014, Fig. 2, Table 3). In 2003–2012, the relative contribution of Chrysophyta to the total algae biomass in Lake Myastro comprised 14–17% of the total algae biomass; the Chlorophyta biomass in all three lakes decreased by almost an order of magnitude compared to the reference period. The total biomass of the phytoplankton in the Naroch Lakes during
486
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 6. The dynamics of the ratio of nSI[TotB]/nSI[Comb] for the Naroch Lakes phytoplankton (the 10-yr annual average and STD values). The horizontal dashed lines indicate the value of the ratio equal to unity. The dynamics of a similar ratio for Lake Kinneret are shown as well.
[TotB]) are of special interest. The decadal average values of these indices dropped in varying degrees, though more gradually for SI[TotB] (Fig. 4). The almost twofold decrease of SI[Comb] during the 40 years that passed since its reference state represents a quantitative estimate of the change in the stability structure of the Naroch Lakes phytoplankton. The decadal dynamics of the SI[TotB] showed a relatively smaller decrease in the stability. The decadal dynamics of the normalized SI[Comb] and SI[TotB] values, (Fig. 4) showed that there were no periods of prolonged destabilization in Lakes Naroch and Batorino. The normalized SI[Comb] and SI[TotB] values calculated for these lakes were close to, or exceeded, unity; thus the structure and abundance of the phytoplankton in these lakes remained within the stability limits during almost the entire period from 1968 to 2012. Statistically significant destabilization events of the phytoplankton structure (SI[Comb] < 1, P < 0.05) were recorded only for the phytoplankton structure of Lake Myastro during 1993–2012 (Fig. 4). The destabilization of the total community stability in Lake Myastro was recorded only in 1993–2002 (SI[TotB] < 1 at P < 0.05). During the rest of the periods, the decadal average values of SI[TotB] obtained for this lake were not less than unity.
1968–2012 varied within wide limits: the 10-year average values of the TotB ranged from 0.62 to 1.36 mg L−1 in Lake Naroch, from 1.62 to 10.18 mg L−1 in Lake Miastro, and from 8.81 to 21.9 mg L−1 in Lake Batorino (Fig. 2). In Lake Naroch, during most part of the observation period the TotB varied within the “natural” limit of its variability (Fig. 2, Table 3). In Lake Myastro, the TotB dropped below the “natural” limit in 1992–2008; after that, there was a trend towards a recovery of the TotB in this lake (Fig. 2 and Table 3).
4.3. Quantifying the stability of the Naroch Lakes phytoplankton 4.3.1. Temporal dynamics of the stability indices The phytoplankton stability was quantified at three different levels: the stability of the five individual phytoplankton groups, the combined stability of the five groups (SI[Comb], Eqs. (3) and (4)) and the stability of the total phytoplankton community (SI[TotB], Eq. (5)). Compared to the reference period values, the normalized stability indices of the individual algal groups (averaged for 10-year periods) in all three lakes decreased (Fig. 3), with two exceptions: the SI values of Cryptophyta and Bacillariophyta in Lake Batorino slightly increased. Note, that the decrease in the SI[Cyano] in the last decade was caused by contrasting reasons: by the increase of the Cyanobacteria biomass above the upper natural variability level in Lake Naroch, while in Lakes Myastro and Batorino the Cyanobacteria biomass decreased below the lower natural variability level of the biomass of this alga group during 1992–2011 (see also Table 3). The dynamics of most of the normalized stability indices did not indicate any destabilization events (i.e., the LSL-normalized SI values were larger than or equal to unity). The statistically significant (at P < 0.05) decrease of the normalized SI was associated with the drastic increase of the Crysophyta biomass and the decrease of the Chlorophyta biomass in Lakes Myastro and Batorino during the last three decades (Figs. 2 and 3). The dynamics of the aggregated stability indices (SI[Comb] and SI
4.3.2. The effect of the selected lower percentile values on estimating of the lower stability limits for SI[Comb] and SI[TotB] In order to study the sensitivity of the results to the selected value of the lower stability limits (LSL) we tested a range of percentile values. Thus we calculated the aggregated stability indices (SI[Comb] and SI [TotB]), for the Naroch lakes, based on the following range of percentile values: 1%, 5%, 10%, 15% and 20%. The resulting stability indices values indicated relatively limited variability in the SI values (Table 4). For Lakes Myastro and Batorino, the errors associated with using of the 1st and 10th percentiles did not exceed 10%; relatively higher relative errors were calculated for SI[Comb] in Lake Naroch. The higher errors were calculated for the 15th and 20th percentiles, however the 487
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
P < 0.005). A statistically insignificant relationship was obtained between the annual average values of SI[Comb] and SI[TotB] for the Lake Kinneret phytoplankton (Fig. 5C). It should especially be noted that for the Naroch Lakes phytoplankton, the SI[TotB] values were larger (on average, by 20%) than SI [Comb] (and thus the ratio of SI[TotB]/SI[Comb] exceeded unity) during each of the 10-year periods for each of the Naroch Lakes (Fig. 6). It means that at the decadal scale, in the Naroch Lakes, the stability of the total phytoplankton community exceeded the stability associated with the phytoplankton structure. 4.4. The relationship between the Naroch Lakes trophic status and the phytoplankton stability After the increase (though within the natural variability limits) in 1968–1987, the TSI decreased (even persistently below the natural variability level, in Lake Myastro), indicating the oligotrophication/ deeutrophication of the Naroch Lakes during 1993–2012 (Fig. 7). Despite the considerable range of the TSI values observed for the Naroch Lakes (36 ≤ TSI ≤ 67), the scatter plots did not indicate any relationship between the aggregated stability indices (SI[Comb], SI [TotB]) and TSI (Fig. 8A & B). Note that the values obtained for Lake Kinneret are within the data “cloud” received for the Naroch Lakes. However, the changes of both aggregated stability index values calculated for the Naroch Lakes (see Eq. (7a,b)) for the decadal average values positively correlated quite well with the corresponding changes of the TSI (Fig. 8 C & D). ΔSI[Comb] linearly correlated with ΔTSI (R2 = 0.75, P < 0.05), while the regression of ΔSI[TotB] versus ΔTSI was non-linear and less statistically significant (R2 = 0.54, P < 0.1). The data for Lake Kinneret were outside of the main tendencies obtained for the Naroch Lakes data. Note, that the decadal average TSI values for Lake Kinneret had a much narrower range compared to the Naroch Lakes: between 47.9 and 49.3. 5. Discussion The results of our study performed using the EuD-approach indicated a decrease of the Naroch Lakes phytoplankton stability. Our results do not contradict the qualitative conclusions by Mikheyeva and Lukyanova (2006) and Ostapenya (2014) about a decrease of the Naroch Lakes phytoplankton stability based on analysis of the longterm changes to the dynamics of the biomass and the structure of the Naroch Lakes phytoplankton (Fig. 2) together with the drop in the TSI values (Fig. 7). We consider the correspondence between previous qualitative estimates and our results as indirect validation of EuD-approach for quantifying the stability of the phytoplankton communities in different lakes. Moreover, the implementation of the EuD-approach made the qualitative estimates more extensive and precise, and allowed us to answer some basic questions outlined by the study objectives. The lowering of the Naroch Lakes stability was indicated by a decline in the decadal average values for both aggregated stability indices: SI[Comb] and SI[TotB] (Fig. 4). On the decadal scale, the stability of the phytoplankton structure (as SI[Comb]) during the period 1978–2012 dropped compared to the reference period by 32%, 54% and 53% for Lakes Naroch, Myastro and Batorino, respectively. These values are comparable with the SI[Comb] decrease obtained for Lake Kinneret (by 50%, according to Parparov et al., 2015). However, the normalized SI[Comb] values were significantly (P < 0.05) smaller than unity only for Lake Myastro during 1993–2012, thus indicating a destabilization of the phytoplankton structure in this lake. The normalized SI[TotB] values did not fall, statistically, below unity (excluding the decadal average value in Lake Myastro in 1993–2002, Fig. 4). Therefore, we conclude that on the decadal scale, the stability of the total phytoplankton community of the Naroch Lakes decreased in 1978–2012, but remained within its natural stability limits. At the same time, our approach revealed some peculiarities of the
Fig. 7. The temporal dynamics of the trophic state index (TSI) in the Naroch Lakes (decadal average ± std values). Horizontal lines indicate the limits of the TSI natural variability.
use of these percentiles for the LSL estimates would be associated with unwanted narrowing of the LSL estimates: 30 and 40% of the LSL values would be lost, respectively. The use of the 1st and 10th percentile would provide close estimates of SI[Comb] and SI[TotB], but we declined their use because the use of the 1st percentile would allow including of almost all stability index values into the reference period while the use of the 10th percentile would require excluding of 20% of the SI values from the reference state estimate.
4.3.3. The relationship between aggregated stability indices: SI[Comb] and SI[TotB] The annual average values of the aggregated stability indices (SI [TotB] and SI[Comb]) correlated quite well (Fig. 5A, R2 = 0.37, n = 118, P < 0.01); a stronger (though not linear) correlation was obtained for the decadal average values (Fig. 5B, R2 = 0.82, n = 11, 488
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 8. A & B: The scatter plot of SI[Comb] and SI[TotB] versus TSI values for the Naroch Lakes and Lake Kinneret (outlined); decadal average values. C and D: The scatter plots and best fit regression lines for the relationships between the changes in the aggregated stability indices ΔSI[Comb] and ΔSI[TotB], and the changes of the TSI (ΔTSI). The data for Lake Kinneret are encircled.
previously regarding Lake Kinneret phytoplankton, was observed in the three other lakes. Thus, we might assume the existence of a generality of a relatively larger stability of the total phytoplankton community in comparison with the stability of its structure. For Lake Kinneret, the exceedance of SI[TotB] over SI[Comb] was interpreted (Parparov et al., 2015; Parparov and Gal, 2016) as evidence of the emergent properties of the phytoplankton community stability, equivalent to the existence of the homeostatic mechanisms that support the stability of the total phytoplankton community despite the changes in its internal structure. According to the emergence principle, the properties of the higher hierarchical levels are distinct from the properties of the lower hierarchical levels (Webster, 1979; Justus, 2008; Jørgensen and Nielsen, 2013). The emergence principle itself is a kind of a hypothesis, which does not explain the nature of the homeostatic mechanisms responsible for the larger stability of the higher hierarchical levels. We should note that for Lake Kinneret, the exceedance of SI[TotB] over SI[Comb] was not observed for the zooplankton community: the stability index of the total zooplankton community was approximately equal to the stability index of its structure (Parparov and Gal, 2016). Further studies are needed to clarify the generality and limitations of this phenomenon. The quantification of the phytoplankton stability for the Naroch Lakes and Lake Kinneret that have different trophic status (36 ≤ TSI ≤ 67) allowed us to test the potential effects of the trophic status on the stability of the producer community. The regression between the phytoplankton aggregated stability indices and TSI did not indicate any relationship between these variables (Fig. 8 A & B). However, such a relationship was revealed for the changes of the aggregated stability indices (ΔSI[Comb], ΔSI[TotB]) and ΔTSI (Fig. 8 C & D), though for the Naroch Lakes only. We interpret these results as indicating the relative independence of the phytoplankton stability in relation to lake trophic status. However, at the decadal scale, the shifts in the stability of the phytoplankton structure and abundance in the Naroch Lakes might be induced by the shifts that occur in their trophic
Naroch Lakes phytoplankton in regards to the quantification of its stability. The Naroch Lakes had no period of relatively limited variability of the algal biomass, as seen in the Lake Kinneret phytoplankton (Parparov et al., 2015). The variability of the individual groups and of the total phytoplankton biomass in the Naroch Lakes during the established reference period (1968–1977) was relatively large (sometimes it ranged by an order of magnitude, e.g., Cyanophyta in Lake Myastro, Table 2). Therefore, due to the large variability observed during the reference period, the “natural variability limits” of the algal biomass in the Naroch Lakes (defined by the 5th and 95th percentiles obtained for the respective reference periods) were relatively wide. An analysis of the decadal dynamics of the individual stability indicated that despite trends of lowering of algal biomasses (or an increase Cyanophyta biomass), only Chrysophyta in Lakes Myastro and Batorino showed a statistically significant persistent (about three decades) destabilization (Fig. 3). For all three Naroch Lakes, the LSL-normalized values of the total community stability index (nSI[TotB]) were systematically larger (by appr. 20%) than nSI[Comb] (Fig. 6). For Lake Kinneret, the SI[TotB] was more than twice the combined stability index (SI[Totb]/SI [Comb] ≈ 2.2, Parparov et al., 2015). Note that in Lake Kinneret the exceedance of SI[TotB] over SI[Comb] was a natural consequence of the relatively higher stability of the total phytoplankton biomass, while in the Naroch Lakes the total biomass variability was comparable with the variability of the individual algal group biomass (Table 3 and Fig. 2). A comparison of the data acquired for the Naroch Lakes and Lake Kinneret indicates that there are two types of dynamics for the aggregated stability indices: 1. Intercorrelated changes of the SI[Comb] and SI[TotB] observed in the Naroch Lakes (Fig. 5A), and 2. Relatively large changes in the SI[Comb], which were independent of smaller changes in the SI[TotB] as seen in Lake Kinneret (Fig. 5C). However, despite the distinction between these two types of stability index dynamics, the exceedance of SI[TotB] over SI[Comb], mentioned 489
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Statistics, Stanford University Fall, 2008, STA 254. http://www.econ.upf.edu/ ∼michael/stanford/. Grimm, V., Wissel, C., 1997. Babel or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia 109, 323–334. Groffman, P.M., Baron, J.S., Blett, T., et al., 2006. Ecological thresholds: the key to successful environmental management or an important concept with no practical application? Ecosystems 9, 1–13. Holling, C.S., 1996. Engineering resilience vs ecological resilience. In: Schultze, P.C. (Ed.), Engineering Within Ecological Constraints. National Academy Press, Washington DC, pp. 31–43. Jørgensen, S.E., Nielsen, S.N., 2013. The properties of the ecological hierarchy and their application as ecological indicators. Ecol. Ind. 28, 48–53. Justus, J., 2008. Complexity, diversity and stability. Chapter 18. In: Sahotra, S., Plutyns, S. (Eds.), A Companion to the Philosophy of Biology. Blackwell Publishing, pp. 321–350. Kindt, R., Coe, R., 2005. Tree Diversity Analysis. A Manual and Software for Common Statistical Methods for Ecological and Biodiversity Studies. World Agroforestry Centre (ICRAF), Nairobi 197pp. Ludwig, D., Walker, B., Holling, C.S., 1997. Sustainability, stability, and resilience. Conserv. Ecol. 1, 7. Available from the Internet. URL: http://www.consecol.org/ vol1/iss1/art7/. Mikheyeva, T.M., Lukyanova, E.V., 2006. The direction and character of multi-year changes of phytocoenotic structure and the indices of quantitative development of phytoplankton communities of the Naroch lakes in the course of trophic state evolution. Izv. Samar. Nauch. Tsentra RAN 8, 125–140 (in Russian). Mikheyeva, T.M., Zhukova, T.V., 2015. Conclusion. In: Mikheyeva, T.M. (Ed.), The Bulletin of Ecological State of Lakes, Naroch, Myastro. Belatus State University, Minsk, pp. 105–107. Odum, E.P., 1971. Fundamentals of Ecology, third edn. W.B. Saunders, London 360 pp. Ostapenya, A.P., Zhukova, T.V., Mikheyeva, T.M., 2011. Bentification as a stage in the Naroch Lakes evolution. Proc. Belarus State Univ. 2 (2), 62–66. Ostapenya, A.P., 1989. Seston and Detritus as Structural and Functional Components of Water Ecosystems. Ref. Diss. Zoological Institute, Minsk 42 pp. (in Russian). Ostapenya A.P., 2014. Benthification of freshwater lakes: exotic mussels turning ecosystems upside down. Chapter 36. In: C. M. Mayer [et al.]//Quagga and Zebra mussels. Biology, impact and control/Edited by T. F. Nalepa, D. W. Schloesser. Second edition. CRC Press London, New York, p. 575–585. Parparov, A., Gal, G., 2012. Assessment and implementation of a methodological framework for sustainable management: Lake Kinneret as a case study. J. Environ. Manage. 101, 111–117. Parparov, A., Gal, G., 2016. Quantifying ecological stability: from community to the lake ecosystem. Ecosystems. http://dx.doi.org/10.1007/s10021-016-0090-z. (accepted for publication). Parparov, A., Gal, G., Zohary, T., 2015. Quantifying the ecological stability of a phytoplankton community: the Lake Kinneret case study. Ecol. Indic. 56, 134–144. Parparov, A.S., 1990. Some characteristics of the community of autotrophs of Lake Sevan in connection with the eutrophication. Hydrobiologia 191, 15–21. Reynolds, C.S., 2006. The Ecology of Phytoplankton. Cambridge University Press 535 p. Oneida Lake: Long-term Dynamics of a Managed Ecosystem and Its Fishery. In: Rudstam, L.G., Mills, E.L., Jackson, J.R., Stewarts, D.J. (Eds.), American Fishery Society, Bethesda, Maryland 541 PP. Rykiel, E.J., 1985. Towards a definition of ecological disturbance. Aust. J. Ecol. 10, 361–365. Scheffer, M., Hosper, S.H., Meijer, M.L., Moss, B., 1993. Alternative equilibria in shallow lakes. Trends Ecol. Evol. 8, 275–279. Simonyan, A.A., 1991. Zooplankton ozera Sevan (Zooplankton of Lake Sevan, in rus). Izdvo AN Armenii, Erevan 299pp. Walker, B., Carpenter, S., Anderies, N., et al., 2002. Resilience management in socialecological systems: a working hypothesis for a participatory approach. Conserv. Ecol. 6 (1), 14. [online] URL: http://www.consecol.org/vol6/iss1/art14/. Webster, J.R., 1979. Hierarchical organization of ecosystems. In: Halfon, E. (Ed.), Theoretical System Ecology. Academic Press, New York, pp. 119–131. Winberg, G.G., 1985. The Ecological System of the Naroch Lakes. Minsk (in Rus.). 289 pp. Zar, J.H., 1998. Biostatistical Analysis. Prentice Hall International, INC, New Jersey p. 147. Zohary, T., Sukenik, A., Nishri, A., 2014a. Lake kinneret: current understanding and future perspectives. Chapter 27. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 417–438. Zohary, T., Yacobi, Y.Z., Alster, A., Fishbein, et al., 2014b. Phytoplankton. Chap. 10. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 161–190.
status. It is obvious that studying the relationship between trophic status and ecological stability requires further accumulation of relevant limnological information. However, in any case, it will be impossible to solve this scientific task without the quantification of stability. The approach of quantification of stability tested and implemented in this study might represent a useful tool for progressing the theory of ecological stability. 6. Concluding remarks
• The results of our study performed for the Naroch Lakes confirmed •
• •
that the approach for quantifying of the ecological stability developed previously for the Lake Kinneret phytoplankton, could be used successfully for lakes of different trophic status. Analyses performed on the dynamics of the stability index values of the Naroch Lakes phytoplankton (of the individual groups, combined and total community) indicated a tendency of decreasing phytoplankton stability in these lakes. Despite this tendency, the total community stability of the Naroch Lakes phytoplankton remained within its stability limits. On the other hand, the dynamics of the stability indices associated with the changes to the phytoplankton structure indicated a persistent destabilization of the phytoplankton community in Lakes Myastro (after 1987) and Batorino (after 1997). For Naroch Lakes, the total community stability exceeded the stability of the phytoplankton structure. Together with similar relation obtained earlier for Lake Kinneret, this observation allowed us to assume the generality of this phenomenon. Quantification of the phytoplankton stability in the Naroch Lakes for the first time allowed us to consider the potential effects of the lake trophic status on the stability of the phytoplankton community and to connect the shifts in the lake tropic status with the shifts in the phytoplankton stability.
Acknowledgments This study was partially supported by the Belarus Republican Foundation for Fundamental Research. We thank two anonymous Reviewers whose valuable comments greatly contributed to improvement of the manuscript. We are very thankful to L. Baumer for her valuable help in preparing and editing this manuscript. References Adamovich, B.V., Zhukova, T.V., Mikheyeva, T.M., Kovalevskaya, R.Z., Lukyanova, E.V., 2016. Long-term variations of the trophic state index in the Narochanskie lakes and its relation with the major hydroecological parameters. Water Resour. 43, 809–817. Carlson, R.E., 1977. A trophic state index for lakes. Limnol. Oceanogr. 22, 361–369. Donohue, I., Owen, L., Petchey, O.L., Montoya, M.L., et al., 2013. On the dimensionality of ecological stability. Ecol. Lett. 16, 421–429. Gal, G., Anderson, W., 2010. A novel approach to detecting a regime shift in a lake ecosystem. Methods Ecol. Evol. 1, 45–52. Gal, G., Hambright, K.D., 2014. Metazoan zooplankton. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 227–245. Goldman, C.R., 2008. Introduction. Five decades of environmental and social change at Lake Tahoe. In: Bachand, T. (Ed.), Lake Tahoe: A Fragile Beauty. Chronicle Books, San Francisco, pp. 14–18. Greenacre, M., 2008. Correspondence Analysis and Related Methods. Department of
490
Contents lists available at ScienceDirect
Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind
Original Articles
The dynamics of freshwater phytoplankton stability in the Naroch Lakes (Belarus)
MARK
⁎
T.M. Mikheyevaa, A. Parparovb, , B.V. Adamovicha, G. Galb, E.V. Lukyanovaa a b
Belarusian State University, Biological Department, 220030, Nezavisimosti Ave., 4, Minsk, Belarus Kinneret Limnological Laboratory, Israel Oceanographic and Limnological Research, POB 447, Migdal, 14950, Israel
A R T I C L E I N F O
A B S T R A C T
Keywords: Naroch lakes Lake Kinneret Phytoplankton community Ecological stability Trophic status Principle of emergence
The phytoplankton community of the Belarus Lakes Naroch, Myastro and Batorino, which have a Trophic State Index of 42.3, 60.7 and 66.8, respectively, underwent drastic changes to their structure during the period between 1968 and 2012. Thanks to an extensive monitoring program, these changes were well-documented and were qualitatively interpreted as signs of the community destabilization. The main objective of this study was the quantification of the ecological stability of the phytoplankton community in the Naroch Lakes. The approach to the quantification of ecological stability was based on defining the stability index as an inverse of the Euclidean Distance between the current and the reference states of the algal community (EuD-approach). The stability of the phytoplankton community was characterized by two indices: a “combined” index (SI[Comb]), and a “total community” index (SI[TotB]). SI[Comb] was calculated based on the individual taxonomic group biomasses and thus characterizes the stability of a community structure. SI[TotB] was calculated based on the values of the total algal biomass. Analyses of the results of this study extended the plausibility of the EuD-approach for the quantification of lake phytoplankton stability and allowed us to identify the dynamics of the stability of the Naroch Lakes phytoplankton. For the Naroch Lakes, we observed relatively larger SI[TotB] values in comparison with the SI[Comb] values. The results enabled us to examine the relationship between the lake trophic status and the stability of the phytoplankton community.
1. Introduction The use of the concept of “stability” to solve ecological problems (“ecological stability”) has a long history. Usually, drastic changes in ecological unit structure are qualitatively interpreted as a destabilization of these units. Examples of such hydroecological disruptions are shifts in the proportion of different types of primary producers (phytoplankton and macrophytes) in shallow lakes (Scheffer et al., 1993), changes to the structure of the phyto- and zooplankton communities in Lake Kinneret (Gal and Anderson, 2010; Gal and Hambright, 2014; Zohary et al., 2014b) and/or drastic changes in the variability and regularity of the dynamics of primary producers and zooplankton in Lake Sevan (Parparov, 1990; Simonyan, 1991). There is no universal definition of ecological stability, and depending on the objectives of a concrete study, there are hundreds of different definitions of this term (Rykiel, 1985; Grimm and Wissel, 1997; Reynolds, 2006). Different stability properties (e.g., resistance and/or resilience) are progressively becoming the object of natural resources management (Holling, 1996; Ludwig et al., 1997; Walker ⁎
et al., 2002; Groffman et al., 2006; Parparov and Gal, 2012). Therefore, the need to quantify ecological stability (i.e., express it in a mathematical form) and its relationships with various driving forces is growing. The concept of “stability” was first defined in the mathematical form by Lyapunov (sensu Justus, 2008). Apparently, the most known implementations of Lyapunov’s stability are the “predator – prey” equations developed by A. J. Lotka and V. Volterra (sensu Odum, 1971). For the objectives of this study, we note that Lyapunov’s concept does not provide a quantitative estimation of stability (“index of stability”). In this study, we used the definition of the stability of an ecological unit as an inverse of the Euclidean Distance between the unit current state and some predefined “reference” state (Parparov et al., 2015). Parparov et al. developed and employed the simple statistical approach to the quantification of the stability of the Lake Kinneret phytoplankton community (hereafter phytoplankton), which was then also applied to the entire lake ecosystem (Parparov and Gal, 2016). This approach (hereafter the “EuD-approach”) consisted of a set of sequential steps (the “ecological checklist”) defined by Grimm and Wissel (1997). For the Lake Kinneret ecosystem, the ecological stability was estimated
Corresponding author. E-mail addresses: [email protected] (T.M. Mikheyeva), [email protected] (A. Parparov), [email protected] (B.V. Adamovich), [email protected] (G. Gal), [email protected] (E.V. Lukyanova). http://dx.doi.org/10.1016/j.ecolind.2017.05.054 Received 1 December 2016; Received in revised form 13 May 2017; Accepted 22 May 2017 1470-160X/ © 2017 Elsevier Ltd. All rights reserved.
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
using a set of stability indices calculated for the individual taxonomic groups of phyto- and zooplankton, for their aggregated (“combined”) value associated with the community structure and for the total abundance (“total community” stability). Parparov et al. (2015) interpreted the exceedance of the total community stability over the combined stability index values as an appearance of the emergence principle (Webster, 1979; Jørgensen and Nielsen, 2013). Or, rephrased, a higher stability of the total community in regard to the aggregated stability of its components (i.e. the combined stability of the populations of the individual taxonomic groups). Using the EuD-approach provided a means for performing a quantitative comparison between the degrees of stability of different ecological units and for identifying stable and unstable periods. The results of the study by Parparov et al. (2015) corresponded well with the qualitative estimates of the stability status of both the plankton communities and the entire Lake Kinneret ecosystem (Zohary et al., 2014a). The implementation of the EuD-approach in other biotic communities and other lake ecosystems is limited due to a lack of available databases. Stability quantification contributes to long-term ecological monitoring; however, respective case studies are relatively scarce and the existing published materials are not sufficient estimation of the stability (e.g., Goldman, 2008; Rudstam et al., 2016). The Naroch Lakes (Belarus) represent a relatively rare example of lake ecosystems with different trophic status, where researchers were able to follow the remarkable changes that occurred in the aquatic ecosystems structure and functioning over the period 1968–2012 due to a long-term monitoring database (Winberg, 1985; Ostapenya, 2014). The main objective of this study was to quantify the phytoplankton stability of the Naroch Lakes and test the applicability of the EuD-approach to the quantification of phytoplankton stability in these lakes. In this paper, we estimated the relationships between the stability of the phytoplankton community and the lake trophic status. We further tested how common is the phenomenon of a higher degree of stability of the total phytoplankton community in relation to the stability of the phytoplankton structure.
Table 1 The major limnological characteristics of the Naroch Lakes during 1968–2014. Average and Min–Max in brackets. Indices
Batorino
Myastro
Naroch
Surface area, km2 Volume, mln. m3 Average depth, m Water retention time, yr Secchi depth, m Total P, μg L−1 Total N, mg L−1 Chlorophyll a, μg L−1 Trophic State Index (avg. 1968–1977)
6.3 18.7 2.4 1.0 1.1 (0.6–1.6) 51 (22–107) 1.2 (0.6–2.3) 30.1 (2.8–164) 66.8 (60.5–72.6)
13.1 70.1 5.4 2.5 3.2 (1.4–4.7) 39 (23–71) 0.9 (0.4–1.5) 12.4 (0.5–63.3) 60.7 (54.6–66.6)
79.6 710.0 8.9 10–11 6.0 (3.6–8.2) 18 (11–44) 0.7 (0.3–1.2) 3.6 (0.3–12.8) 42.3 (39.6–44.6)
spectrum of uses (in comparison with lakes Batorino and Myastro), and therefore this lake has a significant esthetical and recreational potential. Much of the nutrient and pollution load from the watershed is retained in the upper lakes of the system: Batorino and Miastro. Some limnological parameters of the Naroch Lakes are presented in Table 1. The Naroch Lakes were objects of intense hydroecological studies that were based on an extensive monitoring program initiated in the 1960’s (Winberg, 1985; Ostapenya, 1989). The trophic state of the Naroch Lakes ranged from oligo-mesotrophic (Lake Naroch) to eutrophic (Lake Batorino), and underwent remarkable changes during the studied period (1968–2012). The unusual phenomenon of deeutrophication (oligotrophication) in these lake ecosystems was attributed to drastic changes in the economic activities in their watershed, which between 1980 and 1990 resulted in a 45% decrease of the external phosphorus loading and in an invasion of the mollusk Dreissena polymorpha between 1992 and 2005 (Mikheyeva and Lukyanova, 2006; Ostapenya, 2014). The main features of the oligotrophication shift in the Naroch Lakes were as follows: a drop in the nitrogen and phosphorus concentrations, an increase in the water transparency (as Secchi Depth), a decrease in chlorophyll concentration and an increase of the relative significance of macrophytes in the primary production (Ostapenya et al., 2011). Ostapenya (2014) summarized the changes in the Naroch Lakes ecosystem using the term “benthification”, meaning “… an increase in the importance of benthic processes following increased water clarity promoted by nutrient reduction and Dreissena introduction”.
2. Material and methods 2.1. The Naroch Lakes (Belarus) This system of interconnected lakes includes Lakes Batorino, Myastro and Naroch (Fig. 1). These polymictic lakes are located in the north-west part of Belarus, in the glacial landscape. Lake Batorino, the first lake in the system, is a shallow eutrophic waterbody that connects to the mesotrophic Lake Myastro through a relatively narrow (5–6 m) channel. Meso-oligotrophic Lake Naroch is the largest waterbody of Belarus. The water quality of Lake Naroch is high enough for a wide
2.2. The phytoplankton community of the Naroch Lakes The long-term dynamics and structure of the Naroch Lakes phytoplankton have been described in detail (Mikheyeva and Lukyanova, 2006; Mikheyeva and Zhukova, 2015). During the studied period
Fig. 1. The Naroch Lakes and their location in Belarus.
482
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Chrysophyta biomass, from zero values in 1968–1977 to 14–17% of the total biomass, was a clear sign of the destabilization of the phytoplankton community. The major patterns of the evolution of the Naroch Lakes phytoplankton throughout different periods are described in detail in the “Results” section.
Table 2 The steps in the quantification of Naroch Lakes phytoplankton stability. Ecological unit
Phytoplankton community of Lakes Naroch, Myastro and Batorino Biomasses (B ) of the individual taxonomic groups: • Bacillariophyta, Cyanophyta, Chlorophyta, Cryptophyta,
State variables (SV)
i
Reference period Temporal and spatial scales Ecological distance
3. Methodology
Chrysophyta The sum of their biomass (TotB) 1968–1977 Ice free seasonal biomass mean average for the 0–5 m in Lake Batorino, 0–9 m in Lake Myastro and 0–16 m for Lake Naroch
•
EuDi = ABS ⎜⎛ ⎝
In this study, we followed the approach for quantification of stability (EuD-approach) suggested by Parparov et al. (2015). The data used for the current study (state variables) included the mean biomass (Bi), of each of the five main taxonomic groups, for the vegetative (ice-free) period (May–October), as well as their sum (TotB) documented in the Naroch Lakes during the period 1968–2012. There was no regular data for Lake Batorino during the period of 1984–1990. The water samples for the algae counts were collected monthly at specific monitoring points during from six depths (0.5, 3, 6, 8, 12 and 16 m) in Lake Naroch, four depths (0.5, 4, 7 and 9 m) in Lake Myastro and three depths (0.5, 3 and 5 m) in Lake Batorino. The water samples were then mixed in such a way that the water volume of each level (depth) in the mixed sample was proportional to the fraction of the level in the total water volume according to the lake bathymetry (Mikheyeva and Lukyanova, 2006). The state of the algal community was defined as a set of the individual biomass values at time t. The phytoplankton stability was considered to be identical to the constancy of the community state, i.e., the ability to sustain its structure in relation to the respective reference state, where biomasses of individual groups and their sum (Bi and TotB) were used as their reference states (RefBi and RefTotB; Grimm and Wissel, 1997; Reynolds, 2006). The main elements of this approach are summarized in Table 2 (modified from Parparov et al., 2015; Parparov and Gal, 2016). The changes in the state of the phytoplankton were studied based on the temporal variations in the biomass of each taxonomic group as well as the total phytoplankton biomass. The time step used for the stability calculations was a year. However, as mentioned above, the annual mean values were calculated for the vegetation (ice free) period (May–October). We assumed that the phytoplankton biomass variability observed during the reference period corresponded to the “natural variability”. The lower and upper variability limits of the individual groups and the total phytoplankton biomass during the reference period were defined as the 5th and 95th percentiles of the respective biomass time series during the reference period. The stability index values were defined (Eqs. (3) and (5) in Table 2) as an inverse function of the standardized Euclidean Distance (EuDi) of
Bi − Ref (B )i ⎞ ⎟ STD (B )i ⎠
Standardized ecological distance (1) EuD [TotB] = ABS
AggrEuD =
∑
TotB − Ref (TotB ) STD (TotB ) (EuDi)2
(1a) (2)
AggrEuD = (EuDi )2 Where i represents the biomass of the individual phytoplankton groups Stability Indices (SI)i 0 < SI < 1
(SI ) i =
1 1 + EuDi
Combined stability index
SI [Comb] = Total community stability index
(3)
1 1 + AggrEuD
(4)
1 1 + EuD [TotB] (5) The LSL value was defined as the 5th percentile of the stability index value time series during the reference period. If (SI/LSL) > 1, defines a stable state of the ecological unit; If (SI/LSL) < 1, defines a non-stable state of the ecological unit
SI [TotB] = Lower Stability Level (LSL) LSL-normalized stability index (nSI = SI/ LSL)
(1968–2012), the phytoplankton of the Naroch Lakes consisted of five major taxonomic groups: Bacillariophyta (diatoms, the dominating group in terms of algal biomass in Lakes Naroch and Myastro), Cyanophyta (cyanobacteria, the dominating group in terms of algal biomass in Lake Batorino), Chlorophyta (green algae), Cryptophyta and Chrysophyta (golden algae). On average, these five taxonomic groups comprised about 95% of the total algal biomass. The drastic increase in
Table 3 The dynamics of the decadal average biomass of the individual taxonomic groups (mg L−1) of the Naroch Lakes’ phytoplankton during different periods: I – 1968–1977 (the reference period), II – 1978–1987, III – 1993–2002, IV – 2003–2012. The variability limits: the 5th and 95th percentiles are shown in brackets. Cyanophyta
Cryptophyta
Bacillariophyta
Chrysophyta
Chlorophyta
TotB
Naroch I II III IV
0.11 0.18 0.13 0.35
(0.06–0.19) (0.09–0.33) (0.03–0.29) (0.14–0.66)
0.25(0.14–0.38) 0.13(0.03–0.26) 0.14(0.06–0.24) 0.36(0.18–0.58)
0.39(0.22–0.72) 0.66(0.36–1.27) 0.20(0.04–0.35) 0.24(0.10–0.47)
0.02(0.00–0.11) 0.20(0.04–0.34) 0.12(0.03–0.24) 0.18(0.05–0.42)
0.21(0.01–0.76) 0.06(0.02–0.12) 0.01(0.00–0.04) 0.04(0.00–0.13)
1.04(0.67–1.75) 1.36(0.73–2.22) 0.62(0.32–0.84) 1.26(0.85–1.57)
Myastro I II III IV
2.54 1.56 0.20 0.52
(0.51–7.77) (0.67–3.12) (0.02–0.59) (0.05–1.13)
1.78(0.32–3.55) 0.58(0.24–1.08) 0.24(0.12–0.49) 0.49(0.17–0.74)
2.44(1.33–5.15) 4.82(2.43–8.81) 0.90(0.22–2.25) 1.78(0.17–5.33)
0.00(0.00–0.00) 0.45(0.06–1.14) 0.21(0.03–0.49) 0.44(0.04–1.56)
0.55(0.34–0.86) 0.69(0.41–0.93) 0.02(0.00–0.03) 0.32(0.01–1.22)
7.44(3.75–13.0) 10.18(5.70–15.68) 1.62(0.82–3.32) 3.68(2.11–6.92)
12.45 (5.87–19.90)
0.75(0.10–1.59)
4.91(1.76–10.80)
0.00(0.00–0.00)
2.83(1.13–5.40)
21.93(9.90–32.88)
7.09 (3.39–10.41) 2.74 (0.39–6.31)
0.14(0.06–0.25) 0.41(0.22–0.69)
2.39(0.59–4.80) 3.19(1.52–5.73)
0.56(0.19–1.70) 1.24(0.48–3.15)
0.69(0.20–1.41) 0.62(0.26–1.33)
11.07(4.64–17.09) 8.81(3.85–14.27)
Batorino I IIa III IV a
No data.
483
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
the use of different percentile values: 1st, 5th, 10th, 15th and 20th percentile. The LSL-normalized values of the stability indices that were larger than unity indicated that the population or community was stable, while the values smaller than unity indicated a non-stable status of the phytoplankton (Parparov et al., 2015). In this article, further considerations were based mostly on the LSL-normalized (hereafter “normalized”) stability index values (nSI). Preliminary estimates performed for different time scales showed that the decadal scale indicates the main features and trends concerning stability quantification. Therefore, in order to test the EuD-approach, we examined the main features of the temporal dynamics of the 10-yr mean of the state variables and the normalized stability indices during four decadal periods: 1968–1977 (reference period), 1978–1987, 1993–2002 and 2003–2012. The Student t-test (Zar, 1998) was used to test the statistical hypotheses indicating the significance of destabilization events, for instance, the hypothesis that the normalized SI decadal averages are smaller than unity (n = 10, P < 0.05). The lake trophic state was quantified by the lake Trophic State Index (TSI; Carlson, 1977; Adamovich et al., 2016): TSI = (TSISD + TSITP + TSIChl)/3
(6)
where TSISD,TSITP and TSIChl are individual trophic state indices calculated for the Secchi Depth (SD), total phosphorus (TP) and chlorophyll a (Chl) concentrations as follows: TSISD = −14.39·ln(SD) + 59.91
(6a)
TSITP = 14.43·ln(TP) + 4.15
(6b)
TSIChl = 9.76·ln(Chl) + 30.91
(6c)
The TSI natural variability limits were estimated by the 5th and 95th percentiles calculated for the reference period time series. The relationship between the stability indices of the total phytoplankton community and structure was estimated by the ratio of SI [TotB]/SI[Comb]. The trophic status effects on the phytoplankton stability were examined by regressing SI[Comb] and SI[TotB] versus TSI. We also regressed the changes in the combined SIs versus the changes in the TSI. The changes in the aggregated SIs (SI[Comb], SI[TotB]) and TSI were calculated based on their reference values, as follows: Fig. 2. The dynamics of the 10-yr annual average values of the Naroch Lakes phytoplankton biomass. The numbers above the stacked columns represent the values of the total phytoplankton biomass. The dotted horizontal lines represent the limits of the natural variability (the 5th and 95th percentiles calculated for the reference period) of the total phytoplankton biomass.
ΔTSI = (TSI)ref − (TSI)i
(7a)
ΔSI = (SI)ref − (SI)j
(7b)
where indices “ref” correspond to the reference period, while indices “i” and ‘j’ span from 2 to 4, and indicate the decadal periods, i.e., 2–1978–1987, 3–1993–2002, 4–2003–2012. Similar to the Lake Kinneret case study (Parparov and Gal, 2016), we used the qualitative estimates of the stability of the Naroch Lakes phytoplankton (Mikheyeva and Lukyanova, 2006; Ostapenya et al., 2012; Ostapenya, 2014) as an important criterion for the validation of the EuD-approach. The respective data obtained for Lake Kinneret (Parparov et al., 2015; Parparov and Gal, 2016) were used for the analysis of the study results.
the state variable (i.e., the biomass of the individual algae group or total phytoplankton biomass) from the respective reference state (RefBi) at a given point in time, t (Kindt and Coe, 2005; Greenacre, 2008). The stability of the phytoplankton was quantified by comparing the stability indices using two different aggregating schemes. The first scheme used the combined stability index (SI[Comb]), which was calculated based on the stability indices of the individual phytoplankton (SIi) taxonomic groups (Eq. (4) in Table 2). The combined stability index characterizes the stability associated with the changes to the structure of the community. The second scheme used a total community stability index, calculated based on the total abundance of phytoplankton (SI[TotB], Eq. (5) in Table 2). In order to determine the stability status of an ecological unit and to allow a direct comparison between the stability index values of different ecological units, the SI values were normalized to a lower stability limit (LSL). The LSL value was defined as the 5th percentile of the stability index values calculated for the reference period. In order to estimate the effect of the selected percentile value, for SI[Comb] and SI [TotB], we calculated the sensitivity (relative to the 5th percentile) to
4. Results 4.1. Establishing the reference period Defining the reference period is not straightforward, as there is no clearly formulated procedure for its delineation. Therefore, the establishment of the reference period is essentially subjective, depending on the study subject and the researcher’s interest. In this study, we selected the reference period that corresponded to the state of the community that existed in the past and was considered minimally disturbed 484
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 3. The dynamics of the normalized stability indices (nSI) of the individual phytoplankton group in Naroch Lakes (the 10-yr annual average and STD values). The horizontal dashed lines indicate the nSI value equal to unity, and thus separates the stable and non-stable states of the ecological unit. The asterisks above the column indicate values significantly smaller than unity (at P < 0.05).
4.2. Changes to the biomass, species composition and temporal dynamics of the Naroch Lakes phytoplankton
(Grimm and Wissel, 1997; Donohue et al., 2013). We used the decadal dynamics of the phytoplankton biomass and the Naroch Lakes trophic state as an argument for determining the period to be defined as the reference period. The quantification of phytoplankton stability in three different lakes indirectly assumes that the stability of all three lakes will be estimated according to the same reference period; otherwise, the established stability estimates will be incomparable. When selecting the reference period we had to balance the requirements according to the relative constancy of the phytoplankton structure as well as the constancy of the lake trophic state. Therefore, we selected the period of 1968–1977 as the “reference period” for the phytoplankton of the Naroch Lakes based on the following reasons:
Over a period of more than 40 years (from 1968 until 2012), the average biomass of all five phytoplankton taxonomic groups in the Naroch Lakes underwent notable changes. The following major features of the changes to the biomass of the entire phytoplankton as well as the individual taxonomic groups (Mikheyeva and Lukyanova, 2006; Mikheyeva and Zhukova, 2015) could affect the stability of the Naroch Lakes phytoplankton (Fig. 2 and Table 3):
• the most common feature is the considerable increase in the biomass and relative contribution of Chrysophyta; • the increase in the relative contribution of Cyanophyta in Lake
• Major qualitative conclusions regarding changes to the stability of • •
the Naroch Lakes phytoplankton were based on comparisons with the period of 1968–1977 (Mikheyeva and Lukyanova, 2006; Ostapenya, 2014). Immediately after that period, there were drastic changes in the phytoplankton communities of all three lakes (Table 3). This is the period for which a more complete database exists (note the data deficiency for Lake Batorino for the period of 1984–1990).
• • • 485
Naroch was accompanied by a decrease in this algae group biomass in Lakes Myastro and Batorino; irregular though considerable changes in the total biomass of the phytoplankton in all three studied lakes. The following changes to the decadal average algal biomass should be noted as well: the decrease in the Cyanophyta biomass in Lakes Myastro and Batorino after 1982–1984, below the lower “natural” variability (the 5th percentile) level. In contrast, the dynamics of the
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 5. The scatter plot and best fit regression lines of the relationships between the normalized annual average values of the combined (nSI[Comb]) and total community (nSI[TotB]) stability indices for the Naroch Lakes (A) and for Lake Kinneret (C) phytoplankton. B – the same as in A for the decadal average values in the Naroch Lakes.
Fig. 4. The decadal average ( ± std) values of the normalized SI[Comb] and SI[TotB] values (nSI[Comb] and nSI[TotB]) of the Naroch Lakes phytoplankton. The horizontal dashed lines indicate the value of the nSI values equal to unity, and thus separates the stable and non-stable states of the ecological unit. The asterisks above the column indicate values significantly smaller than unity (at P < 0.05).
•
Table 4 Values of the lower stability limits calculated as smaller percentiles for SI[Comb] and SI [TotB]. In brackets: the error (relative to the 5th percentile value). Percentiles
1% 5% 10% 15% 20%
Lake Naroch
Lake Myastro
Lake Batorino
SI [Comb]
SI [TotB]
SI[Comb]
SI[TotB]
SI[Comb]
SI [TotB]
0.211 (18.2) 0.257 (0.0) 0.316 (22.7) 0.341 (32.6) 0.357 (38.8)
0.423 (5.3) 0.447 (0.0) 0.476 (6.6) 0.499 (11.8) 0.521 (16.5)
0.322 (3.1)
0.496 (1.6)
0.301 (1.9)
0.332 (0.0)
0.505 (0.0)
0.307 (0.0)
0.345 (3.9)
0.515 (2.0)
0.315 (2.4)
0.354 (6.4)
0.521 (3.3)
0.320 (4.3)
0.361 (8.6)
0.527 (4.4)
0.325 (5.9)
0.398 (9.3) 0.438 (0.0) 0.489 (11.6) 0.503 (14.7) 0.506 (15.4)
•
•
Cyanophyta biomass in Lake Naroch showed a weak trend of increase in the 1980′s and after 2000, above the 95th percentile of the respective reference period; the periods of increase of the Bacillariophyta biomass (above the level of 95% in Lakes Naroch and Myastro) were followed by a biomass drop below the 5th percentile level (especially in Lake Myastro, Fig. 2); during the reference period, the Chrysophyta biomass in all the Naroch Lakes was close to zero (Table 3 and Fig. 2). Since the late 1970s, the biomass of Chrysophyta underwent a drastic increase, up to well above the upper limit of the “natural” variability, indicating a drop of stability of this taxonomic group. A particularly impressive change in the Chrysophyta biomass was detected in Lake Myastro: within a year (between 1978 and 1979) the biomass of this algal group increased from zero values in 1968–1977 to 0.33 mg L−1 (on average, for 1978–2014, Fig. 2, Table 3). In 2003–2012, the relative contribution of Chrysophyta to the total algae biomass in Lake Myastro comprised 14–17% of the total algae biomass; the Chlorophyta biomass in all three lakes decreased by almost an order of magnitude compared to the reference period. The total biomass of the phytoplankton in the Naroch Lakes during
486
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 6. The dynamics of the ratio of nSI[TotB]/nSI[Comb] for the Naroch Lakes phytoplankton (the 10-yr annual average and STD values). The horizontal dashed lines indicate the value of the ratio equal to unity. The dynamics of a similar ratio for Lake Kinneret are shown as well.
[TotB]) are of special interest. The decadal average values of these indices dropped in varying degrees, though more gradually for SI[TotB] (Fig. 4). The almost twofold decrease of SI[Comb] during the 40 years that passed since its reference state represents a quantitative estimate of the change in the stability structure of the Naroch Lakes phytoplankton. The decadal dynamics of the SI[TotB] showed a relatively smaller decrease in the stability. The decadal dynamics of the normalized SI[Comb] and SI[TotB] values, (Fig. 4) showed that there were no periods of prolonged destabilization in Lakes Naroch and Batorino. The normalized SI[Comb] and SI[TotB] values calculated for these lakes were close to, or exceeded, unity; thus the structure and abundance of the phytoplankton in these lakes remained within the stability limits during almost the entire period from 1968 to 2012. Statistically significant destabilization events of the phytoplankton structure (SI[Comb] < 1, P < 0.05) were recorded only for the phytoplankton structure of Lake Myastro during 1993–2012 (Fig. 4). The destabilization of the total community stability in Lake Myastro was recorded only in 1993–2002 (SI[TotB] < 1 at P < 0.05). During the rest of the periods, the decadal average values of SI[TotB] obtained for this lake were not less than unity.
1968–2012 varied within wide limits: the 10-year average values of the TotB ranged from 0.62 to 1.36 mg L−1 in Lake Naroch, from 1.62 to 10.18 mg L−1 in Lake Miastro, and from 8.81 to 21.9 mg L−1 in Lake Batorino (Fig. 2). In Lake Naroch, during most part of the observation period the TotB varied within the “natural” limit of its variability (Fig. 2, Table 3). In Lake Myastro, the TotB dropped below the “natural” limit in 1992–2008; after that, there was a trend towards a recovery of the TotB in this lake (Fig. 2 and Table 3).
4.3. Quantifying the stability of the Naroch Lakes phytoplankton 4.3.1. Temporal dynamics of the stability indices The phytoplankton stability was quantified at three different levels: the stability of the five individual phytoplankton groups, the combined stability of the five groups (SI[Comb], Eqs. (3) and (4)) and the stability of the total phytoplankton community (SI[TotB], Eq. (5)). Compared to the reference period values, the normalized stability indices of the individual algal groups (averaged for 10-year periods) in all three lakes decreased (Fig. 3), with two exceptions: the SI values of Cryptophyta and Bacillariophyta in Lake Batorino slightly increased. Note, that the decrease in the SI[Cyano] in the last decade was caused by contrasting reasons: by the increase of the Cyanobacteria biomass above the upper natural variability level in Lake Naroch, while in Lakes Myastro and Batorino the Cyanobacteria biomass decreased below the lower natural variability level of the biomass of this alga group during 1992–2011 (see also Table 3). The dynamics of most of the normalized stability indices did not indicate any destabilization events (i.e., the LSL-normalized SI values were larger than or equal to unity). The statistically significant (at P < 0.05) decrease of the normalized SI was associated with the drastic increase of the Crysophyta biomass and the decrease of the Chlorophyta biomass in Lakes Myastro and Batorino during the last three decades (Figs. 2 and 3). The dynamics of the aggregated stability indices (SI[Comb] and SI
4.3.2. The effect of the selected lower percentile values on estimating of the lower stability limits for SI[Comb] and SI[TotB] In order to study the sensitivity of the results to the selected value of the lower stability limits (LSL) we tested a range of percentile values. Thus we calculated the aggregated stability indices (SI[Comb] and SI [TotB]), for the Naroch lakes, based on the following range of percentile values: 1%, 5%, 10%, 15% and 20%. The resulting stability indices values indicated relatively limited variability in the SI values (Table 4). For Lakes Myastro and Batorino, the errors associated with using of the 1st and 10th percentiles did not exceed 10%; relatively higher relative errors were calculated for SI[Comb] in Lake Naroch. The higher errors were calculated for the 15th and 20th percentiles, however the 487
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
P < 0.005). A statistically insignificant relationship was obtained between the annual average values of SI[Comb] and SI[TotB] for the Lake Kinneret phytoplankton (Fig. 5C). It should especially be noted that for the Naroch Lakes phytoplankton, the SI[TotB] values were larger (on average, by 20%) than SI [Comb] (and thus the ratio of SI[TotB]/SI[Comb] exceeded unity) during each of the 10-year periods for each of the Naroch Lakes (Fig. 6). It means that at the decadal scale, in the Naroch Lakes, the stability of the total phytoplankton community exceeded the stability associated with the phytoplankton structure. 4.4. The relationship between the Naroch Lakes trophic status and the phytoplankton stability After the increase (though within the natural variability limits) in 1968–1987, the TSI decreased (even persistently below the natural variability level, in Lake Myastro), indicating the oligotrophication/ deeutrophication of the Naroch Lakes during 1993–2012 (Fig. 7). Despite the considerable range of the TSI values observed for the Naroch Lakes (36 ≤ TSI ≤ 67), the scatter plots did not indicate any relationship between the aggregated stability indices (SI[Comb], SI [TotB]) and TSI (Fig. 8A & B). Note that the values obtained for Lake Kinneret are within the data “cloud” received for the Naroch Lakes. However, the changes of both aggregated stability index values calculated for the Naroch Lakes (see Eq. (7a,b)) for the decadal average values positively correlated quite well with the corresponding changes of the TSI (Fig. 8 C & D). ΔSI[Comb] linearly correlated with ΔTSI (R2 = 0.75, P < 0.05), while the regression of ΔSI[TotB] versus ΔTSI was non-linear and less statistically significant (R2 = 0.54, P < 0.1). The data for Lake Kinneret were outside of the main tendencies obtained for the Naroch Lakes data. Note, that the decadal average TSI values for Lake Kinneret had a much narrower range compared to the Naroch Lakes: between 47.9 and 49.3. 5. Discussion The results of our study performed using the EuD-approach indicated a decrease of the Naroch Lakes phytoplankton stability. Our results do not contradict the qualitative conclusions by Mikheyeva and Lukyanova (2006) and Ostapenya (2014) about a decrease of the Naroch Lakes phytoplankton stability based on analysis of the longterm changes to the dynamics of the biomass and the structure of the Naroch Lakes phytoplankton (Fig. 2) together with the drop in the TSI values (Fig. 7). We consider the correspondence between previous qualitative estimates and our results as indirect validation of EuD-approach for quantifying the stability of the phytoplankton communities in different lakes. Moreover, the implementation of the EuD-approach made the qualitative estimates more extensive and precise, and allowed us to answer some basic questions outlined by the study objectives. The lowering of the Naroch Lakes stability was indicated by a decline in the decadal average values for both aggregated stability indices: SI[Comb] and SI[TotB] (Fig. 4). On the decadal scale, the stability of the phytoplankton structure (as SI[Comb]) during the period 1978–2012 dropped compared to the reference period by 32%, 54% and 53% for Lakes Naroch, Myastro and Batorino, respectively. These values are comparable with the SI[Comb] decrease obtained for Lake Kinneret (by 50%, according to Parparov et al., 2015). However, the normalized SI[Comb] values were significantly (P < 0.05) smaller than unity only for Lake Myastro during 1993–2012, thus indicating a destabilization of the phytoplankton structure in this lake. The normalized SI[TotB] values did not fall, statistically, below unity (excluding the decadal average value in Lake Myastro in 1993–2002, Fig. 4). Therefore, we conclude that on the decadal scale, the stability of the total phytoplankton community of the Naroch Lakes decreased in 1978–2012, but remained within its natural stability limits. At the same time, our approach revealed some peculiarities of the
Fig. 7. The temporal dynamics of the trophic state index (TSI) in the Naroch Lakes (decadal average ± std values). Horizontal lines indicate the limits of the TSI natural variability.
use of these percentiles for the LSL estimates would be associated with unwanted narrowing of the LSL estimates: 30 and 40% of the LSL values would be lost, respectively. The use of the 1st and 10th percentile would provide close estimates of SI[Comb] and SI[TotB], but we declined their use because the use of the 1st percentile would allow including of almost all stability index values into the reference period while the use of the 10th percentile would require excluding of 20% of the SI values from the reference state estimate.
4.3.3. The relationship between aggregated stability indices: SI[Comb] and SI[TotB] The annual average values of the aggregated stability indices (SI [TotB] and SI[Comb]) correlated quite well (Fig. 5A, R2 = 0.37, n = 118, P < 0.01); a stronger (though not linear) correlation was obtained for the decadal average values (Fig. 5B, R2 = 0.82, n = 11, 488
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Fig. 8. A & B: The scatter plot of SI[Comb] and SI[TotB] versus TSI values for the Naroch Lakes and Lake Kinneret (outlined); decadal average values. C and D: The scatter plots and best fit regression lines for the relationships between the changes in the aggregated stability indices ΔSI[Comb] and ΔSI[TotB], and the changes of the TSI (ΔTSI). The data for Lake Kinneret are encircled.
previously regarding Lake Kinneret phytoplankton, was observed in the three other lakes. Thus, we might assume the existence of a generality of a relatively larger stability of the total phytoplankton community in comparison with the stability of its structure. For Lake Kinneret, the exceedance of SI[TotB] over SI[Comb] was interpreted (Parparov et al., 2015; Parparov and Gal, 2016) as evidence of the emergent properties of the phytoplankton community stability, equivalent to the existence of the homeostatic mechanisms that support the stability of the total phytoplankton community despite the changes in its internal structure. According to the emergence principle, the properties of the higher hierarchical levels are distinct from the properties of the lower hierarchical levels (Webster, 1979; Justus, 2008; Jørgensen and Nielsen, 2013). The emergence principle itself is a kind of a hypothesis, which does not explain the nature of the homeostatic mechanisms responsible for the larger stability of the higher hierarchical levels. We should note that for Lake Kinneret, the exceedance of SI[TotB] over SI[Comb] was not observed for the zooplankton community: the stability index of the total zooplankton community was approximately equal to the stability index of its structure (Parparov and Gal, 2016). Further studies are needed to clarify the generality and limitations of this phenomenon. The quantification of the phytoplankton stability for the Naroch Lakes and Lake Kinneret that have different trophic status (36 ≤ TSI ≤ 67) allowed us to test the potential effects of the trophic status on the stability of the producer community. The regression between the phytoplankton aggregated stability indices and TSI did not indicate any relationship between these variables (Fig. 8 A & B). However, such a relationship was revealed for the changes of the aggregated stability indices (ΔSI[Comb], ΔSI[TotB]) and ΔTSI (Fig. 8 C & D), though for the Naroch Lakes only. We interpret these results as indicating the relative independence of the phytoplankton stability in relation to lake trophic status. However, at the decadal scale, the shifts in the stability of the phytoplankton structure and abundance in the Naroch Lakes might be induced by the shifts that occur in their trophic
Naroch Lakes phytoplankton in regards to the quantification of its stability. The Naroch Lakes had no period of relatively limited variability of the algal biomass, as seen in the Lake Kinneret phytoplankton (Parparov et al., 2015). The variability of the individual groups and of the total phytoplankton biomass in the Naroch Lakes during the established reference period (1968–1977) was relatively large (sometimes it ranged by an order of magnitude, e.g., Cyanophyta in Lake Myastro, Table 2). Therefore, due to the large variability observed during the reference period, the “natural variability limits” of the algal biomass in the Naroch Lakes (defined by the 5th and 95th percentiles obtained for the respective reference periods) were relatively wide. An analysis of the decadal dynamics of the individual stability indicated that despite trends of lowering of algal biomasses (or an increase Cyanophyta biomass), only Chrysophyta in Lakes Myastro and Batorino showed a statistically significant persistent (about three decades) destabilization (Fig. 3). For all three Naroch Lakes, the LSL-normalized values of the total community stability index (nSI[TotB]) were systematically larger (by appr. 20%) than nSI[Comb] (Fig. 6). For Lake Kinneret, the SI[TotB] was more than twice the combined stability index (SI[Totb]/SI [Comb] ≈ 2.2, Parparov et al., 2015). Note that in Lake Kinneret the exceedance of SI[TotB] over SI[Comb] was a natural consequence of the relatively higher stability of the total phytoplankton biomass, while in the Naroch Lakes the total biomass variability was comparable with the variability of the individual algal group biomass (Table 3 and Fig. 2). A comparison of the data acquired for the Naroch Lakes and Lake Kinneret indicates that there are two types of dynamics for the aggregated stability indices: 1. Intercorrelated changes of the SI[Comb] and SI[TotB] observed in the Naroch Lakes (Fig. 5A), and 2. Relatively large changes in the SI[Comb], which were independent of smaller changes in the SI[TotB] as seen in Lake Kinneret (Fig. 5C). However, despite the distinction between these two types of stability index dynamics, the exceedance of SI[TotB] over SI[Comb], mentioned 489
Ecological Indicators 81 (2017) 481–490
T.M. Mikheyeva et al.
Statistics, Stanford University Fall, 2008, STA 254. http://www.econ.upf.edu/ ∼michael/stanford/. Grimm, V., Wissel, C., 1997. Babel or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia 109, 323–334. Groffman, P.M., Baron, J.S., Blett, T., et al., 2006. Ecological thresholds: the key to successful environmental management or an important concept with no practical application? Ecosystems 9, 1–13. Holling, C.S., 1996. Engineering resilience vs ecological resilience. In: Schultze, P.C. (Ed.), Engineering Within Ecological Constraints. National Academy Press, Washington DC, pp. 31–43. Jørgensen, S.E., Nielsen, S.N., 2013. The properties of the ecological hierarchy and their application as ecological indicators. Ecol. Ind. 28, 48–53. Justus, J., 2008. Complexity, diversity and stability. Chapter 18. In: Sahotra, S., Plutyns, S. (Eds.), A Companion to the Philosophy of Biology. Blackwell Publishing, pp. 321–350. Kindt, R., Coe, R., 2005. Tree Diversity Analysis. A Manual and Software for Common Statistical Methods for Ecological and Biodiversity Studies. World Agroforestry Centre (ICRAF), Nairobi 197pp. Ludwig, D., Walker, B., Holling, C.S., 1997. Sustainability, stability, and resilience. Conserv. Ecol. 1, 7. Available from the Internet. URL: http://www.consecol.org/ vol1/iss1/art7/. Mikheyeva, T.M., Lukyanova, E.V., 2006. The direction and character of multi-year changes of phytocoenotic structure and the indices of quantitative development of phytoplankton communities of the Naroch lakes in the course of trophic state evolution. Izv. Samar. Nauch. Tsentra RAN 8, 125–140 (in Russian). Mikheyeva, T.M., Zhukova, T.V., 2015. Conclusion. In: Mikheyeva, T.M. (Ed.), The Bulletin of Ecological State of Lakes, Naroch, Myastro. Belatus State University, Minsk, pp. 105–107. Odum, E.P., 1971. Fundamentals of Ecology, third edn. W.B. Saunders, London 360 pp. Ostapenya, A.P., Zhukova, T.V., Mikheyeva, T.M., 2011. Bentification as a stage in the Naroch Lakes evolution. Proc. Belarus State Univ. 2 (2), 62–66. Ostapenya, A.P., 1989. Seston and Detritus as Structural and Functional Components of Water Ecosystems. Ref. Diss. Zoological Institute, Minsk 42 pp. (in Russian). Ostapenya A.P., 2014. Benthification of freshwater lakes: exotic mussels turning ecosystems upside down. Chapter 36. In: C. M. Mayer [et al.]//Quagga and Zebra mussels. Biology, impact and control/Edited by T. F. Nalepa, D. W. Schloesser. Second edition. CRC Press London, New York, p. 575–585. Parparov, A., Gal, G., 2012. Assessment and implementation of a methodological framework for sustainable management: Lake Kinneret as a case study. J. Environ. Manage. 101, 111–117. Parparov, A., Gal, G., 2016. Quantifying ecological stability: from community to the lake ecosystem. Ecosystems. http://dx.doi.org/10.1007/s10021-016-0090-z. (accepted for publication). Parparov, A., Gal, G., Zohary, T., 2015. Quantifying the ecological stability of a phytoplankton community: the Lake Kinneret case study. Ecol. Indic. 56, 134–144. Parparov, A.S., 1990. Some characteristics of the community of autotrophs of Lake Sevan in connection with the eutrophication. Hydrobiologia 191, 15–21. Reynolds, C.S., 2006. The Ecology of Phytoplankton. Cambridge University Press 535 p. Oneida Lake: Long-term Dynamics of a Managed Ecosystem and Its Fishery. In: Rudstam, L.G., Mills, E.L., Jackson, J.R., Stewarts, D.J. (Eds.), American Fishery Society, Bethesda, Maryland 541 PP. Rykiel, E.J., 1985. Towards a definition of ecological disturbance. Aust. J. Ecol. 10, 361–365. Scheffer, M., Hosper, S.H., Meijer, M.L., Moss, B., 1993. Alternative equilibria in shallow lakes. Trends Ecol. Evol. 8, 275–279. Simonyan, A.A., 1991. Zooplankton ozera Sevan (Zooplankton of Lake Sevan, in rus). Izdvo AN Armenii, Erevan 299pp. Walker, B., Carpenter, S., Anderies, N., et al., 2002. Resilience management in socialecological systems: a working hypothesis for a participatory approach. Conserv. Ecol. 6 (1), 14. [online] URL: http://www.consecol.org/vol6/iss1/art14/. Webster, J.R., 1979. Hierarchical organization of ecosystems. In: Halfon, E. (Ed.), Theoretical System Ecology. Academic Press, New York, pp. 119–131. Winberg, G.G., 1985. The Ecological System of the Naroch Lakes. Minsk (in Rus.). 289 pp. Zar, J.H., 1998. Biostatistical Analysis. Prentice Hall International, INC, New Jersey p. 147. Zohary, T., Sukenik, A., Nishri, A., 2014a. Lake kinneret: current understanding and future perspectives. Chapter 27. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 417–438. Zohary, T., Yacobi, Y.Z., Alster, A., Fishbein, et al., 2014b. Phytoplankton. Chap. 10. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 161–190.
status. It is obvious that studying the relationship between trophic status and ecological stability requires further accumulation of relevant limnological information. However, in any case, it will be impossible to solve this scientific task without the quantification of stability. The approach of quantification of stability tested and implemented in this study might represent a useful tool for progressing the theory of ecological stability. 6. Concluding remarks
• The results of our study performed for the Naroch Lakes confirmed •
• •
that the approach for quantifying of the ecological stability developed previously for the Lake Kinneret phytoplankton, could be used successfully for lakes of different trophic status. Analyses performed on the dynamics of the stability index values of the Naroch Lakes phytoplankton (of the individual groups, combined and total community) indicated a tendency of decreasing phytoplankton stability in these lakes. Despite this tendency, the total community stability of the Naroch Lakes phytoplankton remained within its stability limits. On the other hand, the dynamics of the stability indices associated with the changes to the phytoplankton structure indicated a persistent destabilization of the phytoplankton community in Lakes Myastro (after 1987) and Batorino (after 1997). For Naroch Lakes, the total community stability exceeded the stability of the phytoplankton structure. Together with similar relation obtained earlier for Lake Kinneret, this observation allowed us to assume the generality of this phenomenon. Quantification of the phytoplankton stability in the Naroch Lakes for the first time allowed us to consider the potential effects of the lake trophic status on the stability of the phytoplankton community and to connect the shifts in the lake tropic status with the shifts in the phytoplankton stability.
Acknowledgments This study was partially supported by the Belarus Republican Foundation for Fundamental Research. We thank two anonymous Reviewers whose valuable comments greatly contributed to improvement of the manuscript. We are very thankful to L. Baumer for her valuable help in preparing and editing this manuscript. References Adamovich, B.V., Zhukova, T.V., Mikheyeva, T.M., Kovalevskaya, R.Z., Lukyanova, E.V., 2016. Long-term variations of the trophic state index in the Narochanskie lakes and its relation with the major hydroecological parameters. Water Resour. 43, 809–817. Carlson, R.E., 1977. A trophic state index for lakes. Limnol. Oceanogr. 22, 361–369. Donohue, I., Owen, L., Petchey, O.L., Montoya, M.L., et al., 2013. On the dimensionality of ecological stability. Ecol. Lett. 16, 421–429. Gal, G., Anderson, W., 2010. A novel approach to detecting a regime shift in a lake ecosystem. Methods Ecol. Evol. 1, 45–52. Gal, G., Hambright, K.D., 2014. Metazoan zooplankton. In: Zohary, T., Sukenik, A., Berman, T., Nishri, A. (Eds.), Lake Kinneret: Ecology and Management. Springer, Heidelberg, pp. 227–245. Goldman, C.R., 2008. Introduction. Five decades of environmental and social change at Lake Tahoe. In: Bachand, T. (Ed.), Lake Tahoe: A Fragile Beauty. Chronicle Books, San Francisco, pp. 14–18. Greenacre, M., 2008. Correspondence Analysis and Related Methods. Department of
490