Reaction to pesticides 

In the course of the set experiments according to the capability for growth and the viability under the conditions of the action of different preparations the following relationships were revealed: Lemna minor shows a sudden and strong reaction, which is continued for a while, after which it ceases to react to the toxin; therefore it can be considered a Type 1 Bioindicator (a bioindicator that shows a sudden and strong reaction) according to the classification; Lemna trisulca reacts to contaminants more intensively, however, the distinctive maximum reaction gradually attenuates and therefore it can be categorized as a Type 4 Bioindicator (under the effect of the contaminant the reaction of the bioinidicator gradually becomes of ever of more intensive; however, after reaching/achieving a maximum it gradually attenuates); with Spirodela polyrhiza after an immediate, strong reaction, the response dampens, first sharply, then gradually, which is characteristic of a Type 3 Bioindicator (after immediate, strong reaction of the bioindicator, it is observed to dampen, first sharply, then gradually). 

Thus, on the basis of the carried out observations the forms being investigated can be identified as bioindicators of different types. 

Use of such qualitative indices as specific pigmentation and specific reaction gives the possibility to estimate the specific individual reaction of bioindicators to each separate herbicide.  For the index specific pigmentation we have isolated the following criteria: 

• the yellowing of fronds (change of green pigmentation to yellow) is encountered as reaction practically to all herbicides; 
• browning fronds (change of green pigmentation to brown, Fig. 1 -7) is observed in all species tested as a reaction to the herbicide Roundup; the retention/maintaining the constant pigmentation of the test object with all concentrations of the herbicide burifena. 

Fig. 1.Lemna minor Nurel - 0,04 mg/L (reaction to 2 days).  Fronds are assembled into the  groups, light green. 40% die off.	Fig. 2. Lemna minor Detsis - 0,01 mg/L (reaction to 2 days).  Fronds are light green. 30% die off.
Fig. 3. Lemna minor By sumi - L.- 0,03 mg/L  (reaction to 2 days).  Fronds are light  green.  40-50% die off.  The loss of painting/color goes from pereferii to the center.	Fig. 4.  Lemna minor Anthios - 0,005 mg/L (reaction to 2 days).  The penetration of the preparation occurs on apoplasts, caused lightening entire fronds, 50% of fronds green, 50% lost pigmentation.
Fig. 5. Lemna minor Derozal - 0,12 mg/L (reaction  to 2 days).  Fronds green, are assembled into groups.  Bleaching of the tissue occurs along the vascular bundles, and fronds become "perforated".	Fig. 6. Spirodela polyrhiza Derozal - 0,12 mg/L (reaction to 2 days).  Fronds are light green;  30% die off.  Bleaching of the tissue occurs along the vascular bundles, and fronds become "perforated".
Fig. 7. Lemna minor Omayt - 0,1 mg/L (reaction to 2  days).  Fronds sluggish,  40% die off.

In addition to this the following types of specific reaction were isolated: 

• reticulated pigmentation (it was removed/taken upon consideration of the index of the percentage of damage by the method of vital staining) is characteristic for all concentrations of gradil and kharnes in all indicator-species.

A	B	C
Fig. 8.  Application of a method of vital staining for studying the degree of the phytotoxicity of herbicides based on the example of lesser duckweed (Lemna minor).  Note:  red color designated dead (damaged) cells A - "reticulated" staining, connected with the penetration of dye to the apoplast to networks;  B - "reticulated" staining, which is combined with damage of fronds on the edges, and also with the partial damage to young fronds;  C - the combination of "reticulated" staining with the local penetration of stain.

• Death from the edges of fronds, and the fading of fronds - to the preparations acryl - M bazagran; 
• Disconnection of fronds from the groups - this reaction is observed in the lesser duckweed with concentrations 2,5 and 5,0 herbicides of burifen.

In the last stage of work for us it was necessary to analyze the changeability of all available measurable characteristics.  In the work of Wang (1991) where the basic factor was the growth process, the methods of the statistical analysis were used.  In our work we used the methods of multivariate statistics, where the influence of each factor can be investigated. 
 

Table 1  Qualitative indices of the evaluation of the specific the individual reactions of bioindicators to the herbicides  Signs  Herbicide 1 2 3 4 5 6 7 Specific pigmentation Yellowing + + + + + + + Browning - - - - - + - Maintenance of green pigmentation - - - - - - + Specific reaction Reticulated pigmentation + + - - - - - Death from the edges, fading - - - + + - - Disconnection of fronds from groups  - - - - - - + Note:1) 1 - kharnes, 2 - gradil, 3 - poast, 4 - acryl- M,                                                 5 - bazagran, 6 - raundap, 7 - burifen;                                      2) + the presence of reaction;- the absence of reaction.

Three-factor analysis of variance for the variability of the coefficients of instantaneous growth made it possible to reveal statistically reliable differences on the factors "form", "herbicide" and "concentration" with the portions in general/common variances of 61,9% and 10%, respectively.  Interaction of factors was not considered, since data were organized into the nonrepetitive dispersion complex.  As can be seen from the results of analysis of variance , according to this criterion it was possible to reveal high species specificity (61,0%), which gives grounds for investigating this index for each form separately.
 

Table 2  Results of the three-factor variance analysis the variation of the coefficient of instantaneous growth Variation SS df mS Fraction/portion % Association 0,5141 69   100 Between the concentrations 0,0639 4 0,016 9 Between the forms 0,2425 1 0,2425 61 Between the herbicides 0,079 6 0,0132 10 Residual 0,1287 58 0,0022 20

Then we obtained the table of average values by the index the "coefficient of instantaneous growth" (Table 3).  This seemed interesting, since by this indication the factors "form", "herbicide" and "concentration" were identified as statistically reliable differences.
 

Table 3  Average/mean values of the coefficients of instantaneous growth for  different types, concentrations and the preparations  Levels of the factors Volume Average value Standard error 95% confidence interval  (lower limit) 95% confidence interval  (upper limit) Concentration  1 14 0,13 0,02 0,11 0,16 2 14 0,12 0,02 0,10 0,15 3 14 0,12 0,02 0,09 0,14 4 14 0,10 0,02 0,08 0,13 5 14 0,05 0,02 0,02 0,07 Species  Lemna minor 35 0,16 0,01 0,15 0,18 Lemna trisulca 35 0,05 0,01 0,03 0,06 Herbicide A 10 0,06 0,02 0,03 0,09 B 10 0,07 0,02 0,04 0,10 C 10 0,14 0,03 0,11 0,17 D 10 0,15 0,04 0,12 0,18 E 10 0,09 0,03 0,06 0,12 F 10 0,13 0,04 0,10 0,16 Note: 1) numbers designated the concentrations of preparations in the ascending order  (1 - the lowest, 5 - the highest);  2) they are designated by the letters of Latin alphabet  A - gradil, B - kharnes, C - poast, D bazagran, E - acryl- M, F - Round-Up, G - burifen.

It is evident from table 3 that the differences are observed only between the lowest and highest by concentrations.  The difference between the forms is present.  Moreover Lemna minor is characterized by the greatest coefficient of instantaneous growth.  The herbicides gradil and kharnes possess the maximum phytotoxicity (inhibiting the increase in fronds), while Roundup and burifena are of average phytotoxicity.  Poast, acryl- M and bazagran  inibit growth less.  The appearance of these three groups can explain, on one hand, the degree of phytotoxicity of the tested preparations and, on the other hand, the basic requirements, presented by the herbicides - degree of specificity during the destruction of weeds in the environment (K. Fedke, 1985).

Two-factor analysis of variance revealed, what in the structure of the changeability of horizontal chloroplasts was represented only correlation with the factor of herbicide.  The differences between the forms, and also the effect of interaction are acknowledged to be statistically uncertain. 

Concentration and Herbicide:  The results of three-factor variance analysis revealed statistically reliable differences for the factors of concentration and herbicide, constituting 15% and 23% of the variance, respectively, and it was revealed that the effect of their interaction constituted 34% of the variance.

Species:  Statistically reliable differences were not identified between the species.  Effects of the following  interactions also were not detected:  concentration - form, herbicide - form, and concentration - herbicide - form.

The percentage of the damage related to statistically-reliable differences was not revealed according to the indices, which can be connected to similar insignificant "damaging action" of all herbicides there and with the fact that herbicide concentrations apparently were distant from their PDK for the species data. 

Conclusions:

1. Undoubtedly, the species Lemna minor possesses this advantages as a test subject:  It is very sensitive to pollution/contamination.  Thus, Lemna minor can be categorized as a Type 1 Bioindicator.  Lemna trisulca conforms to the definition of Bioindicator Type  4, one which accumulates contaminants, and only then reacts to them.  Spirodela can be categorized as a Type 3 Bioindicator, which reacts well to a change in the concentration of contaminant, but which will die at high concentrations.
2. Visual analysis of the reactions to the herbicides of the representatives of the family Lemnaceae identified as the criterion the "coefficient of instantaneous growth".  The statistical influence of interspecies differences comprises a 61% portion of the overall variability.