=Paper=
{{Paper
|id=Vol-2030/HAICTA_2017_paper74
|storemode=property
|title=Microclimates and their Stochastic Effect on Olive Fruit Fly Evolution: Modeling and Simulation
|pdfUrl=https://ceur-ws.org/Vol-2030/HAICTA_2017_paper74.pdf
|volume=Vol-2030
|authors=Romanos Kalamatianos,Markos Avlonitis
|dblpUrl=https://dblp.org/rec/conf/haicta/KalamatianosA17
}}
==Microclimates and their Stochastic Effect on Olive Fruit Fly Evolution: Modeling and Simulation==
https://ceur-ws.org/Vol-2030/HAICTA_2017_paper74.pdf
Microclimates and their Stochastic Effect on Olive Fruit
Fly Evolution: Modeling and Simulation
Romanos Kalamatianos1, Markos Avlonitis2
Department of Informatics, Ionian University, Greece, e-mail: {1c14kala, 2avlon}@ionio.gr
Abstract. Climate variables play an important role in the development and
general activity (diffusion, oviposition etc.) of the olive fruit fly. These
variables can fluctuate from area to area due to the topography of the area and
several other factors. Due to this fluctuations microclimates are created.
Through simulation runs we investigate how the population dynamics of the
olive fruit fly are affected in four distinct microclimates, from an olive grove
area in Corfu, Greece with environmental data collected from environmental
sensors. Finally, we investigate how current spraying practices affect the
population of the olive fruit fly in each microclimate.
Keywords: olive fruit fly, microclimate, simulation model
1 Introduction
Bactrocera oleae also known as olive fruit fly is a widespread pest of olive trees in
many Mediterranean countries, such as Greece, as well as other olive producing
countries around the world. Olive fruits are the growth habitat of all three premature
stages of the insect. Female olive fruit flies oviposit the eggs inside the fruit and the
larva that hatches feeds from it. At the last premature stage the insect pupates either
inside the fruit or in the ground until the perfect adult emerges (Vossen et al, 2004).
The total number of olive fruit fly generations per year varies from region to region
due to changes in the climate. Up to five generations were reported in Southern
California, with the authors stating that under ideal conditions the olive fruit fly
could reach seven generations (Rice et al, 2003). While, in Jordan up to three
generations have been documented (Mustafa and Alzaghal, 1987).
The development and activity of the olive fruit fly is heavily dependent upon
environmental conditions. Both temperature and relative humidity are important
enviromental factors that affect the population dynamics of the olive fruit fly.
Specifically, the development rate of the immature stages increases as the
environmental temperature increases within the developmental thresholds (Tsitsipis
1977 & 1980). When exposed to high temperatures, even for a couple of hours, it
affects all development stages in terms of longevity, reproduction and survival
(Pappas et al, 2010). Furthermore, high temperatures affect the flight performance of
adult olive fruit flies (Xin-Geng Wang et al, 2009). Likewise, low values of relative
610
humidity, below 33%, reduces ovarian maturation of female olive fruit flies, as well
as the longevity of the adult flies. For higher values percentages of the
aforementioned parameters increase (Broufas et al, 2009).
Both aforementioned environmental parameters are affected by the topography of
a region. For instance temperature is known to decrease with increasing altitude, also
known as the lapse rate, which ranges from 5.5oC, to 6.5oC, per 1000m gain in
altitude (Hodkinson, 2005). Temporal and spatial variations of solar radiation, cloud
coverage, the presence of shading and proximity to the sea, or to mountains, among
others, can also have an impact on environmental conditions. All these factors
constitute in the creation of a microclimate. Thus, an olive grove field depending on
the topography of its location and size can exist in the same microclimate, or it can
be divided into numerous microclimates.
The population size of the olive fruit fly inside an olive grove if left unchecked
can reach high numbers, infesting the whole grove as a result. Therefore, several
methods are used to monitor and control the population, which can be divided into
two main categories (Vossen and Devarenne, 2006).
The first category is mass trapping, where some type of trap is placed on every
tree inside the grove. Types of traps include the Attract and Kill device, which is a
cardboard, coated with pesticide killing olive fruit flies coming in contact with it.
Other types of traps are the McPhail trap, which is a liquid trap that baits olive fruit
flies and the Sticky panel trap which is a yellow cardboard with a non-drying sticky
coating that lures olive fruit flies. Finally the OLIPE trap which is a plastic bottle
with small holes, filled with the same liquid solution as the McPhail trap.
The second category of population control is spraying. One method of spraying is
the Kaolin clay, which creates a fruit barrier preventing olive fruit flies from laying
eggs inside it. Another method of spraying is the Spinosad bait, which is sprayed
onto every other tree and lures olive fruit flies to feed from it and eventually die.
Aim of this paper is to investigate, via simulation runs, how the evolution of the
olive fruit fly is affected by the microclimate it resides in. As well as, the effect
current patterns of spraying have on the population of the olive fruit fly when the
existence of microclimates is disregarded. In an earlier version of the simulation
model (Voulgaris et al, 2013) a comparison of the spatial distribution of the olive
fruit fly of a good spraying scenario versus a bad spraying scenario was presented.
2 Material and Methods
2.1 Simulation Model
In order for the simulation to be executed, information about the field must be given
as input. Therefore, different layers of information are passed, the first layer provides
the location of olive trees and urban areas, as well as the location of olive fruit fly
traps (Voulgaris et al, 2013). The second layer provides the percentage of olive fruit
that is present at the olive grove locations (Kalamatianos & Avlonitis, 2015). Finally,
611
we upgraded the model to identify one or more microclimates that the field is divided
to and attach to each a different temperature set. All areas inside a microclimate have
the same temperature. Given the aforementioned information the simulation model
creates a grid representing the real life field and populates its cells with a starting
population based on olive fruit fly trap measurements.
The population is structured into five transformation stages. Namely, egg, larva,
pupa, sexually immature adult and fully mature adult. All adults are considered
female olive fruit flies and each of them can lay up to three eggs in its lifetime.
The degree-day model (Wilson & Barnett, 1983) is utilized to simulate the
development of the pre-imaginal stages (egg, larva, pupa) of the olive fruit fly. Since,
the time resolution is hourly the following equation is used to compute the
accumulated degree hour units (Kalamatianos & Avlonitis, 2015).
DH(ti) = ((ti – TL) * (1 – (1 / (1 + exp(-10 * (ti – TU)))))) / 24. (1)
Where ti is the temperature present in the i-th simulation step, TL and TU are the
lower and upper developmental thresholds, respectively, of each olive fruit fly which
differs from one transformation stage to the other.
For the spatial dispersion of the olive fruit fly the first major parameter is olive
fruit percentage. The average speed of an olive fruit fly based on the olive fruit
percentage of its current position is calculated using Equation 2 (Kalamatianos &
Avlonitis, 2017).
AvgSpeed(fx,y) = (-4.513* fx,y + 451.8) / wh. (2)
Where fx,y is the olive fruit percentage on (x, y) coordinates of the grid and when
the total daytime hours in the current week. The second parameter is temperature,
which affects the final speed of the olive fruit fly. Equation 3 calculates the speed of
the olive fruit fly based on the temperature present at its current position
Fig. 1. Map of the study area in Corfu, Greece. (Map data: Google, TerraMetrics, Data SIO,
NOAA, U.S. Navy, NGA, GEBCO, CNES/Airbus)
612
Fig. 2. Study area microclimate division. Blue colored is microclimate A, red colored is
microclimate B, green colored is microclimate C and yellow colored is microclimate D. (Map
data: Google, TerraMetrics, Data SIO, NOAA, U.S. Navy, NGA, GEBCO, CNES/Airbus)
(Kalamatianos et al., 2015).
Speed(fx,y,ti) = AvgSpeed(fx,y) * (1 – (1 / (1 + exp(-10 * (ti – TU))))). (3)
Where ti is the temperature present in the i-th simulation step, TU is the upper
movement threshold. The upper movement threshold was set to 35 oC, beyond this
temperature the olive fruit flies stop drifting (Avidov, 1954; Johnson et al., 2011).
2.2 Study Area
The study area was the municipality of St. George (Figure 1), which is located on the
northwestern part of the island of Corfu, Greece. High temperatures and almost no
rainfall in general characterize the climate type of the study area and the island
during the summer months. The total area is 9km x 7,5km.
Although the study area may consist of many microclimates, the data we collected
wasn’t enough to distinguish the different microclimates. Thus, for simplicity
reasons we divided the area arbitrarily into four microclimates as they are depicted in
Figure 2. The resulting microclimates from left to right are designated as
microclimate A, B, C and D.
2.3 Data
Figure 3 depicts the locations where environmental sensors were installed.
Microclimates where more than one sensor was installed, the average value was
calculated from all sensors and was used to represent the temperature for all areas
613
Fig. 3. Environmental sensors. (Map data: Google, TerraMetrics, Data SIO, NOAA, U.S.
Navy, NGA, GEBCO, CNES/Airbus).
inside the microclimate. For microclimate C, where no environmental sensor was
installed, we calculated the average value from both neighbouring microclimates,
namely microclimates B and D.
The installed environmental sensors recorded temperatures from 19th June 2015
until 9th November 2015 at a 15-minute interval. We choose to use the recorded
temperatures for the time period from 21st June 2015 to 20th October 2015 for our
Fig. 4. Average daily temperatures for each microclimate.
614
experiments. Figure 4 depicts the average daily temperatures of each microclimate
for the selected time period.
Microclimate A seems to have a warmer climate, as well as microclimate B which
seems to be a bit cooler than the aforementioned microclimate. On the other hand,
microclimate D has a colder climate than the rest microclimates. While microclimate
C is between microclimate B and D since the temperatures provided are a
combination of the aforementioned microclimates.
2.4 Simulation Scenarios
For both of the following simulation scenarios the simulated field was the former
municipality of St. George (Figure 1) which was divided into four microclimates
(Figure 2). The temperature sets for each microclimate was for the time period from
21st June 2015 to 20th October 2015 and the fruit bearing percentage was set to 75%
for all olive grove areas inside the simulated filed. We run the following two
simulation scenarios, in the first scenario we let the olive fruit flies to develop
without any interference i.e. spraying, in order to observe the temporal evolution of
the population of the olive fruit fly. In the second scenario, a single spray is
conducted, with a mortality rate of 75%, on the 20th July, which is generally the time
when local agricultural authorities commence the first universal spray. Generally,
local authorities conduct three sprays in total. The first spray is conducted between
10th to 20th July, the second is conducted between 10th to 15th August and finally the
last spray is conducted during the last days of September up until 10th October. The
first spray is always a universal spray while the two latter are local sprays, meaning
only specific areas are treated.
Fig. 5. Olive fruit fly evolution in each microclimate without spraying.
615
Fig. 6. Olive fruit fly evolution in each microclimate after single spray.
3 Results
Figure 5 displays the evolution of the population of the olive fruit fly for each
microclimate without conducting any spray during the simulated time. It is obvious
that the temperature set of each microclimate has an effect upon the population
dynamics of the olive fruit fly. For microclimate A and B two generations are visible
Fig. 7. Population evolution comparison for Microclimate A before and after spray.
616
Fig. 8. Population evolution comparison for Microclimate B before and after spray.
and the 3rd generation starts to emerge at the end of the simulated period. On the
other hand, for microclimate C and D only two generations are visible. In addition,
there seems to be a delay to the development and emergence of the first generation,
and consequently for each following one, between each microclimate. The smallest
delay is seen between the neighboring microclimates A and B of four days while the
biggest delay is observed between microclimates A and D with a delay of the latter
of 19 days.
The effect of a single spray one month after the start of the simulation, namely at
20th July, is depicted in Figure 6. The population dynamics of the olive fruit fly in
the areas corresponding to microclimates A and B are greatly affected. However,
that's not the case for the population in microclimates C and D which seem to be
unaffected by the conducted spray.
The following figures compare the population evolution of the olive fruit fly
before and after spraying for each microclimate. In Figure 7, a comparison for
microclimate A can be seen. It is obvious that the spray had a devastating effect upon
the evolution of the olive fruit fly. It is important to note that, the population size
until the end of the simulation was below the infestation threshold.
The comparison for microclimate B is depicted in Figure 8. The spray that was
conducted seems to have greatly affected the evolution of the olive fruit fly.
Although, the effect is not as devastating as is the case for microclimate A, the
population size of each subsequent generation has greatly decreased. The population
size for the first generation is below the infestation threshold and for the second
generation the population size is almost halved when compared to the population
evolution of the first simulation scenario.
On the other hand, microclimate C (Figure 9) is slightly affected by the conducted
spraying. Since, there is only a minor decrease in the size of the population between
the two simulation scenarios.
617
Fig. 9. Population evolution comparison for Microclimate C before and after spray.
Finally, areas belonging to microclimate D seem to be impervious to the spraying
conducted in the second simulation scenario (Figure 10), as if the spraying had never
happened.
4 Conclusions
The effect of microclimates on the population evolution of the olive fruit fly is
investigated. Therefore, we modified our simulation model to be able to divide the
Fig. 10. Population evolution comparison for Microclimate D before and after spray.
618
simulated field into one or more microclimates as specified by the user, where each
microclimate has its own distinct temperature set. The study area was the former
municipality of St. George, Corfu, Greece, which we divided into four microclimates
while the temperature sets for each microclimate were provided from environmental
sensors installed in the area.
The results of the simulation scenarios showed that due to the temperature
difference the warmer a microclimate is, within development thresholds, the faster
adult olive fruit flies emerge. The time difference of first adult olive fruit fly
emergence between the warmer (microclimate A) and the colder (microclimate D)
microclimate was nearly a month.
When a singular universal spray, based on local authorities’ practices, was added
in the simulation run the olive fruit fly population in microclimates A and B is
greatly affected, while in microclimate C the population is ever so slightly affected.
Whereas in microclimate D the spray has no effect whatsoever on the overall
population.
In conclusion, based on the results of the conducted simulation scenarios, as far as
universal-spraying practices goes the existence of microclimates should be taken into
account. After all, for microclimates with a colder climate the chemicals used in the
spray as well as the time to apply it go to waste since it will have a slight effect on
the olive fruit fly population or none at all.
Acknowledgments. The financial support of the European Union and Greece
(Partnership Agreement for the Development Framework 2014-2020) under the
Regional Operational Programme Ionian Islands 2014-2020, for the project “Olive
Observer” is gratefully acknowledged.
References
1. Avidov, Z., 1954. Further investigations on the ecology of the olive fly
(dacus oleae, gmel.) in israel. Ktavim 4, 39-50.
2. Broufas, G., Pappas, M., Koveos, D., 2009 Effect of relative humidity on
longevity, ovarian maturation, and egg production in the olive fruit fly
(Diptera: Tephritidae. Annals of the Entomological Society of America,
102, 70-75.
3. Hodkinson, I. D. (2005) Terrestrial insects along elevation gradients:
species and community responses. Biological Review 80, 489-513.
4. Johnson, M., Wang, X., Nadel, H., Opp, S., Lynn-Paterson, K., Stewart-
Leslie, J., Daane, K., 2011. High temperature a#ects olive fruit fly
populations in californias central valley. Calif Agr 65 (1), 29-33.
5. Kalamatianos, R., Avlonitis, M., 2015. The role of tree distribution and
olive fruit bearing in olive fruit fly infestation. In: Proceedings of 7th
International Conference on Information and Communication Technologies
in Agriculture, Food and Environment. Vol. 1498. CEUR Workshop
Proceedings, pp. 720-729.
619
6. Kalamatianos, R., Avlonitis, M., 2017. Modeling the effect of olive fruit
bearing percentage on Dacus oleae Gmel. (Dipt., Tephritidae) stochastic
dispersion. International Journal of Agriculture Innovations and Research
(submitted).
7. Kalamatianos, R., Avlonitis, M., Stravoravdis, S., 2015. Complex networks
and simulation strategies: An application to olive fruit fly. In: 6th
International Conference on Information, Intelligence, Systems and
Applications (IISA), 2015. IEEE.
8. Mustafa TM, Alzaghal K. 1987. Frequency of Dacus oleae (Gmelin)
immature stages and their parasites in seven olive varieties in Jordan. Insect
Sci. Appl. 8:165–69
9. Rice R, Phillips P, Stewart-Leslie J, Sibbett G. 2003. Olive fruit fly
populations measured in Central and Southern California. Calif Agr
57(4):122-127. DOI: 10.3733/ca.v057n04p122.
10. M.L. Pappas, G.D. Broufas, N. Koufali, P. Pieri, D.S. Koveos, Effect of
heat stress on survival and reproduction of the olive fruit fly Bactocera
(Dacus) oleae. Journal Applied Entomology, 135 (2011), pp. 359–366
11. Tsitsipis JA. 1977. Effect of constant temperatures on eggs of olive fruit fly,
Dacus oleae (Diptera: Tephritidae). Ann. Zool. Ecol. Anim. 9:133–39
12. Tsitsipis JA. 1980. Effect of constant temperatures on larval and pupal
development of olive fruit flies reared on artificial diet. Environ. Entomol.
9:764–68
13. P. Vossen, L. G. Varel, D. Alexandra, Olive fruit fly, Tech rep, University
of California Cooperative Extension - Sonoma County,
http://cenapa.ucanr.edu/files/52578.pdf (2004).
14. P. Vossen, A. Devarrene, 2006, Controlling olive fruit fly at home, Tech.
Rep., University of California Cooperative Extension - Sonoma County,
http://cesonoma.ucanr.edu/files/27230.pdf
15. S. Voulgaris, M. Stefanidakis, A. Floros, M. Avlonitis, 2013, Stochastic
modeling and simulation of olive fruit fly outbreaks. Procedia Technology 8
580-586.
16. Wang, X. G., M. W. Johnson, K. M. Daane, and S. Opp. 2009. Combined
effects of heat stress and food supply on flight performance of olive fruit ßy
(Diptera: Tephritidae). Ann. Entomol. Soc. Am. 102: 727-734
17. L. Wilson, W. Barnett, Degree-days: an aid in crop and pest management,
California Agriculture 37 (1) (1983) 4-7.
620