Baranov S., Zykov I., Fyodorova L. INFLUENCE OF CLIMATIC FACTOR ON PLASTICITY AND ASYMMETRY OF LAMINAS OF GENUS TILIA L. SPECIES // Principy èkologii. 2019. № 1. P. 4‒18. DOI: 10.15393/j1.art.2019.8243

Issue № 1

Original research

pdf-version

INFLUENCE OF CLIMATIC FACTOR ON PLASTICITY AND ASYMMETRY OF LAMINAS OF GENUS TILIA L. SPECIES

Introduction

The index of fluctuating asymmetry (FA) is a popular indicator, which is characterized by a slight deviation from zero of the difference between the values of the right and left parts of the homologous bilaterally symmetric trait. The dimensional or countable bilaterally symmetric traits with a wide range of response to stress factors are used to determine the value of FA and developmental stability / instability (Palmer, Strobeck, 2003; Baranov et al., 2015). They include the most genotypically or phenotypically variable traits. In representatives of the genus Tília the dimensional, bilaterally symmetrical traits of leaf plates have high variability, which is associated with species-specific features. It is known that developmental stability (DS) is the ability of an organism to compensate and reduce deviation from normal ontogenetic development along a certain channeling pathway (Debat, David, 2001; Lens et al., 2002; Klingenberg, 2016).

Fluctuating asymmetry is of interest both in ecological terms and in the aspect of evolution and taxonomy of plant organisms. Representatives of the Birch and Linden families are widely used to determine the level of stability of development (Zorina, Korosov, 2015; Zykov et al., 2015), however, the causes of fluctuating changes in the morphological structures of bilateral organs remain poorly understood. The use of tree species as an indicator of development stability in points faces difficulties related to the genotypic purity of the used species, population heterogeneity, the influence of factors, including lighting, water-salt composition and soil regime, terrain, other physical-chemical and other features of the environment.

Methodologically, the determining the level of developmental stability raises some questions. Methods of measurement and preparation of herbarium material occupy an important part and imply a "blind" unbiased approach (Kozlov, Niemelä, 2003; Kozlov, Zvereva, 2015). Traditional methods, such as normalizing difference, are supposed to work with the use of individual traits with a normal distribution of the difference between the right and left values. This approach still finds application, although the evaluation using one or two or many traits does not often provide a representative recording of the signal in the form of an increased value of FA. The rapidly developing method of geometric morphometry allows giving a more harmonious integrative assessment, bringing researchers closer to the true causes for the increase or decrease in FA. The development of epigenetics has led to a number of new studies. For example, works carried out on several generations indicate genotypic conditionality of the processes of decrease-increase in FA (Clarke, 1993; Freeman et al., 1993).

It remains to be recognized that "the health of the environment» is a term which hardly pave the way in bioindication. This opinion is derived from the results of numerous studies in the field of dynamics of the value of FA, modeling experiments in vitro, comparative studies on FA in different species, populations, works with negative results (Zverev et al., 2018), finally, works with justification of paradoxical effects of a dose of stress on FA and developmental stability (Erofeeva, 2015). Thus, studies of FA and DS depending on the gradient of pollution by technogenic toxins show contradictory results: both reverse (Kaligarič et al., 2008, Gostin, 2015) and direct relationship (Belyaeva, 2013; Sobchak et al., 2013; Koroteeva et al., 2015; Ivanov et al., 2015), as well as uncertain or conditionally positive results (Leung, 1996; Lung, 2000). At the same time, the study of FA of leaf plates in intact ecosystems showed a serious impact of terrain height and other abiotic (Zvereva et al., 1997; Wuytack et al., 2011; Baranov, 2014a; Baranov, 2014b) and biotic factors (Zvereva, 1997; Kozlov et al., 2017).

The influence of climatic factors was studied in detail on the example of oak (Hódar, 2002), pine (Kozlov, Niemelä, 2003), white birch (Valkama, Kozlov, 2001) and some herbaceous plants (Alados et al., 2001). The combined effect of many factors, such as humidity and low temperature remains unclear, so these factors against the background of anthropogenic stress are of interest. The logical continuation of the research is background monitoring, studying the stages of morphogenesis of leaf plates (Anastasia, 2014) and work in the field of evolutionary development along a certain, channeling path. The latter term is associated with the property of plasticity of development (plasticity of development; Graham, 2010). In Russian literature, the familiar terms continue to be used – phenotypic or ecological plasticity. It is known that the ecological plasticity of plants is caused by the buffer capacity of morphological structures, which allows them to actively adapt to environmental conditions. The relationship between plasticity and developmental stability continues to be debated. At the same time, the question of fluctuating asymmetry, as an adaptive signal manifested in the trajectory of population development, is debated (Clarke, 1993; Debat, David, 2001; Graham et al., 2010). The modular concept of ecological plasticity is detailed in many sources, for example, De Kroon (De Kroon et al., 2005). According to this concept, the norm of reaction, as the main ecological characteristic of the organism, covers the sum of the plasticity modules of the root, leaf and stem. Evolution and adaptive properties of plastic variability (PV) under the conditions of environmental perturbation are also described in many review papers (Lande, 2009; Gilbert et al., 2015; Abley et al., 2016), including paleontological studies of the size and shape of leaf plates on the example of maple subspecies in the temperature gradient (Royer et al., 2009). A well-established opinion is the correlation between developmental stability, phenotypic plasticity and adaptability at the genetic level (Tonsor et al., 2013). Therefore, plastic variability is of interest as a companion of fluctuation variability both under anthropogenic stress and under severe climatic conditions. Despite the high variation in the characteristics of linden leaves and the frequent presence of directional asymmetry that prevents the determination of FA compared to species of genus Betula, representatives of the genus Linden are more convenient due to a more stable chromosomal karyotype. It is known that close species differ in FA value in response to carbon dioxide content in the environment and emissions of industrial enterprises that confirms the genotypic basis of development stability (Hochwender, Fritz, 1999; Zorina, Korosov, 2015; Koroteeva et al., 2015).

The aim of this work was to determine the stability of development and phenotypic plastic variability in two close species in two urbanized stations with different degrees of anthropogenic load. The task was to study the influence of climate-temperature factors on the stability of development and plastic variability of leaf plates of close species of the genus Linden. The hypothesis being tested was as follows: fluctuation variability is associated with plastic variability, but in different species this relationship is manifested differently, which is explained by the genotypic features of the species.

Materials

Collecting places and preparation of herbarium material

It is known that the big-leaf linden (Tilia platyphyllos Scop.) is a more eurybiontic species in relation to soil and climatic conditions. Both species are equally demanding to light, which is convenient for studying genotypic and phenotypic properties in the context of asymmetry and plastic variability of leaf plates. Populations of both species grow in the Eastern part of Moscow, in Izmailovsky Park and on Izmailovsky Boulevard. Izmailovo natural and historical Park (55°47'N; 37°47'E) is a biocenosis with more than 500 species of plants. Based on this, individuals of trees from different parts of the park, i.e. from different phytocenoses, were selected.

Izmailovsky Boulevard (55°48 '13.8"N; 38°58'23.8" E) is the center of a busy highway, the area of the station on the Boulevard was about 5 km². The characteristic features of this area were the uniformity of lighting, aeration, lack of competition and a high level of atmosphere pollution by carbon oxides, nitrogen, sulfur and hydrocarbons.

The leaf plates were collected at the generative stage of trees development in 2016-2018. In each station, leaf plates 3-4 cm wide were uniformly collected from the lower parts of the crowns of ten trees of the same age, 10 samples from each of the 10 trees. For measurement we used the traits originally developed by the laboratory of V.M.Zakharov (Zakharov, Chubinishvili, 2001). As previous studies have shown (Baranov et al., 2015; Zykov et al., 2015), the angular trait was clearly inconvenient to measure due to the high degree of curvature of the first lateral vein. It was replaced by an alternative to it – linear: the distance between the base of the second vein of the 1st order and the base of the first vein of the 2nd order on the first vein of the 1st order (Fig. 1).

 

Fig. 1.  The traits used to determine the fluctuating asymmetry of species of the genera Tilia L.: 1) width in the middle of a leaf; 2) distance between the bases of the first vein of the 1st order and the second vein of the 2nd order; 3) distance between the bases of the second and third veins of the 2nd order; 4) distance between the bases of the first and second veins of the 1st order; 5) distance between the base of the second vein of the 1st order and the base of the first vein of the 2nd order on the first vein of the 1st order.

 

The leaves were dried under a paper press, and after 2-3 days were measured with a measuring ruler with an accuracy of 0.5 mm (GOST 427-75), the data were stored in Excel tables. The triple measurement showed an extremely low error, less than 1 % of the FA value, which is considered acceptable when testing the asymmetry (Palmer, Strobeck, 2003). Taking into account the data of Velicković (Velicković, 2010) on the greatest sensitivity of the trait of leaf length and previous results, in which another trait was the most suitable (No. 4), we decided to use the whole set of traits, which seems to be an advantage of the integrative approach.

Methods

Preparatory screening of L and R values (values of the left and right traits) plays a significant role in testing FA of leaf plates, no less important than the preparation of herbarium material and improving the accuracy of measurement. First, the data were inspected for the presence of clearly asymmetric features; for example, if L and R differed more than 8-10 times, then such (unit) data were rejected. Then the normality of the distribution (L – R) was checked by the Kolmogorov – Smirnov test with the Lilliefors correction, since this difference is the starting point of the FA testing. The presence of antisymmetry was checked by the kurtosis of values (L-R), since the value of the kurtosis γ < -2 indicate a deviation from the normal distribution and the presence of antisymmetry as a possible form of bilateral asymmetry, which affects the value of FA. The directional asymmetry (DA) was determined by a t-test with the null hypothesis H₀: (L – R) = 0.

Taking into account the possible relationship between the trait value (L + R)/2 and the fluctuating asymmetry, correlation analysis of pair values by Spearman was carried out (Palmer, Strobeck, 2003). Testing of statistical significance of the factors influencing fluctuation and plastic variability was carried out by factorial dispersion analysis. Plastic variability was determined using the formula PL = 1-x/X, where PL was the value of plastic variability, and x and X corresponded to the minimum and maximum values of the trait (L + R)/2 of the leaf plate from each tree (Bruschi, 2003). The average values of PL for each tree were used, i.e. the individual played the role of an experimental unit (n = 10). Most of the statistical analyses were carried out in STATISTICA10 (StatSoft Ink) using the level of statistical significance α = 0.05 %.

Results

Weather conditions

 The 2017 vegetative season was colder and wetter. The average temperature in May-August 2017 was 16-18 % lower, and humidity 2-3 % higher compared to 2016 and 2018. According to the Russian research Institute of hydrometeorological information, May 2017 was the coldest since the beginning of the XX century, and June – the coolest in the last 70 years (table. 1).

Table 1. Climatic data, 2016–2018 

	2016	2017	2018	2016	2017	2018
	H			T
May	60.30	60.21	57.96	15.03	10.91	16.19
June	62.36	69.20	58.87	18.19	14.45	16.32
July	69.54	73.51	72.63	20.96	17.94	20.47
August	72.92	69.82	63.51	19.48	18.84	19.64
Average	66.28	68.18	63.24	18.41	15.53	18.15

 Note. T – average air temperature Celsius at an altitude of 2 m above ground level; H – relative humidity in% at an altitude of 2 m above ground level.

The average air temperature in May was below the climatic norm by more than three degrees, temperature drops below zero and the appearance of snow in May and June were noted, which affects the phases of meristematic activity, with following budding, leafing and the beginning of Linden flowering in Russia (Kishchenko, 2015; Erofeeva, 2011; Ufimtseva, Terekhina, 2017).

  

Verification of preliminary data

The samples (L – R) deviated from the normal distribution, so to determine the fluctuating asymmetry the formula |Log L – Log R|was used. After normalization, only two samples did not satisfy the normal distribution conditions and showed directional asymmetry (t-test; p = 0.01). Considering the slight deviation from the criterion t, it was decided to include these data in the study. In 90% of samples (L – R) the kurtosis value γ was above zero, in the rest - the kurtosis was less than zero, but not less than -0.2. The critical value γ indicating antisymmetry was -0.68, when the number of observations n = 100, therefore, no antisymmetry was detected in the samples (L – R). Paired correlation coefficients, indicating the strength of the relationship between the magnitude of the trait and FA, showed a weak correlation dependency in Tilia cordata (r = -0.11 ÷ 0.1; p < 0.05). The weak correlation between the FA values of the traits allowed us to conclude that the traits were independent and suitable for determining development stability. The influence of various factors on the value of FA is shown in table. 2.

 

Table 2. Effect of species, year, gathering place and combined effect of factors on FA value 

Sources of variation	SS	df		MS	F	p
Species	0.161	1		0.161	149.77	0.000
Year of collection	0.016	2		0.008	7.62	0.001
Species x year of collection	0.012	2		0.006	5.79	0.003
Collecting place	0.003	1		0.003	2.74	0.098
Species x collecting place	0.005	1		0.005	4.31	0.038
Error	1.168		943	0.001

 Note. Conventional signs: SS – sum of squares; df – degree of freedom; MS – mean square; F – F-test; p – level of statistical significance

 

In fact, the value of FA depended on the species, the year of collection of leaf plates, but did not depend on the location of populations.

Fluctuation variability

As shown in Fig. 2, on the Boulevard, the FA value of only one species – cut-leaf linden– was significantly higher compared to the control.

 

Fig. 2. Dependence of FA on collecting place, F (1;945) = 4.31; p = 0.04. Vertical segments represent a 95 % confidence interval.

 

Thus, the cut-leaf linden had fluctuation variability explained by the action of 2 factors: the year of collection and collecting place. Observations of 2017 revealed critical FA values only for cut-leaf linden (Fig. 3).

  

Fig. 3. Dependence of FA on the year of collection. In the field of the diagram – average annual tempera­ture and humidity, average value for May – August. F (2;943) = 5.79; p = 0.003. Vertical segments represent a 95 % confidence interval

 

 

Plastic variability

Significant factors affecting the plastic variability of leaf plates were species and year of collection of leaf plates (table. 3).

    Table 3. Effect of some factors on phenotypic plasticity

Sources of variation	SS	df	MS	F	p
Species	0.037	1	0.037	12.351	0.001
Year of collection	0.053	2	0.027	8.780	0.000
Collecting place	0.001	1	0.001	0.403	0.527
Species x year of collection	0.039	2	0.019	6.412	0.003
Species x collecting place	0.016	1	0.016	4.501	0.037
Error	0.58	178	0.007

    Note. Conventional signs: SS – sum of squares; df – degree of freedom; MS – mean square; F – F-test; p – level of statistical significance.

 

Plastic variability was not the same in the two species (df = 62; p = 0.0002) and depended on year of collection (df = 62;    p = 0.006). In the big-leaf linden indicators PV, as well as FA, were higher than in small-leaved linden (p < 0.05; Fig. 4).

 

Fig. 4.  Plastic variability, F (2; 88) = 6.41; p = 0.003. Vertical segments mean a 95 % confidence interval

 

2017 led to an increase in FA only in big-leaf linden, this year the leaf plate was significantly smaller: 44.34 ± 0.83 – in 2017; 48.74 ± 1.04 - in 2016 and 46.85 ± 0.62-in 2018 (p < 0.001). There was no statistically significant correlation between PV and the mean value of the trait. Characteristics of the leaf plate of small-leaved linden remained invariant during the entire observation period.

 

Verification of results

 Since the leaf plates were unequal in size (about 30 % of samples), the verification task was to determine whether the standard error ME in determining the FA significantly changed depending on the value of the trait. If the FA–trait relationship was stronger than the error–trait relationship, this gave the right to positively verify the statistical significance of the FA value (Palmer, Strobeck, 2003). A regression analysis was carried out to estimate the Pearson correlation coefficient, since the samples containing the average values per tree had normal distribution. The ratio of the difference between correlation coefficients to the measurement error of FA was tested using a t-test based on the assumption of the normality of the distribution of such a ratio:

t_s = (r_FA – r_ME)/ SE_FA,

where t_s is statistical test of differences of correlation coefficients; r_FA– value of r correlation between FA and (R+ L)/2; r_ME – r correlation between FA and ME; SE_FA – standard error of the FA.

 For both species, the statistical value of the t_s criterion exceeded the tabular criterion value t_s = 2.02; n = 40 (small-leaved linden – 62.4, big-leaf linden– 43.4). Thus, the effect of measurement error on the value of FA at the level of individual variability was insignificant. The generalized linear model (GLM) showed no difference in the obtained dependence in the samples grouped by year of collection or by place of collection (p > 0.05), i.e. in different years and in different populations, the measurement error did not affect the FA equally.The authors consider the described verification of the statistical significance of the standard error to be fundamental in the case of working with samples of plates of different sizes, which is often found in such studies.

Discussion

The choice of individual traits to determine the stability of development in contrast to the integrative path is questionable and can be justified only if there is an accurate assessment of the gradient of factors on the value of FA of the trait (Baranov, 2014b; Baranov et al., 2015, Zykov et al., 2015). In our case, normalization with the use of logarithmation allowed to operate with normally distributed values in all 5 traits, which can be considered, rather, as luck. At the same time, an intrigue to use separate traits (one or two) the most sensitive to pollutants remains (Baranov, 2014a; Koroteeva et al., 2015). It is known that the plasticity of plants is caused by the buffer capacity of morphological structures, which allows them to adapt actively to environmental conditions (Debat, David, 2001; Klingenberg, 2003; Tonsor et al., 2013; Venâncio et al., 2016). Consequently, stressful climatic effects and chemical pollution on the Boulevard affected the PV, adapting the plant. It is known that on the Kola Peninsula in the cold summer there was an increase in the FA of the leaves of the white birch, while the temperature was a more significant factor than pollution with sulfur dioxide and nickel (Valkama, Kozlov, 2001). As in the case of birch, the climatic factor played a more important role than the action of chemical pollutants.

Big-leaf linden is a more thermophilic species with more pronounced stenobionticy to climatic factors and physical and to chemical properties of the soil. We see the reason for the increase in FA of leaf plates in the decrease in the growth of plates and the decrease the photosynthesis activity during the cold summer. This is confirmed by the indicator values of thermophilic big-leaf linden, which are shifted to the right positive side (8-13) compared to small-leaf linden (6-12 on the Tsyganov scale).

The authors attribute the adaptation of big-leaf linden to the conditions of the Central Russian plain to one of the phases of the reaction to stress with an increase in the FA of leaf plates. Consequently, the hypothesis of the relationship of 2 types of variability was confirmed. We consider the specific genotypic feature of this introducer in the form of increased plastic and fluctuation variability as a "fee" for adaptation in the condition of climatic stress factors.

Conclusions

The phenotypic plasticity influences fluctuation variability (Houle, 2000; Sultan, 2003; Scheine, 2004; Lajus, Alekseev, 2003; Tucić et al., 2018). Stability of development is a category that explains the processes caused by the peculiarities of the genotype and the environment. PV is an optional indicator correlated with FA, but a frequent manifestation accompanying FA. In this case, the increased plastic variability was a phenotypic deviation caused by low air temperatures, while the increased fluctuating asymmetry was due to the small size of leaf plates. The growing plate is known to develop intermittently in the right, then in the left direction and has increased asymmetry (Freeman et al., 1993). It is known that the oscillation of the variance of the difference (L-R) within the same value of the trait should be the attribute of destabilization of development (Palmer, Strobeck, 2003). In our case, the plasticity of the development of leaf plates was increased, however, FA, although increased, was not evidence of a decrease in the stability of development as a deviation in the homeostasis of development at the biochemical level (Klingenberg, 2003; Graham et al., 2010; Klingenberg, 2016).

Increased plastic variability in the leaf plates of the big-leaved linden as a species-specific property, without imbalance in the homeostasis of development, should be taken into account in studies related to long-term monitoring of the stability of the development of woody plants.

References