1. Introduction
The growing negative human impact on the environment calls for continuous monitoring of its condition, for the assessment of which more bio-indication research is increasingly being used. At the end of the 20th and in the beginning of the 21st century, a number of researchers, such as Cabagna et al. [1] , Davis et al. [2] , Korzh et al. [3] , Lajmanovich et al. [4] , Zhelev [5] , have substantiated the successful application of haematological parameters of tailless amphibians (Order Anura) in biomonitoring. They are one of the first life-forms to react to changes caused by negative environmental factors, including anthropogenic ones.
Nowadays, despite the controversial attitude towards using them as good biomarkers, the changes in the values of haematological parameters in the marsh frog Pеlophylax ridibundus Pallas, 1771 are used successfully; one of the three species of Pelophylax kl. esculentus Linnaeus, 1758. They are very attached to a specific breeding basin (They rarely leave it and spend almost their entire lives in it), they reproduce quickly in large quantities and have an extensive range. This makes it possible to conduct research on large areas using the same signs, obtaining comparable results [6] -[9] . From Pelophylax kl. esculentus the marsh frog P. ridibundus is a widely spread species in Bulgaria being found throughout the country, P. esculentus Linnaeus, 1758 can be found in the north-along the Danube, while P. lessonae Camerano, 1882, despite its presumed presence, has not been reliably documented [10] .
The presence of alternative morphs of P. ridibundus by the presence or absence of a central light stripe on the back has been used since fairly recently as a bio-indicative marker for evaluation of the condition of the outer environment. Analysing the genetic nature of this morph in Ranidae Rafinesque-Schmaltz, 1814 family showed that the striata morph is a monogenic mutant. The dominant allele of the diallelic autosomal striata gene determines the presence of a stripe (complete dominance). Such mode of inheritance has been found for P. ridibundus [11] and Rana arvalis Nilsson, 1842 [12] . For the various species from the Ranidae family it has been found that in the individuals with striata morph the total level of the reduction-oxidation processes and the amount of haemoglobin are higher, and the sodium penetrability of the skin is twice lower than that of the maculata morph [13] . We found that in juvenile [14] and adult [5] [15] individuals of P. ridibundus of striata morph the hepaticsomatic index (HSI) was higher than that of maculate morph. Studies on the bioaccumulation of heavy metals in R. arvalis found that in animals of striata morph strontium values (Sr-90) are five times lower than those in animals of maculata morph [16] . According to these authors’ opinion, the relatively weak ability to bioaccumulate is associated with a higher frequency of occurrence in populations of R. arvalis, living in areas with natural and artificial geochemical anomalies in individuals of striata morph. According to the views of Sils and Vershinin [17] , the lower skin permeability in individuals of striata morph leads to an increasing role of pulmonary breathing, which, in turn, is linked with hemopoiesis intensification (the number of erythrocytes and the amount of haemoglobin get higher).
A new large-scale study was conducted in 2012 in the same biotopes 3 along the rivers Sazliyka and Topolnitsa in populations of P. ridibundus, in view of analysing a wider range of haematological parameters in individuals of both morphs striata and maculate.
We set the following objectives:
• Obtaining data on the main haematological parameters (including any seasonal variations) in adult, sexually mature individuals of both colour morphs (striata and maculata) in the populations of P. ridibundus, inhabiting biotopes with different levels of anthropogenic pollution in Southern Bulgaria.
• Assessment of the most informative haematological parameters that objectively reflect the changes in the functional status of animals in habitats with anthropogenic stress.
• Output of possible differences in the values of the parameters in individuals of both morphs in P. ridibundus in the water basins polluted with different types of toxins and assessment of their role in the survival and environmental adaptation of the morph to the peculiarities of the environment.
• Analysis of the possibilities of using these haematological parameters in individuals of both morphs of P. ridibundus as markers for assessing the ecological status of the environment in the system of biomonitoring.
A schematic overview of the work-flow showing the design of the study is present in Figure 1.
Figure 1. Schematic overview of the work-flow showing the design of the study.
2. Materials and Methods
2.1. The Area of Investigation
The samples were gathered during three seasons (spring, summer and autumn) of 2012 in three biotopes (for convenience labeled as A, B and C) located along the courses of two rivers in Southern Bulgaria: the river Sazliyka in its upper reaches near the village of Rakitnitsa (N 42˚19'58.8'', E 25˚31'1.2'', 200 m above sea level—biotope A), in the middle part in the region of the town of Radnevo, after it flows into the river Blatnitsa (N 42˚18'0'', E 25˚55'58.8'', 113 m above sea level-biotope B) and the river Topolnitsa near the village of Poibrene, below where the river Medetska enters the Topolnitsa water reserve (N 42˚30'0'', E 24˚0'0'', 511 m above sea level-biotope C). Physiographic map of the investigated biotopes are presented in Zhelev et al. [1]">8] .
2.2. Data from Physicochemical Analysis of the Water Ecosystems Studied
The average annual data relating to the water basins studied for 2012 are extracted from the reports of the Basin Directorate for water management—East Aegean Region—Plovdiv (http://www.bg-ibr.org) and the Ministry of the environment and water resources (data not shown). Physico-chemical monitoring is based on 21 indicators. Details of the physicochemical analysis for the period 2009-2011 in each of the biotopes are presented in Zhelev et al. [1]">8] . For Biotope A there is no data about anthropogenic pollution (all 21 indicators are standard for Category I (clean) and Category II (slightly polluted) water basins, under Order No. 7/8.8.1986 (State Gazette, No. 96.12.12.1986). In our study it is viewed as control. The remaining two biotopes B and C are polluted. In the biotope B the main pollution is of sewage-domestic type (higher Biologic Consumption of Oxygen (BOD: 17.2 mg O2/dm3, MAC (maximum admissible concentrations) for Category I, II and III (normally contaminated) respectively 5.0, 15.0 and 25.0 mg O2/dm3),), Nitrite-Nitrogen (N-NO2: 0.2 mg/dm3, MAC for Category I, II and III respectively 0.002, 0.04 and 0.06 mg/dm3) and Total Nitrogen (N 5.3 mg/dm3, MAC for Category I, II and III respectively 1.0, 5.0 and 10.0 mg/dm3)), while the biotope C is polluted with heavy metals (Iron-total (Fe): 0.20 mg/dm3, above 0.1 mg/dm3—very poor condition, Manganese (Mn): 0.33 mg/dm3, above 0.05 mg/dm3—very poor condition, Copper (Cu): 0.031 mg/dm3, above 0.022 mg/dm3—very poor condition, Lead (Pb): 0.015 mg/ dm3, above 0.0072 mg/dm3—very poor condition, Nickel (Ni): 0.09 mg/dm3, above 0.02 mg/dm3—very poor.
2.3. Subject of Study and Methods of Analysis
The subject of our study is the marsh frog P. ridibundus. To achieve maximum objectivity in the research and to avoid selectivity in terms of the number of individuals of the two morphs in the test group (30 animals per season), the catches were performed in the following way: once in the spring, summer and autumn of 2012 (the dates are listed below), in the evening after sunset, we spotlighted each of the habitats using 2 Petromax lanterns placed on the shore at a few square metres. Then in the spotlighted area we caught adult individuals randomly in the water (using fishing net) and on land-along the bank. (Snout-Vent Length > 60.0 mm) []">19] , there was no selectivity with respect to the morph. We caught altogether 270 individuals from the three biotopes and the animals were divided as follows: Biotope A [spring (24 April 2011)-striata = 14, maculata = 16; summer (17 July 2011)-s = 12, m = 18; autumn (4 October 2011)-s = 11, m = ]">19]; Biotope B [spring (28 April 2011)-s = 23, m = 7; summer (22 July 2011)-s = 21, m = 9; autumn (9 October 2011)-s = 22, m = ]">8]; Biotope C [spring (4 May 2011)-m = 5, s = 25; summer (28 July 2011)-m = 7, s = 23; autumn (17 October 2011)-s-24, m-]">6]. To prevent eventual recapture in spring and summer the animals were marked through taking individual photographs [20] .
The frogs were transported from the place of capture to the laboratory. The live frogs were anaesthetised with ether and blood (0.20 ml) was drawn by means of cardiac ventricular puncture using small heparinized needles (20 mm length) via heparinized haematocrit capillaries [1] . The erythrocyte (RBC) and leukocyte (WBC) count was determined according method of Wierord using a Burker chamber [21] . For dilution standardised Hayem solution was used for erythrocytes via Thoma pipettes, while for the leucocytes we used Turck’s solution. Dilution was carried out 200 times for the erythrocyte count and for 20 times for the leukocyte count. The haemoglobin concentration (Hb) was determined with the cyan-haemoglobin method-reading 540 nm [21] . The haematocrit value (PCV) was determined with heparinized haematocrit capillaries. Blood was centrifuged for 5 min at 3000 rpm constant-rotation, in thin-walled capillary tubes and the value obtained was read from the scale and recorded in L/l [22] . The derivative haematological parameters (MCH-mean cell haemoglobin, MCHC-mean cell haemoglobin concentration, and MCV-mean cell volume) were calculated mathematically from the results above, according to Wintrobe’s formula [23] . MCV was calculated by dividing haematocrit per litre of blood by total RBC count. The differential blood formula (St—stab neutrophils; Sg—Segmented nuclei neutrophils; Ba— Basophils; Еo—Eosinophils; Mo—Monocytes; Ly—Lymphocytes) was determined on the basis of 100 leukocytes per slide, using Shiling’s microscopic method [20] . After the analysis, the animals were released back to the same locations in the biotopes.
2.4. Statistical Procedures
The mathematical data processing was performed using standard statistical procedures with the help of the software package STATISTIKA 7.0 [24] . The normality in the distribution of the examined haematological parameters was checked using a Shapiro-Wilk test that indicated normal distribution: p > 0.05. The haematological parameters have been analysed (in total) in the individuals of both morphs of P. ridibundus using principal component analysis-PCA and standard discriminant analysis—DA. The statistical reliability of the differences in the values of the haematological parameters received for the individuals from both morphs in the compared biotopes was proven with a one-way ANOVA. As a post-hoc test was used an LSD-test. Results with p < 0.05 [α = 5%] were considered significant.
3. Results and Discussion
In preliminary testing of our data (individuals of the homonymous morphs-♂_s/♀_s и ♂_m/♀_m) with one-way ANOVA, we found only two statistically significant differences during spring in the control group: RBC in ♂_s was higher than ♀_s, and the number of Ba cells in ♀_m was higher than in ♂_m. There were no gender differences in the haematological parameters in the populations of the two anthropologically polluted biotopes (p > 0.05). For this reason, comparing the biotopes with various levels and types of pollution, the sample of the relevant population was analysed as a whole (♂ + ♀) for the individuals of the respective morph.
3.1. Multi-Variational Statistics—PCA
The sum of the first three variables includes 75.46% of the variation in the individuals from the two morphs of P. ridibundus: eigenvalue was fixed ≥ 1. The parameters: RBC (−0.941), Hb (−0.859), PCV (−0.891), WBC (−0.805) and Mo (−0.883) show a high degree of correlation in reference to the first axis. The parameters Sg (−0.875), Eo (−0.826), Ba (−0.671), Ly (0.796) have a high correlation in reference to the second axis. The positive dependence on F3 is indicated only by MCV (0.796). The grouping of parameters in reference to the first two main axes is shown in Figure 2.
3.2. Multi-Variational Statistics—DA
DA-discriminant analysis defines the difference between the comparable groups of individuals from the two
Figure 2. Graph of the correlations of the 12 haematological parameters (factor weights) in the individuals from the two morphs of Pelophylax ridibundus and the first two main axes of the biotopes with various degrees of anthropogenic pollution in Southern Bulgaria.
morphs in the biotopes with various degrees of anthropogenic pollution in Southern Bulgaria as statistically reliable, using the parameters: RBC (Wilks’ Lambda = 0.012; F = 3.681; p = 0.003), Hb (Wilks’ Lambda = 0.013; F =9.621; p = 0.000), PCV (Wilks’ Lambda = 0.014; F = 16.916; p = 0.000), Sg (Wilks’ Lambda = 0.015; F = 18.908; p = 0.000), Ba (Wilks’ Lambda = 0.012; F = 3.678; p = 0.003), Eo (Wilks’ Lambda = 0.012; F =5.230; p = 0.000) and Mo (Wilks’ Lambda = 0.011; F = 3.628; p = 0.003). The derived Mahalanobis distances are shown in Table1 The graph of the factor weights of the individuals of both morphs in the six comparable groups, from the biotopes with various degree of anthropogenic pollution, depicts three differentiated clouds (Figure 3). The animals of both morphs from the biotope polluted with heavy metals were clearly distinguished from those in the other two biotopes, while for the animals that inhabit the relatively clean biotope and that of the domestic sewage pollution type, there was an overlapping zone, where most of the individuals are with the striata morph; in our opinion, it is an expression of its higher flexibility in the conditions of specific type of toxins.
3.3. Descriptive Statistics and ANOVA
Due to the insignificant influence of the season factor on the change of the values of the haematological parameters (Figure 2), the data for each of the biotopes was combined (year-round). The ratio between the individuals of the maculata and striata morphs is: Biotope A: 1/0.7; Biotope B: 1/2.8; Biotope C: 1/4.0.
Our findings (descriptive statistics) for the haematological parameters in the groups of P. ridibundus from the three biotopes in Southern Bulgaria, with one-way ANOVA, are shown in Tables 2-4. The differences with a high degree of statistical reliability were found for all examined haematological parameters of the P. ridibundus individuals and from both morphs (striata and maculata) in the groups inhabiting the biotopes with different degree and nature of anthropogenic pollution (Table 5 and Table 6). Statistically significant differences in comparisons of the individuals from the two morphs (striata and maculata):
a) In each type of biotope In the relatively clean biotope (biotope A), the three parameters RBC, St and Sg of the thirteen we studied, have higher values in the individuals of the striata morph, and the parameter Eo—in those of the maculata morph. In the biotope with domestic sewage pollution (biotope B), the parameters RBC, Hb and Ly have higher values in the individuals of the striata morph and it is the same for MCHC, Ba and Eo in those of maculata morph. In the biotope polluted with heavy metals (biotope C), the values of RBC, Hb and PCV are higher in the individuals of the striata morph, while MCH and MCV are higher in those of the maculata morph.
b) In the polluted biotopes (domestic sewage pollution biotope-B and heavy metal pollution biotope-C) compared with the control sample (biotope A)
There was an increase of RBC, Hb and PCV in the individuals of both morphs in the two polluted biotopes. MCH value was lower in both morphs in the biotope B, while in the individuals of the striata morph only it was
Figure 3. Discriminant coordinates for the ten haematological parameters for the individuals of both morphs of Pelophylax ridibundus from the six comparable groups inhabiting biotopes in Southern Bulgaria with various degree of anthropogenic pollution. [Biotopes: A—the river Sazliyka below the village of Rakitnitsa, B—the river Sazliyka below the town of Radnevo; C—the river Topolnitsa below the village of Poibrene and morphs s-striata, m—maculata of individuals].
Table 1. The Mahalanobis distance between the six groups of individuals of both morphs (striata and maculata) of Pelophylax ridibundus from the biotopes in Southern Bulgaria with various degree of anthropogenic pollution using 10 haematological parameters and the corresponding statistically reliable differences between them.
A_s, A_m (Population A—striata and maculata morphs of individuals), B_s, B_m (Population B—striata and maculata morphs of individuals); C_s, C_m (Population C—striata and maculata morphs of individuals). *p < 0.00~1.
lower than those of the control sample biotope with the maculate morph in the biotope C. МCHC decreases in both morphs in the biotope with domestic sewage pollution. MCV value is lower in the individuals of the maculata morph in the biotope B and in these of the striata morph in the biotope C, when compared with the control ones of the maculate morph. In the polluted biotopes there was an increase of WBC in the individuals of both morphs, compared with the control group, regardless of the nature of toxins, however, there were various changes in the character of the differential blood formula: St neutrophils increase proportionally in both morphs; in Biotope B that increase refers only to the individuals of the maculata morph in Biotope A, while in Biotope C to the individuals of both morphs of the control group; the number of Sg neutrophils in the biotope C is greatly reduced in the individuals of both morphs, whereas their number increases in both morphs in the biotope B; the number of Ba cells in biotope B is higher in the individuals of both morphs, but those of the maculata morph are more than these of both morphs in the control group, and these of the striata morph are more only than those of the maculata morph in the control group. In the biotope C, the number of Ba cells decreases in the individuals of
Table 2 . Results from the comparison of the haematological parameters in the groups of Pelophylax ridibundus from the three biotopes in Southern Bulgaria, with one-way ANOVA.
Regression SS = Total SS-Residual SS; Regression DF = regression degrees of freedom = number of independent variables (factors); Regression MS = Regression SS/Regression D; Regression F = Regression MS/Residual MS, Standard Error = (Residual MS) 0.5 * p < 0.00~1.
Table 3. Descriptive statistics of haematologic parameters of circulating blood in the populations of Pelophylax ridibundus from the researched biotopes [RBC: erythrocyte count (N/µl); Hb: haemoglobin concentration (g/dl); PCV: haematocrit value (L/l); MCH: mean cell haemoglobin (pg); MCHC: mean cell haemoglobin concentration (g/l); MCV: mean cell volume (fl)].
Table 4. Descriptive statistics of haematologic parameters of circulating blood in the populations of Pelophylax ridibundus from the researched biotopes [WBC: white blood cell count (N/µl); blood differential formula (N/100 WBC). St: stab neutrophils; Sg: Segmented nuclei neutrophils; Ba: Basophils; Еo: Eosinophils; Mo: Monocytes; Ly: Lymphocytes].
Table 5. Summary presentation of statistically reliable differences in the comparison of the haematological parameters in the groups of Pelophylax ridibundus from the three biotopes in Southern Bulgaria with one-way ANOVA and LSD-test (the signs > and < compare mean values of haematological parameters in Table 3). *p < 0.05, **p < 0.01, ***p < 0.001.
both morphs; in the biotope B, the number of Еo cells increases in the individuals of both morphs, but their number sharply decreases in the individuals of both morphs in the biotope C; the number of Mo cells increases proportionally in the individuals of both morphs in the biotopes B and C; in the polluted biotopes, the number of Ly decreases in the individuals of both morphs.
C) In the two polluted biotopes—B and C
Table 6 . Summary presentation of statistically reliable differences in the comparison of the haematological parameters in the groups of Pelophylax ridibundus from the three biotopes in Southern Bulgaria with one-way ANOVA and LSD-test (the signs > and < compare mean values of haematological parameters in Table 4). *p < 0.05, **p < 0.01, ***p < 0.001, ns − p > 0.05.
RBC, Hb and PCV values are higher in the individuals of the striata morphs in the biotope C. MCH value in the individuals of maculata morph in the biotope C is higher than this in the individuals of both morphs in the biotope B. MCHC value is higher in the individuals of both morphs in the biotope C, compared with B. MCV value is higher in the individuals of both morphs in the biotope B than that of the striata morph in the biotope C. The number of St neutrophils is lower in the individuals of both morphs in the biotope B. The number of Sg neutrophils is lower in the individuals of both morphs in the biotope C. The number of Ba and Eo cells is lower in the individuals of both morphs in the biotope C. The number of Mo and Ly cells is higher in the individuals of both morphs in the biotope C.
The results of this research support the view from our previous work that the increasing values of RBC, Hb and WBC in populations of P. ridibundus (for individuals of both morphs) are sufficiently reliable markers of anthropogenic pollution. The same is true for the changes in PCV values, and partially in MCV. Also the changes in the differential blood formula are sufficiently informative, but the enhanced specificity of their alterations in individuals of the two different morphs to a specific type of toxin must be taken into account.
4. Conclusions
Based on the data obtained in this research, we can make the following conclusions:
• In populations of P. ridibundus (for individuals of both morphs striata and maculata), living in conditions of long-term anthropogenic pollution, there is an increase in the values of the haematological parameters: RBC, Hb, PCV, WBC, St and Mo, as well as reducing Ly and diverse changes in individuals of both morphs in the values of MCH, MCHC, MCV, Sg, Ba and Eo. In the anthropogenically polluted biotopes, they do not undergo seasonal changes and they have constant high values compared to the control group.
• The haematological parameters: RBC, Hb, PCV, Sg, Ba, Mo, to a lesser degree MCV, WBC, Eo, and Ly are most informative, regarding the establishment of changes (of adaptive nature or damages as a result of toxicosis) in the individuals of both morphs (striata and maculata) in P. ridibundus living in conditions of anthropogenic pollution.
• The higher values of RBC (at a high exit level of the parameter in the control group), as well as of the other two basic haematological parameters (Hb and PCV) in individuals of striata morph, in comparison with those of maculata morph, in the two polluted biotopes, are probably one of the reasons for the increased resistance and better survival of the individuals of this morph in anthropogenically polluted biotopes. The adaptive changes in the basic haematological parameters in individuals of striata morph are more distinctive in conditions of domestic sewage pollution.
• The haematological parameters RBC, Hb, PCV, WBC and differential blood formulae in adult P. ridibundus, of both sex and both morphs (striata and maculata), are suitable as markers for assessing water basins’ conditions with a different level and nature of anthropogenic pollution; they can also complement the data from physicochemical analysis and be applied for performing an independent initial general assessment of ecosystems.
Acknowledgements
The P. ridibundus is listed in Appendix 4 to the Bulgarian Biodiversity Act (Prom. SG. 77, August 9th 2002). According to Article 42, Article 41 and Appendix 2 for Article 41 of the same law, capture permits for P. ridibundus are not issued if in use for scientific research.
NOTES
*Corresponding author.