Algorithm for Experimental Estimation of the Conditional Threshold
for the Duration of Low Temperatures Exposure on the Example of the
Laboratory-Reared Population of Lymanrtia Dispar
SEDELNIKOV A.V.
Department of Higher Mathematics
Samara National Research University
Samara, Moscow highway, 34, 443086
RUSSIA
Abstract: The paper substantiates the introduction of a new parameter into the development model of the
laboratory-reared population of Lymantria dispar, formulates and mathematically formalizes the parameter, and
develops an algorithm for its experimental evaluation. It increases the correctness and adequacy of the
mathematical description of the population development in terms of assessing the main parameters of its
development used earlier in the model. The obtained results can be used to study the development of
laboratory-reared populations of Lymantria dispar, as well as to understand the dynamics of population
development in the natural environment.
Key-Words: Lymantria dispar, laboratory-reared population, development model, algorithm for experimental
estimation.
Received: April 21, 2022. Revised: October 13, 2022. Accepted: November 22, 2022. Published: December 15, 2022.
1 Introduction
Modeling the development of laboratory-reared
populations requires a rigorous mathematical
formalization of the development process itself. At
the same time, it is necessary to formulate model
parameters that allow taking into account the impact
of certain factors on the process of population
development in order to build a correct and adequate
model. Modern models do not describe this complex
multifactorial process in sufficient detail even under
conditions of long-term observation of the
population. Therefore, their clarification and
addition is an important and urgent task in studying
the development of populations. This clarification
can take various forms.
For example, there is no general opinion
regarding the temperature threshold from which the
sum of effective temperatures should be calculated
when hatching caterpillars from eggs, when
studying the development of Lymantria dispar
caterpillars. The article in [1] presents different
approaches to this issue. At the same time, the
approaches are fundamentally different from each
other. The concept of some fixed threshold value
can be found in most works like [2], [3], [4] and
others. However, there are more complex ideas
about the threshold values, for example, that they
vary depending on the period of eggs stay at low
temperatures, [5], or some other factors, [6], [7]. In
this case, it is necessary to supplement the
composition of the model parameters and
experimentally study the temperature threshold, [8],
[9]. At the same time, the introduction of a critical
value of the parameter will make it possible to
combine these concepts, considering them within
the framework of one model. The temperature
threshold can be considered constant before
reaching the critical value. It can be considered
variable after reaching the critical value.
The workd in [10], [11], [12] and [13] consider
the development of the laboratory-reared population
of Lymantria dispar in the framework of the Eigen
quasispecies model. [14], [15]. Here, the main
parameters are singled out for studying their
dynamics when fixing external conditions that affect
the development of the population. At the same
time, the main parameters should tend in probability
to some optimal values corresponding to fixed
external conditions according to the Eigen model. In
this case, the introduction and experimental study of
the dynamics of a new model parameter will
increase confidence or, conversely, doubts about the
correctness of using the Eigen quasispecies model in
describing the development of a laboratory-reared
population.
Thus, the presented work is aimed at
supplementing the development model of the
laboratory-reared population of Lymantria dispar
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
215
Volume 19, 2022
with a new parameter and developing an algorithm
for its experimental evaluation. It will increase the
adequacy and correctness of the development
model.
2 Conditional threshold for the
duration of exposure to low
temperatures as an important factor
in the population development
It is known that under natural conditions the
formation and development of Lymantria dispar
caterpillars in the egg occurs in autumn, [7]. They,
as a rule, go into diapause after full formation. In the
literature, cases of caterpillars' non-diapause
development are known both in nature, [16], and
under laboratory conditions, [1]. The introduced
parameter will not make sense for a non-diapause
population since during the development of such a
population there is no effect of low temperatures on
eggs. However, non-diapause individuals, as a rule,
do not complete their development and die under
natural conditions, [1]. Laboratory-reared
populations in some cases are quite viable, [16]. In
this regard, it should be noted that we are talking
about diapause populations of Lymantria dispar.
It has been established that for the formation of
caterpillars in eggs and the normal course of
diapause, a sufficient amount of heat in autumn and
moderate cold at the beginning of winter are
necessary, [7]. It is possible to fix one of the
parameters in laboratories. Thus, the sum of
effective positive summer-autumn temperatures was
recorded for all generations at the level of +580 ± 10
0С for the studied laboratory-reared population. It
was experimentally established that with such a
sum, the onset of embryonic and temperature
diapauses occurs almost simultaneously. The
embryonic diapause precedes the temperature
diapause in nature. Otherwise, the caterpillars in the
eggs are underdeveloped and often die. However,
the significantly more complex temperature
distribution in nature compared to laboratory
conditions does not make it possible to obtain a
relatively simple mathematical model.
Further, the egg-laying of the laboratory-reared
population is placed in a refrigerator at a
temperature of –2 +6 0С. An important issue for
constructing a development model is the duration of
the period of exposure to low temperatures, during
which normal development of caterpillars will be
observed. It is obvious that this duration should
have upper and lower bounds. Moreover, the
concept of a lower bound is not so obvious and can
be interpreted in different ways. Indeed, in fact, we
are talking about the minimum value of the period
of exposure to low temperatures, at which normal
development of caterpillars is possible. However,
non-diapause populations of Lymantria dispar
indicate a zero boundary, especially since we are
talking not only about laboratory-reared
populations, but also about natural phenomena.
Therefore, there is no unconditional duration
threshold. On the other hand, the existence of non-
diapause populations has not been described in
nature, and the habitat of Lymantria dispar in nature
is limited in the west by 30–60 parallels north
latitude, in the east by 50 parallel north latitude and
by the northern tropic, [7]. Therefore, we can
assume that the normal development of Lymantria
dispar in nature with a zero unconditional threshold
is not observed. At the same time, the unconditional
duration threshold is a low-value parameter for
formulating a population development model.
Let us introduce the concept of a conditional
threshold for the duration of the period of exposure
to low temperatures. Various restrictions can be
selected as a condition. For example, it is proposed
to use the condition of hatching at least a third of
viable eggs. Thus, we can write:
min
0
3need
D
N
NtD
, (1)
where D is the duration of the period of
exposure to low temperatures; t is the time
parameter; N+ is the number of hatched caterpillars;
N0 is the total number of eggs involved in the
experiment; Dneed min is the period duration threshold
for exposure to low temperatures.
It should be understood that Dneed min is a
parameter that depends both on the sum of effective
positive summer-autumn temperatures and on the
temperature values during the period of low
temperatures. Therefore, it is more correct to
characterize it as a conditional threshold for the
duration of the period of exposure to low
temperatures (provided that at least 1/3 of viable
eggs are hatched). This threshold corresponds to a
certain value of the sum of effective positive
summer-autumn temperatures (for example, +580
0С as in the experiments) and the average
temperature during the period of low temperatures
(for example, +4 0С). All fertilized eggs can be
considered viable under laboratory conditions.
It should also be noted that when the conditional
threshold is reached, the hatching rate is an
increasing function. Then the increase is replaced by
a decrease. Starting from some value of D,
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
216
Volume 19, 2022
caterpillars will stop hatching altogether. The
condition for the hatching of at least a third of viable
eggs will be fulfilled in a certain range of D values
since hatchability is a continuous function of D. Of
course, Dneed min will be the minimum value of this
segment.
The dynamics of the conditional threshold is an
important indicator of population development. In
nature, global warming is observed, which leads to a
reduction in the period of low temperatures and its
gradual disappearance. It means either a reduction in
the habitat or the adaptability of the population to
new conditions. The first scenario is the most
probable with the Dneed min value not changing
significantly. If, on the contrary, its value tends to
decrease, then this indicates a gradual adaptability
and development towards a non-diapause
population. Global warming can lead to the
formation of a stable natural non-diapause
population, especially since such laboratory-reared
populations exist and are successfully developing.
3 Algorithm for estimating the
conditional threshold for the duration
of exposure to low temperatures
3.1 Estimating the value of the conditional
threshold for the duration of exposure to low
temperatures
Based on the problem being solved, two approaches
to estimation are possible: according to the required
accuracy and according to the available source
material. The first approach assumes the presence of
a large amount of biological material (eggs) that can
satisfy the required estimation accuracy. We choose
a confidence level β (usually 0.95 or 0.99). Then the
average value of hatched caterpillars will be in the
confidence interval:
t
N
Nt
N
N
00
;
,
with the probability β, where σ is the sample mean-
square deviation;
2
1
0
t
, and
are the
Laplace's function.
In this case, the required sample size (the number
of eggs required for the experiment) is determined
as follows:
2
0
t
N
,
(2)
where ε is the prescribed accuracy.
Next, we should determine experimentally or
assign the value of the time step. Apparently, a
uniform step can be used initially. In fact, this value
has a certain meaning, besides a simple time scale.
It denotes the period of time during which changes
are so serious that they significantly affect the
hatching rate. However, most likely, its value is not
constant and, at least, depends on the duration of
exposure to low temperatures. The following
algorithm can be used when experimentally
determining the time step. It is necessary to select a
certain sufficient number of early egg-laying at the
end of the generation development. We divide them
into separate batches and, using a different time step
value for different batches, estimate the statistically
most probable step size. At the same time, we
assume that this value will not change significantly
for later egg-laying. You can also use the experience
of previous generations, if the conditions for their
development and the main indicators of the
population do not differ notably from each other. In
order to insure yourself against possible outliers of
early egg-laying in the presence of previous
experience, it is possible to estimate the step in two
ways, followed by a comparison of these estimates.
You can build an integral estimate using weighting
factors if it’s necessary.The biological material
should be removed from the refrigerator and the
number of hatched caterpillars should be recorded
using the sample size calculated by formula (2) and
the time step. As a result, a correlation dependence
of the conditionally average number of hatched
caterpillars on the duration of the period of exposure
to low temperatures will be obtained. Next, it is
necessary to identify the trend of this correlation
dependence. The trend will be a linear or non-linear
functional relationship. Substituting condition (1)
into this dependence will give the estimate Dneed min.
However, it should be understood that the accuracy
of this estimate will decrease due to the addition of
an error in the approximation of the correlation
dependence by a functional trend. This fact should
be taken into account especially when using a linear
trend as the simplest functional relationship.
The second approach assumes the presence of a
limited amount of biological material (eggs). From
here it follows, using formula (2), to determine the
maximum accuracy of estimating the conditional
average value of hatched caterpillars:
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
217
Volume 19, 2022
0
N
t
.
(3)
Most likely, we will have to be content with a
uniform time step in this case. Its experimental
estimation may not be available, as in the first
approach. Therefore, we will have to use previous
experience, if available, and simple considerations.
For example, it hardly makes sense to use a step size
of less than one day. In this case, it will be difficult
to calculate the effective temperature, which can
lead to a decrease in the estimate accuracy.
Otherwise, the algorithm for estimating the
conditional threshold for the duration of exposure to
low temperatures will coincide with the first
approach.
3.2 Estimation of the error in the value of the
conditional threshold for the duration of
exposure to low temperatures
An important issue is not only the estimation of the
value of the conditional threshold for the duration of
exposure to low temperatures, but also the error
estimation. There are two types of errors to
consider. The first of them is a statistical error
associated with the limited sample N0. The second
error, as mentioned above, is related to the
approximation error. Then the total error will be the
sum of these two errors. For the number of hatched
caterpillars, the statistical error will be determined
by inequation (3). Table 1 provides some data on the
estimation accuracy and the sample size required to
achieve this accuracy.
Table 1. There are data on accuracy, volume of
biological material (eggs) and confidence interval
ε
β
σ
tβ
N0
Confidence interval
1
0.95
10
1.645
271
1
N
1
0.99
10
2.326
542
1
N
The following considerations can be used to
estimate the approximation error. It is necessary to
calculate the part of the variance of the correlation
dependence that is not explained by the functional
dependence. For this purpose, we need to calculate
the regression residuals
i
:
i
i
iDNN ˆ
,
(4)
where
i
N
are the experimental values of the
hatched caterpillars number for the duration of the
exposure period to low temperatures Di;
i
DN
ˆ
are
the same values approximated by the functional
dependence.
Then the unexplained part of the variance will be
equal to:
n
ii
ESS
1
2
,
(5)
where n is the total number of experimental
points.
The approximation error can be estimated as
follows using (4) and (5):
n
ESS
ˆ
.
(6)
Then for the value of hatched caterpillars we can
write:
)
ˆ
(
N
. We use the functional dependence
DfN
ˆ
to estimate the approximation error in the
value of the conditional threshold for the duration of
exposure to low temperatures. Substituting the values of
)
ˆ
(
N
and
)
ˆ
(
N
for
N
ˆ
, we obtain,
respectively, the values of Dmin and Dmax. Then we have:
.
2
;
2
minmax
minmax
min
DD
DD
Dneed
.
(7)
δ is the required error in estimating the value of
the conditional threshold for the duration of
exposure to low temperatures in system (7). It also
includes the statistical error (limitation of the
sample) and the error in approximating the
experimental correlation dependence
i
iDfN
by the functional dependence
DfN
ˆ
.
Thus, a new indicator is introduced in the work
in the form of a conditional threshold for the
duration of the period of exposure to low
temperatures. This indicator allows you to control
the stability of external conditions when growing a
laboratory population of Lymantria dispar.
Permissible values of the percentage of hatching of
caterpillars from eggs with the help of this indicator
are transformed into the minimum required duration
of the period of exposure to low temperatures. It is
very useful when developing a plan for rearing
generations of the laboratory population. Since this
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
218
Volume 19, 2022
indicator is reasonable and is associated with the
main parameters describing the development of the
laboratory population.
4 Experimental part
4.1 Experimental material
14 egg-laying of the semiannual laboratory-reared
population of Lymantria dispar were selected, [12],
[13], [17] for the experimental study. This is the
seventh generation of the population grown in
artificial laboratory conditions. Data on the studied
parameters for previous generations are shown in
Table 2.
Table 2. There are data on the studied parameters
of six generations of the semiannual laboratory-
reared population of Lymantria dispar
Year
Genera-
tion
designati
on
D, days
N0
Hatching,
%
Thatmin,
0С
hat, 0С
2018
F0
No
data
280
89
90
100
2018
0,5F1
49
62
82
153
230
2019
0,5F2
133
100
95
72
106
2019
0,5F3a
0,5F3b
42
42
22
27
50
89
170
133
306
224
2020
0,5F4a
0,5F4b
153
153
100
100
66
45
89
89
119
119
2020
0,5F5a
0,5F5b
0,5F5c
35
38
86
50
50
50
84
88
94
151
135
73
216
229
95
2021
0,5F6a
0,5F6b
97
102
50
50
94
48
110
110
130
131
Table 2 uses the designations: Thatmin is the
hatching threshold, which denotes the minimum
sum of effective positive temperatures from +6 0С,
after which caterpillars hatch, [17]; hat is the
average sum of effective temperatures, which
denotes the average value of the sum of effective
positive hatching temperatures of all hatched
caterpillars from +6 0С, [17].
At the same time, the sums of effective positive
temperatures from +6 0С at the egg stage fluctuated
in the range of +580...+590 0С. This parameter can
be considered constant taking into account the
estimation errors. The ambient temperature was
+20...+25 0С during this period. A total of 768
fertilized eggs were used in the experiments.
4.2 Experimental results
All experimental material was divided into 10 parts
and was exposed to low temperatures in the range of
–2...+2 0С for different periods of time. Then the
eggs were taken out of the refrigerator for hatching
at a temperature of +20...+25 0С. The hatched
caterpillars developed normally to the adult stage in
all ten batches. The results of the evaluation of the
parameters studied during the experiment are
presented in Table 3.
Table 3. There are data on the studied parameters of the semiannual
laboratory-reared population of Lymantria dispar
Population
labeling
D, days
N0
Hatching,
%
Thatmin,
0С
hat,
0С
0,5F7a
34
51
39
175
289
0,5F7b
44
77
83
173
262
0,5F7c
56
83
37
160
198
0,5F7d
64
137
94
137
169
0,5F7e
74
89
97
130
148
0,5F7f
84
84
95
135
152
0,5F7g
94
36
97
160
171
0,5F7h
105
85
75
124
146
0,5F7i
116
46
78
130
167
0,5F7j
148
90
32
150
171
Figure 1 shows the experimental data (Table 3)
and the approximation of the experimental data by a
third-order polynomial.
Fig. 1. There are experimental data and approximation:
1 is hatching, 2 is approximation of the hatching, 3 is minimum sum of
effective positive temperatures from +6 0С, after which caterpillars, 4 is
average sum of effective positive temperatures from +6 0С, after which
caterpillars
In this case, the estimated value is
daysDneed 29
min
or, taking into account errors
(3) and (6), is
daysDneed 829
min
. One can
obtain an estimate of the upper bound on the
duration of exposure to low temperatures arguing in
a similar way. We calculate
daysDneed 147
max
using the approximation in Figure 1. Taking into
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
219
Volume 19, 2022
account errors (3) and (6), we obtain:
daysDneed 8147
max
.
Thus, it was found that for the duration of
exposure to low temperatures
daysD 139,37
,
the condition
3
0
N
N
will almost certainly be met
in the experiment course.
5 Conclusion
Thus, this paper describes a new parameter for the
development of the laboratory-reared population of
Lymantria dispar, which is a conditional threshold
for the duration of exposure to low temperatures. An
algorithm for estimating this parameter using
experimental data is given. The calculation of the
estimation error is presented. This parameter can be
used to study the applicability of the Eigen quasi-
species model for the correct description of the
development of the laboratory-reared population of
Lymantria dispar.
The presented parameter can be interpreted more
broadly than it is described in this paper. Depending
on the problem being solved, condition (1) can be
replaced by another condition. At the same time, the
presented studies can be considered as an example
of the parameter formation for a specific task posed
in this paper.
References:
[1] Meshkova V.L., Seasonal development of
eating needles and leaves wreckers of the
forest, Novoje Slovo, Kharkiv, Ukraine, 2009.
[2] Furuta K., Developmental zero temperature and
eff ective thermal constant in Lymantria dispar
collected from Sapporo, Hokkaido, Forest
Pests (Sinrin Boeki), Vol. 22, 1973, pp. 120
123.
[3] Johnson P.C., Mason D.P., Radke S.L.,
Tracewski K.T., Gypsy moth Lymantria dispar
(L.) (Lepidoptera: Lymantriidae), egg eclosion:
degree-day accumulation, Environ. Entomol.,
Vol. 12, 1983, pp. 929–932.
[4] Hihashiura Y., A method to forecast the dates
of hatching and dispersal of larvae in
Lymantria dispar, Kosunai kihou, Vol. 74,
1989, pp. 19–24.
[5] Rubtsov I.A., Influence of constant and
variable temperatures on the development of
eggs of a Lymantria dispar (L.), Plant
protection, No. 17, 1938, pp. 25–37.
[6] Ilyinsky A.I., Supervision, registration and
forecast of mass reproduction of pine and leaf-
gnawing insects in the forests of the USSR,
Lesnaya promyshlennost’, Moskow, USSR,
1965.
[7] Gamayunova S.G., Novak L.V., Voitenko Yu.
V., Kharchenko A.E., Mass needles and leaf-
eating forest pests, Publishing house of the
Kharkiv State Agrarian University, Kharkiv,
Ukraine, 1999.
[8] Castedo-Dorado F., Lago-Parra G.,
Lombardero M.J., Liebhold A.M., Álvarez-
Taboada M.F., New Zealand Journal of
Forestry Science, Vol. 46, 2016, 18.
[9] Islam M.S., Barr N.B., Braswell W.E.,
Martinez M., Ledezma L.A., Molongoski J.,
Mastro V., Schuenzel E.L., A Multiplex Real-
Time PCR Assay for Screening Gypsy Moths
(Lepidoptera: Erebidae) in the United States for
Evidence of an Asian Genotype, Journal of
Economic Entomology, Vol. 108, 2015, pp.
2450–2457.
[10] Sedelnikov A.V., Lymantria dispar biocycle
formalization within the model of quasispecies:
theoretical framework, Scientific Life, No. 7,
2016, pp. 98–104.
[11] Sedelnikov A.V., Estimation of the average
number of eggs in laying of Lymantria dispar
in Samara region, International Research
Journal, No. 1, 2016, pp. 109–111.
[12] Sedelnikov A.V., Khnyryova E.S., Estimate of
parameters changing of the Lymantria dispar
artificial population in different environmental
conditions with using the quasispecies model,
International Journal of Modeling, Simulation,
and Scientific Computing, Vol. 11, No. 5, 2020,
2050053.
[13] Sedelnikov A.V., Mathematical formalization
of the process of crossing individuals of
different populations within the framework of
the quasispecies model, International Journal
of Biomathematics, Vol. 14, No. 1, 2021,
2150022.
[14] Eigen M., P. Schuster, The Hypercycte. A
Principle of Natural Self-Organization,
Springer-Verlag, Berlin-Heidelberg-New York,
1979.
[15] Musso F., A stochastic version of Eigen
model, Bulletin of Mathematical Biology, Vol.
73, No. 1, 2011, pp. 151–180.
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
220
Volume 19, 2022
[16] Kozhanchikov I.V., Peculiarities of wintering
and diapause of the Gypsy moth (Ocneria
dispar L.), Reports of the Academy of Sciences
of the USSR, Vol. 73, No. 3, 1950, pp. 605–
607.
[17] Sedelnikov A.V., The effect analysis of the
diapause duration on the hatching parameters
of the artificial semi-annual population of
Lymantria dispar caterpillars, International
Journal of Biomathematics, Vol. 14, No. 8,
2021, 2150064.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
There is no funding provided to prepare the
manuscript.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
your ID is: WF-1245
WSEAS TRANSACTIONS on BIOLOGY and BIOMEDICINE
DOI: 10.37394/23208.2022.19.23
Sedelnikov A.V.
E-ISSN: 2224-2902
221
Volume 19, 2022