Ijaer

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283474312

A relative analysis on sound of red palm weevil based on field and lab
recordings

Article  in  International Journal of Applied Engineering Research · January 2015

CITATIONS READS

4 1,340

1 author:

Betty Martin
Sathyabama Institute of Science and Technology
92 PUBLICATIONS   572 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Betty Martin on 28 September 2016.

The user has requested enhancement of the downloaded file.


International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

A relative analysis on sound of Red Palm Weevil based on field


and lab recordings
1
Betty Martin
1. Associate Professor - Sathyabama University, Chennai-119,India

Abstract tissues [2-4]. A visible symptom such as crown loss or leaf


In this investigated work, an effort has been made to detect wilt is seen when the palm has come to a stage of irreversible
the existence of the insect by studying its aural activity and damage. In such cases, sounds of larvae burrowing and
then based on the outcome of the study made, several chewing can be heard with one ear placed to the trunk of palm.
conventional and progressive models of a new system for the The genetic variation of RPW has been previously found using
idea of detecting the presence of pests, has been completed. In random amplified polymorphic DNA analysis. High similarity
the research carried out, the significant acoustical features of was found in random amplified polymorphic DNA (RAPD)
the pest activity have been considered as a main key. Based on
patterns between RPW populations of Kingdom of Saudi
that study some design analysis, modeling, simulation and
methods have been expertized for detection of pests. Using this Arabia and the weevil of Indonesia and no similarity was
foundation, a trial has been made to detect the red palm weevil detected between Egyptian RPW population and other
by using comparative analysis between lab and field based Kingdom of Saudi Arabia or Indonesian population. There is a
sounds of the insect. positive correlation between genetic distances and
geographical distances of RPW.
Keywords: Red Palm Weevil (RPW), acoustical activity,
acoustic features

1. Introduction

The different pests on coconut palms are Rhinoceros beetle


(RB) (Orycytes Rhinoceros) which feeds on tender opening
leaves. Leaf eating caterpillar (Opirina Arenosella walk)
destructs leaves. Red Palm Weevil (RPW) (Rhynchophorus Fig. 1. Image of red palm weevil (India)
Ferrugineus) feeds on tissues of trunk. Eriophyid Mite (Aceria
Guerreronis) affects nut surface[8,10,11]. Coconut white grub Due to the destruction caused by these pests, detecting this
(Leucopholis Coneophora) affects growing roots of coconut insect species living in the interior part of the stem mainly by
palm. Mealy bugs and Scales (Pseudo Coccus longispinus and means of a sensor using its acoustic activity was of growing
Aspidiotus destructor) infest spindle and unopened spathe. importance. [1-3] used acoustic system with vibration sensor
Coreid Bug (Paradasynus Rostratus) sucks sap from to monitor the activities of hidden insects on soil and interior
developing buttons and causes immature nuts to fall and can structure of plants in lab as well as field conditions. [8] used
cause malformation of nuts. Rat (Rattus Rattus Wroughtonni) acoustic techniques to monitor and detect hidden insect
lives on palm crown and damages tender nuts. In the above infestation of plants in lab. Here, an acoustic sensor in the
listed group of insect pests on palms, it has been found that the form of digital recorder to record the activities of the hidden
red palm weevil (RPW) is the most influential pest and leads pest was implemented. Other than these methods of detection
to the death of tree [2]. The main objective of the research is to and control, acoustic detection gives immediate results of
detect the red palm weevil which is the most deadly pest. Due existence of RPW. The acoustic activities of RPW captured
to its high reproduction rate, RPW prefers to live with no other from pest infesting farms were analyzed. Data was acquired
insect in one trunk. The RPW known as Asian or Sago palm by digital voice recorder which could record low frequency
weevil affects palm of age 3-15 years. As the name suggests, signals.
they are red coloured and 2 to 4 cms long. The larva of red
palm weevil tunnels through the trunk of palms weakening the To sum up, the only way of detection of the pest inside lab in
stem thereby causing mortality of the tree trunk. The male an early stage is by acquiring its acoustic patterns which it
RPW is characterized by the presence of a series of black hairs generates during crawling, eating, or gnawing inside the
on the dorsal and frontal parts of snout. Female RPW do not trunk. This is feasible by capturing acoustic signals by using a
exhibit hairs. The maximum span of adult RPW for female is suitable recorder. With this background, the next section deals
76 days and 113 days for male. The RPW has 3 with materials and methods involved in detecting the pest
generations/year. The shortest generation is of 100 days and with the sensor.
longest of 127 days. Figure1 shows the image of RPW.
Symptoms of infestation are mostly yellowing and wilting of 2. Materials and Methods
palms which can lead to death of affected plant. The crown
wilts first and lower leaves follow due to damage of vascular

5261
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

Analysis on Laboratory Recordings From these observations, it is clear that from the spectrum
of Figure 1 and Figure 2 the sound level of slow biting signal
is -24.2 dB lesser than the normal biting signal which is -19.6
Details on the techniques used towards analyzing the
dB. This infers that slow biting signal has lesser sound level
acoustic signals generated by RPW are presented. Acoustic
than normal biting signal.
signals generated by RPW in the palm and in the laboratory
have been taken for this analysis.
2.1.2. Analysis on eating pattern of RPW
2.1Categorization of Acoustic Pattern of RPW from
Eating can be defined as the process of chewing the food
Laboratory Recordings
substance and collectively swallowing it into the mouth which
is subsequent to biting action. Eating action is divided
During recording in laboratory the acoustic activity of
into Clear Eating and Slow Eating. At a frequency peak of
RPW actions inside the sound proof chamber, was vigilantly
2219 Hz with -17.2dB the spectrum exhibits clear eating, the
watched. The acoustic activity recorded in the sound proof box
spectrum presented a frequency 1163 Hz at -15.2 dB for slow
is classified as biting, eating and moving. In these sorting
eating and is tabulated in Table 2. Figure 4 - Figure 5 show the
methods, the sound level and frequency of each spectrum are
spectrum for different eating patterns of RPW.
the key factors which decide the presence of RPW.
Table 2. Frequency of eating pattern of RPW
2.1.1. Analysis on biting pattern of RPW
Sound Level
Activity Frequency (Hz) (dB)
Biting is the process of breaking the food substance into
smaller pieces when the upper and lower mandible meets each Clear Eating 2219 -17.2
other. Biting action is further divided into Normal Biting and Slow Eating 1163 -15.2
Slow Biting. With normal biting, the observed frequency
peak in the spectra displayed is 1651 Hz at a sound level of –
19.6 db, and at a frequency peak of 1588 Hz with -24.2db the
spectrum exhibited slow biting as tabulated in Table 1. Figure
2 and Figure 3 show the biting patterns of RPW[4-6].

Table 1. Frequency of Biting pattern of RPW

Activity Frequency (Hz) Sound Level


(dB)
Normal biting 1651 -19.6 Fig. 4. Spectrum for clear Eating
Slow Biting 1588 -24.2

Fig. 2. Spectrum for Normal Biting;


Fig.5. Spectrum for Slow Eating

From these observations, it is clear that from the spectrum


of Figure 4 and Figure 5, the sound level of clear eating signal
is -17.2 dB less than the slow eating signal which is -15.2 dB.
This infers that clear eating signal generates less sound than
slow eating signal.
2.1.3. Analysis on moving pattern of RPW
Movement of an insect is the process of crawling or
changing position or moving from one place to another.
Movement action is further divided into Normal movement ,
Fig. 3. Spectrum for slow biting

5262
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

Slow movement and Lengthy movement .While in normal the sound level of eating signal produces more sound than that
movement the observed frequency peak in the spectrum of biting signal. This study made on sound level emanating
displayed a range of 1652Hz - 2168 Hz at a sound level of + from the red palm weevil could help in the design and
9.9 db, and at a frequency range of 1651 Hz to 2201 Hz with development of a system to detect red palm weevil.
+10.5db to 9.7 dB the spectrum exhibited slow movement, the To summarize the recordings done in laboratories, the
spectrum presented a frequency range of 1003 Hz to 1600 to presence of the insect pest could be detected by its unique
2151 Hz at 0.5 db for lengthy movement as tabulated in Table traits namely eating, biting or crawling actions. Hence these
3. The spectrum for movement did not exhibit consistent activities have been categorized as biting, eating and
frequency pattern. Figure 6 – Figure 8 shows different patterns movement. Among these actions the sound level for eating
of movement of RPW. signal is high when compared to the biting sound signal. But
movement of RPW signal could not be compared since no
Table 3. Frequency of movement of RPW evenness existed for evaluation. Comparison between the
sound level of RPW`s signal activity, helped in detecting the
Activity Frequency (Hz) Sound Level (dB) presence of hidden RPW inside the trunk.
1652-2168 9.9 Analysis on Field Recordings
Normal movement
Slow movement 1651-2201 9.7 2.1.4 Discrimination of RPW Sounds with other Sound
Lengthy movement 1600-2151 0.5 Existing Patterns

A recording from the field includes the sound made by


moving vehicle, human whisper, wind sound and RPW signal
if available. By observation, it was found that the spectrum
from a recording with a moving vehicle, wind or whisper does
not include peaks. As can be seen from the observation of
figures given, the waveform remains within the recording
depending on the time for the noise to pass out of the range of
detection. Hence these signals do not appear as clicks. Figure
Fig. 6. Spectrum for slow movement 9 shows the typical pattern that appears for an infested sound
of RPW. Figure 10 shows the pattern that exists for a
noninfested sound of RPW. Figure 11 shows the disturbed
recorded pattern that comprises the sound of wind, human
whisper, moving vehicle and infested sounds of RPW done on
coconut plantations.

Fig .7. Spectrum for lengthy movement

Fig.9. Infested pattern of RPW

Fig. 8. Spectrum for crawling movement

From the observed spectrum, comparison of movement signal


cannot be made on biting and eating signals since there is no
Fig. 10. Noninfested pattern of RPW
uniformity in the results obtained. But it can be concluded that

5263
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

Fig. 12. Waveforms of different sound patterns in field


Fig. 11. Recording containing patterns of A)RPW B)moving recording
vehicle C)human whisper

Through the advanced methods of signal processing to identify The initial recording during field visit captures the
spectral and temporal patterns that distinguish targeted pests activity of the hidden pest in time domain, which is then
from other animals and background noises, human listener transformed into frequency domain. The spectrum determines
remains significantly better in successful discrimination of the frequency of particular acoustic pattern. The sound spectra
insect sounds for identifying chewing, crawling and moving for each input is constructed. It varies between the minimum
activities [8]. The sound patterns for an infested, noninfested level of frequency peaks and maximum level of sharp peaks.
and disturbed recording confer more information in The minimum and maximum level of frequency is tabulated in
discriminating the sound of RPW with other existing sound Table 4. The activities of all sound patterns exhibit significant
patterns[11-12]. This experimental pattern can be used as a peaks. Figure 13 to Figure 16 show the frequency spectra of
reference for comparing the signal recorded in test fields. This different inputs. The results obtained give the frequency of
field study enables to design and develop a particular system RPW and other back ground noises observed during recording.
used to detect the presence of pests.

2.2.2. Spectral Profile Obtained in Field Recordings

Based on multiple recordings it is inferred that the


background noise is distinguishable from the acoustic activity
of the weevil. The sound recording of different patterns is used
as a profile to match against individual sound and then identify
Fig. 13. Spectrum for RPW
the source of sound. Some profiles are constructed for
background sounds which include wind, whisper and trucks.
Noises from vehicles, wind blow, birds chirp, and human
whisper contain long continuous signals rather than the short
broadband impulses of weevils [2]. Both the signals appear as
impulses, but weevil signals occur in longer trains with more
impulses. The waveforms are sliced for each pattern from
Figure 10 and represented as separate acoustic pattern as
shown in Figure 12 [6].

Fig. 14. Spectrum for moving vehicle

5264
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

3.1 Acoustic Pattern of Red Palm Weevil

Acoustic emission is a sound wave or a stress wave within the


material due to sudden release of strain energy[9-10].
Acoustic emission monitoring is a nondestructive means of
evaluation by which the sound waves generated by stresses
within a material are being passively listened or heard. Sound
Fig. 15. Spectrum for wind emission emitted by larva is represented by clicks. They occur
at intermittent intervals with different intensity. The activity
may vary for a few hits every 1-2 sec. Figure 17 a,b shows the
activity of a single acoustic hit of RPW in a few millisecond
obtained during the recording compared against survey[7].

Fig. 16. Spectrum for human whisper

Table 4. . Range of frequency and Sound level of different


sounds

Minimum Value Maximum Value


Name of sound
Hz dB Hz dB Fig. 17.(a). Acoustic activity of RPW
RPW 1370 -13.9 2019 -17.2
Moving vehicle 68 -3.3 450 -11.7
Wind 22.5 -0.1 155 -5.0
Human Whisper 43 -4.2 692 -0.2

From the spectral profile of each input the frequency


ranges are noted along with its sound level in dB. From the
Table 4, it can be observed that each input has its own
frequency range and the back ground noise has comparatively
lower frequencies when compared to the RPW species. From
the observation, it can be stated that the recordings obtained in
both the laboratory and the test in the field exhibit almost the
same frequency for a RPW signal. From the comparison made
between the laboratory and field recordings of Table 1 to 4, it
can be concluded that the frequency for RPW signal in both Fig. 17.(b) Acoustic pattern of RPW
cases are within the similar ranges. The above study made will
help in design and development of a novel system for detection 3.2 Frequency Spectrum for RPW
of RPW.
The frequency spectrum of a time domain signal is the
representation of the signal in frequency domain. In the Figure
3. Validation of Results 18a[9], frequency spectrum distinctly shows the spikes ranging
between 4000Hz – 6800Hz and the peak is noticed at
Validation of acoustic data acquisition can be done by a few 4591Hz.as compared to figure 18 b.
methods based on the literature survey done on RPW and also
by destructive sampling. They are,
 Acoustic Pattern of Red Palm Weevil
 Frequency Spectra of Red Palm Weevil
 Dissection of palm.

5265
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

The recorder set was easy to use, battery operated and


portable. It was simple to carry it from one palm to another
without much difficulty. It involved no invasive inspection of
trunk to insert the sensor which otherwise becomes a loophole
to RPW to enter into the tree easily and to bore the tree tissues.
All recordings were done in laboratories and on the footing of
trees with an average age of 3-15 years in coconut plantations.
Laboratory recordings made on RPW and further analysis done
on their acoustic activity proved their habitation and their
Fig. 18. (a) Spectrum for infested pattern of RPW effect on plantation. From experimental observations made in
fields it is understood that RPW do not allow other coconut
pests to live together in the same tree trunk. The RPW signal
obtained in laboratory was compared against the field signal of
RPW and is being correlated. Having validated the sounds of
RPW by the above analysis, this study is further used for the
design and development of an electronic system for other
insects.

References
Fig. 18. (b) Spectrum for infested pattern of RPW
[1] Betty Martin and Vimala Juliet, (2012) “A Novel
Approach to Identify Red Palm Weevil on Palms” in the
proceedings of 2nd International Conference on Chemical,
3.3 Dissection of palm Material and Metallurgical Engineering ICCMME, China
Dec 14-15, 2012.
For further validation, the confirmation of weevil [2] Mankin et al (2000) “Eavesdropping on Insects Hidden in
Soil and Interior Structures of Plants”.Journal of Economic
presence in palms was done by dissecting the tree. In every Entomology,Vol. 93 No 4, pp.1173-1182, Aug. 2000.
field visit, after getting the consent of plantation owner, the [3] Betty Martin, Vimala Juliet, (2011)“ Detection of red
particular affected coconut tree was cut and to check for the palm weevil against RB in coconut plantations of India by
presence of pest. Though the affected trees look outwardly means of signal processing” International Journal of Digital
fresh, the recorder identified it`s unhealthiness. When cut Signal Processing-IJDSP CiiT pp no.450 Vol No9 Oct
2011.,0974-9705.
open, it is found that the inside of the tree is totally affected by [4] Betty Martin, Vimala Juliet, Sankaranarayanan and Gopal
lots of infestation by red palm weevil comprising cocoons, Aravamudan “Wireless Implementation Of Mems
neonates, a few adults deceased with their larvae as shown in Accelerometer To Detect Red Palm Weevil On Palms" in
Figure 19. The dissected tree has been instantly burnt and the proceedings of International Conference on Advanced
removed from the site to avoid further pest spread. Electronic Systems (ICAES-2013) ,CSIR-CEERI Pilani,
Rajasthan Sept. 21-21, 2013.
[5] Betty Martin, P.E. Shankaranarayanan, Vimala Juliet and
A. Gopal’ Identifying Sound of RPW In-Situ from External
Sources’ in Artificial Intelligence and Evolutionary
Algorithmsin Engineering Systems, journal of Advances in
Intelligent Systems and Computing’, Vol. 324, ISBN 978-81-
322-2126-5,Jan 2015. Springer
[6] Betty Martin, S.Maflin Shaby , M.S. Godwin Premi
,”Studies on acoustic activity of red palm weevil the deadly
pest on coconut palms, in the proceedings of second
international conference on nanomaterials and technologies
CNT 2014 , 17th -18th Oct 2014, Hyderabad, India
[7] Ilyas(2007)`Automatic bioacoustic detection of
Rhynchoporous ferrugineus’
[8] R. W. Mankin, (2002) ` Acoustic Detection of Termite
Infestations in Urban Trees’ J. Econ. Entomology. 95(5):
981to 988
[9] Abelardo Gutiérrez et al(2010), Development of a
bioacoustic sensor for the early detection of Red Palm
Fig. 19. Dissected tree with larva inside Weevil Crop Protection 29,671 to 676.
[10] Vidyasagar, P.S.P.V. and S. Keshava Bhat. (1991).
Pest management in coconut gardens. Journal of Plantation
Conclusion Crops 19(2):163-182.

5266
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

[11] F.Fleurat-Lessard, et al() Acoustic detection and


automatic identification of insect stages activity in grain
bulks by noise spectra processing through classification
algorithm 9th International Working Conference on Stored
Product Protection 476 PS5-13 – 6304
[12] V. Soroker et al. (2005) Phytoparasitica 33(1) : 97-
106 Current Status of Red Palm Weevil Infestation in
Date Palm Plantations in Israel.

5267
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 6 (2015)
© Research India Publications ::: http://www.ripublication.com

5268

View publication stats

You might also like