=Paper=
{{Paper
|id=Vol-1152/paper49
|storemode=property
|title=Modelling Movements of Root-knot Nematodes Meloidogyne spp. Juveniles when Encumbered with Spores of Pasteuria penetrans
|pdfUrl=https://ceur-ws.org/Vol-1152/paper49.pdf
|volume=Vol-1152
|dblpUrl=https://dblp.org/rec/conf/haicta/VagelasPG11
}}
==Modelling Movements of Root-knot Nematodes Meloidogyne spp. Juveniles when Encumbered with Spores of Pasteuria penetrans==
https://ceur-ws.org/Vol-1152/paper49.pdf
Modelling movements of root-knot nematodes
Meloidogyne spp. juveniles when encumbered with
spores of Pasteuria penetrans
Ioannis Vagelas1, Barbara Pembroke2 and Simon R. Gowen2
1
Department of Plant Production, Technological Education Institute of Larissa, 41110
Larissa, Greece, e-mail: vagelas@teilar.gr
2
School of Agriculture, Policy and Development, University of Reading, Earley Gate, PO
Box 236, Reading, RG6 6AR, UK, e-mail: b.pembroke@reading.ac.uk
Abstract. A system for monitoring movement of the free-living second stage
juveniles of root-knot nematodes (Meloidogyne spp.) using digital image
analysis is described. The method is based on the analysis of video sequences
of movement of individual random nematodes encumbered with or without
Pasteuria penetrans endospores. Software packages were used to grab video
images, to process images and to monitor the movement of selected body part
positions over time. Methods include the study of nematode locomotion based
on (a) the geometric center (b) the centroid body point, (c) tracking of two or
four selected body points and (c) tracking a rectangular shape area produced by
the nematode’s body. Data showed that (a) the normal sinusoidal movement of
nematodes is changed when individuals are encumbered with spores of P.
penetrans and (b) in all cases a significant greater motility was observed by
nematodes without P. penetrans spores attached.
Keywords: Nematode movement, digital image analysis, motion analysis.
1 Introduction
Pasteuria penetrans (Thorne, 1940) is a mycelial, endospore forming bacterial
parasite of plant parasitic nematodes (Mankau, 1975; Imbriani and Mankau, 1977)
showing promising results in a biocontrol strategy of root-knot nematodes
(Meloidogyne spp.) (Stirling, 1991). The endospores attach to the outside nematode
body wall (cuticle) of the infective stage, the second-stage juveniles (J2) of
Meliodogyne populations (Mankau, 1980). After the J2 penetrates a plant root and
begins to feed, the bacterium penetrates the nematode body wall and begins to grow
and develop in the developing nematode (Mankau and Imbriani, 1975; Imbriani and
Mankau, 1977; Sayre and Wergin, 1977). Eventually, the female nematode body
becomes completely filled with spores (Sayre and Wergin, 1977; Stirling, 1991).
Each infected female may contain up to 2.5 million spores (Darban et al., 2004),
which are eventually released into the soil.
BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB
&RS\ULJKWE\WKHSDSHU¶VDXWKRUV&RS\LQJSHUPLWWHGRQO\IRUSULYDWHDQGDFDGHPLFSXUSRVHV
,Q 06DODPSDVLV$0DWRSRXORV HGV 3URFHHGLQJVRIWKH,QWHUQDWLRQDO&RQIHUHQFHRQ,QIRUPDWLRQ
DQG&RPPXQLFDWLRQ7HFKQRORJLHV
IRU6XVWDLQDEOH$JULSURGXFWLRQDQG(QYLURQPHQW +$,&7$ 6NLDWKRV6HSWHPEHU
559
The potential of P. penetrans to control of root-knot has been widely studied
(Gowen et al., 2008) including distribution, host range, and specificity. Successful
parasitism depends on the attachment of 5-10 spores per juvenile, which is sufficient
to initiate infection without reducing the ability of the nematode to invade roots
(Davies et al., 1988; Rao et al., 1997). There may be little or no root invasion if there
are greater than 15 spores attached, inferring that spore attachment will affect the
ability of a J2 to locate and/or invade a root (Davies et al., 1988). Few attempts have
been made to quantify the effect of P. penetrans spore attachment on the movement
of infective root-knot nematode juveniles.
Nematodes move by undulations or wave-like motions through dorsal/ventral
contractions of the body (Buchsbaum et al., 1987; Storer et al., 1979) similar to an
undulatory swimming motion of eels (Tytell, 2004) and as larval chironomids
(Brackenbury, 2003). As one segment of the body contracts, it “pulls” the remainder
of the body forward along the body in a head to tail direction (Brackenbury, 2000).
There are no previous experimental data to model root-knot nematode movement
encumbered with or without P. penetrans spores. In this paper a study using digital
image analysis is made of the movement of the root-knot nematode (Meloidogyne
spp.) and how this is affected when sporesof the bacterium Pasteuria penetrans (a
naturally occurring nematode parasite) are attached to the cuticle.
2 Materials and methods
2.1 Nematode cultures
Nematode cultures (M. javanica) were maintained on tomato plants in the glasshouse
and fresh second stage juveniles (J2) were collected from infected tomato roots using
the methods described by Hussey and Barker (1973).
2.2 Preparation of Pasteuria penetrans spores
A commercial product of Pasteuria penetrans (Pp) (Nematech Co. Ltd., Japan) was
used in this study. Fresh J2 were encumbered with Pp spores as described by Darban
et al. (2004). Nematodes with 5-10 spores attached were considered as the Low Pp
and J2 with 15-25 spores were the High Pp treatment.
2.3 Acquisition of the video images
Nematode locomotion was tracked with an inverted microscope (MICROTEC 200)
mounted with a digital camera (Aptiva 3.2 Megapixel). In all cases a nematode’s
movement was observed in water in a 9 cm Petri dish. Nematode movement was
recorded in 30 second video sequences. The microscope magnification was x 100 for
Figure 1, x 200 for Figures 5, 6 and 7 and the highest magnification for Figure 3. All
video sequences showing nematode locomotion were observed on Movie Maker 2
(Microsoft software) before any further analysis.
2.4 Image extraction
Before analysis, images (frames) grabbed from selected 30 sec videos were obtained.
We used a video decompiler (SC Video decompiler software program) to extract
560
frames. A total of 390 frames saved in jpg format were obtained from a 30 sec video.
Original images were 320 x 240 pixels or 3.3 x 2.5 inches.
2.5 Measurement of nematode movement
Measurement of movement (in inches) was performed using image analyzer software
Scion Image for Windows (Scion Corporation, www.scioncorp.com).
2.6 Image processing and analysis
All frames were saved in a tiff format before importing to the Scion Image software
package. For image analysis the 39 frames were aligned in ranks of 10 as the 1st, 11th,
21st and the 31st frame. When an image file (*.tiff) was imported to the Scion Image
software program, measurements were performed with the Manual Area
Measurement selecting the Measure command of the program. In order to understand
the aspects of nematode body posture, measurements were taken of the body
movement. For each frame, five measurements were taken; (a) tracking of the
nematode geometric center; (b) tracking the nematode centroid body part; (c)
tracking of two or four selected body points and (d) tracking a rectangular shape area
produced by the nematode’s body. In detail, the measurements of the five above
measurements are:
2.6.1 Nematode locomotion
Using the rectangular selection tool of the Scion Image Program and moving the
mouse on the nematode image, we fitted all faraway nematode body segments in a
rectangle shape in order to estimate the X-Y geometric center (using the
measurement options Area, X-Y Center of the Scion Image program). With the
geometric center (X-Y center of the rectangle area) we were able to track the
geometric point of the nematode body, or very close area matches in that point.
Further, we extracted metrics to an Excel data spread-sheet and presented all data to
GenStat 7th Edition to create a scatter plot. The trends of two individual J2/treatment
movement directions were randomly selected and shown in a scatter plot.
2.6.2 Tracking the nematode centroid point and two or four nematode body points
The centroid points, of four random J2 per treatment were tracked using the Elliptical
Collection Tool of the Scion Image Program follow the procedure described above.
Using the Cross Hair Tool, and moving the mouse on the nematode image we
marked four equal-distance nematode body points and their X-Y coordinates were
recorded for each frame and analysed in the Program Results window. The points
were: 1 the head; 2 the esophagus; 3 the centre of the gut: 4, the tail. All data were
extracted as X and Y values on the program Result window and exported to an Excel
data spread-sheet.
Based on those data we displayed the nematode per treatment movement, with the
centroid body point and the relationship of four nematode body parts or the
relationship of the two intermediate (Ym1 and Ym2) nematode body points
movement in all Pp treatments.
Moreover, the equation of nematodes movement based to the centroid point were
obtained by fitting the data sets (X, Y values) to CurveExpert 1.3, a curve fitting
system for Windows by using the Program CurveFinder command.
561
Further, the equation of nematodes movement based to four or to two intermediate
Ym1 and Ym2 nematode body parts were obtained by fitting the data sets values to
GenStat by using the Standard Curves command of the Nonlinear Regression
Analysis.
2.6.3 Tracking the nematode rectangular shape area
Using the Rectangular Selection Tool of the Scion Image Program and moving the
mouse on the nematode image, we fitted all faraway nematode body segments in a
rectangular shape in order to estimate the rectangular shape area, using the
measurement options Area of the Scion Image Program. Further, we extracted
measurements to Excel and represented the different rectangular shaped areas
produced by nematodes encumbered or unencumbered with Pp endospores.
2.6.4 An estimation of J2 motility
The locomotion of nematodes treated with low and high or without P. penetrans
spores were estimates based on the nematode wavelength (Ȝ) and the distance moved
over time. The body lengths of each nematode were measured using the Straight Line
Selections Tool of the Scion Image Program and moving the mouse on the nematode
image marked the nematode head and the tail each time. The same procedures were
used to measure the distance covered by the nematode head over time t1 and t2
(frames 1 and 2) (up to 15 frames) moving the mouse from X1,Y1 (frame 1) to
X2,Y2 position on frame 2. Data were exported to an Excel data spread-sheet as
described above. The absolute body length of a J2 in this research is equal to a 105
pixels as presented in Figures 6 and 7 with a dotted line. Measurements were based
on 10 nematodes per treatment and with 15 frames per individual nematode. The
time between two frames in sequence was 11.5 sec.
2.7 Statistical analysis
All regression analyses were performed to GensStat 7th Edition statistical program.
Scatter line plots (Figures 1-3) were performed to MinTab 13th Edition statistical
program and all other graphs were made to GraphPad Prism5 statistical program.
3 Results
3.1 Nematode locomotion
Over the same time period, J2 without P. penetrans spores attached can move further
than those J2 encumbered with low numbers of spores (Figure 1). That means that
the body of a J2 without Pp spores moves in a direction explained by a linear
regression r2 96-98%. Attachment with P. penetrans spores probably disrupts this
natural behaviour.
3.2 Tracking the nematode centroid point and two or four nematode body
points
When data sets based on the nematode centroid body point were fitted to
CurveExpert, the best equation to explain the J2 body movement without P.
562
penetrans spores(Figures 2.1a-2.4a) is the sinusoidal fit, y=a+bcos(cx+d) with
correlation coefficient r1=0.9368, r2=0.9396, r3=0.9345 and r4=0.9090 (R2
parameters based to four random nematodes). When nematodes were encumbered
with P. penetrans sporesthere was no sinusoidal movement (Figures 2.1b-2.4b and
Figures 2.1c-2.4c).
Our simple tracking system for extracting data from the four or the two
intermediate nematode body part points has shown that nematodes without P.
penetrans spores attached have a sinusoidal movement producing a double Fourier
curve (equation 1). This is represented in Figure 3A where the Ym1 line presents the
esophagus and the Ym2 line the gut (R2 Ym1+Ym2 = 86.0, P<0.001) and in Figure 3B
where the tracking points are the head, the esophagus, the centre of the gut and the
tail (R2 head +esophagus + gut +tail = 78.4, P<0.001).
J2 without Pp end point
J2 without Pp start point
.
Fig. 1. Nematode travel position without P. penetrans spores (circle solid dots) and with low
P. penetrans spores (square solid dots, top). The arrows indicate the direction of movement (X,
Y axis units in pixels). The nematodes without P. penetrans spores move in a line explained by
2
a Linear regression fit y=a+bx with a correlation coefficient (R ) equals to 0.9868 and 0.9679
for the two random individual nematodes. Axis scales x, y are measured in pixels (200_250
pixels), value for a straight J2 body length is 10 pixels.
Based on the movement of the four body points, the best curve to explain our data
is also the double Fourier curve. This is a compound of two sine waves as presented
in Figure 4, one having half the cyclic period of the other.
Y = a + b*sin(2*pi*(X-e)/w) + c*sin(4*pi*(x-f)/w) equation 1
563
A J2 encumbered with A J2 encumbered with
A J2 free of Pp spores fits low Pp spores fits in a high Pp spores, fits in box
a rectangle shape with a smaller box with a slow with no forward
forward movement* movement (figures 1-4b in movement (figures 1-4c in
(figures 1-4a in a a column) a column)
column)
Box a Box b Box c
6
Figure 2.1a, t=0sec Figure 2.1b, t=0sec Figure 2.1c, t=0sec
Figure 2.2a, t=10sec Figure 2.2b, t=10sec Figure 2.2c, t=10sec
Figure 2.3a, t=20sec Figure 2.3b, t=20sec Figure 2.3c, t=20sec
Figure 2.4a, t=30sec Figure 2.4b, t=30sec Figure 2.4c, t=30sec
*Note that the grid lines show that there is a significant movement of J2 without
Pp spores compared to those encumbered with Pp spores Figs 1-4b and 1-4c.
Fig. 2. Nematode body wave formation presented as a rectangle shape movement over time. J2
were encumbered with Pp spores column b (low Pp) and c (high Pp) or without Pp spores
column a. Arrows indicate the J2 speed.
564
Fig. 3. Motion analysis and position of nematode body parts over time. Nematodes were not
encumbered with P. penetrans endospores.. Analysis was performed using GenStat Statistical
Package based to (a) Ym1 and Ym2 data sets and (b) to four nematode tracking points (head,
esophagus, gut and tail). Y axis, in promotion to nematode body length which is equal (in this
study) to 1.8. and X axis, in frames where 39 frames are equal to 30 sec.
Ym1,Ym2 & Tail
Head
Fig. 4. Motion analysis (4a) and position (4b) of nematode body parts (XY) over time.
Nematodes were free P. penetrans spores. Analysis was performed based to the Head, Ym1,
Ym2 and to the J2 tail data sets. R2 = 74.9, P<0.001. Axis scales x, y are measured in
promotion to nematode body length which is equal (in this study) to 2.0.
565
The observation that a nematode body produces a sine wave is further tested using
the MotionPro software program ver. 4.4.2., by PDSofTec (www.pdsoftec.com),
based on head region turns. The directions of head region rotation were observed
based on nematode movement and displayed as a set of points (dots) in real time
when nematodes moved (Figure 4).
With nematodes free of spores we observed that the 2nd, 3rd and 4th body part
follows the head movement. This does not happen when the nematode is encumbered
with P. penetrans spores at low or high density (Figures 2.1b-2.4b and Figures 2.1c-
2.4c).
3.2.1 Tracking the nematode rectangular shape area
The analysis showed that the nematode without P. penetrans spores produced a
rectangular shape with a significantly greater area compared to nematodes
encumbered with spores (Figure 5). This was confirmed with the Movie Maker 2
software program where each nematode video was observed (Figure 2). We conclude
that nematodes without P. penetrans spores fit a rectangle shape with a strong
forward movement equal to the rectangle shape in Figure 2 marked as box a, whereas
nematodes encumbered at a low or high density of spores could fit to rectangles
equal to boxes b and c in Figure 2. There was little movement of J2 encumbered with
high P. penetrans attachment and several times they were seen to collide.
2
1,6
J2 body lenght
1,2
0,8
0,4
0
Major Minor Major Minor Major Minor
J2 without Pp J2 with low Pp J2 with high Pp
Fig. 5. Rectangle estimation results for J2 motion treated with (left) or without P. penetrans
spores (middle and right). The major (long) rectangular side presents J2 forward movement
where the minor (short) presents the width of the sinusoidal J2 body motion. The value of two
(2) in Y axis is equal to a value for a straight J2 body length.
3.2.2 An estimation of J2 motility
The measurements based on the nematodes locomotion shows a significantly greater
wavelength (Ȝ) and distance movement values for nematodes without spores attached
compared with nematodes encumbered at a low or high density of spores (Figures 6
and 7 respectively). Nematodes encumbered at a high spore density showed
insignificant movement (Figure 7) confirming observations shown in Figure 2 where
566
nematodes’ faraway body segments can be fitted in a box with an insignificant no
forward movement.
Fig. 6. Differences in J2 body length [= a nematode wavelength (Ȝ)] during motion in
treatments with or without P. penetrans spores. Dotted line represents a straight J2 body length
which is equal (in this study) to the value 105 in Y axis.
Fig. 7. Distance moved by J2 with or without P. penetrans spores (N=10) in 15 sequential
frames (=11.5 sec). Dotted line represents a straight J2 body length which is equal to the value
105 in Y axis.
567
3 Discussion
In this paper we present a technique to track a plant parasitic nematode movement
using a digital camera and an inverted microscope. Similar techniques where shown
in Baek et al. (2002) and Cronic et al. (2005) where the authors used a microscope
fitted with a camera and a videtaping (VCR) system to record movement of
Caenorhabditis elegans.
Using free Internet software packages such as the SC Video Decompiler we
extracted images (frames) from video files (.avi format) for further analysis. This is
similar to Cronic et al. (2005) who took their data to a PC using a Matrox Meteor-
II/Standard video frame grabber hardware and the Recognizer 2.1 software package.
Further, with the commercial software package Scion Image we collected data
taking selected nematode body points. Similarly, Cronic et al. (2005) presented the
same ideas for data extraction and processing in Matlab called Wormproc; they
showed many C. elegans measurements and histograms grabbed at 13 points and
from the centroid point of the body respectively. Greng et al. (2004) identified and
tracked separately the head and tail movement of C. elegans. In our studies we
employed the GenStat statistical program and the Standard Curve Routine, a tool of
the regression analysis and we described effectively with the double Fourier curve
the motion of four or two body points in nematodes without Pp endospores. Wallace
(1958; 1959), Baek et al. (2002); Cronic et al. (2005) concluded that nematodes
produce a sinusoidal movement. In our studies we showed that the nematode motion
is a compound of two sine waves, one having half the cyclic period of the other
(double Fourier curve).
A natural characteristic is that all nematode body parts follow the movement of
the nematode head (Niebur and Erdos, 1991) and this is demonstrated. However this
did not occur when the nematodes were encumbered with Pp endospores, probably
because the spores impeded forward movement. Moreover we show that J2
encumbered with high numbers of sporesshow no forward movement and several
times were observed to collide with other nematodes.
The nematode rectangular shape area produced by a nematode’s body parts proved
a very good estimator to describe nematode locomotion for nematodes without spores
and those encumbered with low and high numbers.
Nematodes without spores move faster than those that are encumbered. Moreover
those nematodes moved in a straight line in the same direction and covered a longer
distance than those with endospores. Those nematodes pulse the body with a
wavelength (Ȝ) equal to a straight body position, probably this is a random walk with
memory as suggested by Hapca et al. (2007). Wallace (1958) observed the same for
Heterodera schachtii migration in soil and he concluded that the maximum speed of
the H. schachtii J2 is attained when there is no lateral movement and each part of the
body follows the part immediately in front of it.
Finally it can be suggested that this research could be improved with more
emphasis on mathematics developing codes e.g. on the Matlab software package.
Further the methods (the technique to track a plant parasitic nematode movement
using a digital camera and a microscope) could be improved in wider range.
568
References
1. Baek, J-H., Cosman, P., Feng, Z., Silver, J and Schafer, W.R. (2002) Using
machine vision to analyze and classify Caenorhabditis elegans behavioral
phenotypes quantitatively. Journal of Neuroscience Methods, 118, 9-21.
2. Brackenbury, J. H. (2000) Locomotory modes in the larva and pupa of
Chironomus plumosus (Diptera, Chironomidae). Journal of Insect Physiology,
46, 1517–1527.
3. Brackenbury, J. H. (2003) Swimming kinematics and wake elements in a worm-
like insect: the larva of the midge Chironomus plumosus (Diptera). Journal of
Zoology London, 260, 195-201.
4. Buchsbaum, R., Buchsbaum, M., Pearse, J and Pearse, V. (1987) Animals
Without Backbones. Chicago: University of Chicago Press.
5. Cronic, C.J., Mendel, J.E., Mukhtar S., Kim, Y-M., Stirbl, R.C., Bruck J and
Sternberg. P.W. (2005) An automated system for measuring parameters of
nematode sinusoidal movement. BMC Genetics, 6, 1471-2156.
6. Darban, D.A., Pembroke, B and Gowen, S.R. (2004) The relationships of time
and temperature to body weight and numbers of endospores in Pasteuria
penetrans-infected Meloidogyne javanica females. Nematology, 6, 33-36.
7. Darban, D.A., Pembroke, B and Gowen, S.R. (2004) The relationship of time
and temperature to body weight and numbers of endospores in Pasteuria
penetrans-infected Meloidogyne javanica females. Nematology, 6, 33-36.
8. Davies, K.G., Kerry, B.R and Flynn, C.A. (1988) Observations on the
pathogenicity of Pasteuria penetrans, a parasite of root-knot nematodes. Annals
of Applied Biology, 112, 491-501.
9. Gowen, S.R., Davies, K.G and Pembroke, B. (2008) Potential use of Pasteuria
spp. in the management of plant parasitic nematodes, in Integrated management
and biocontrol of vegetable and grain crops nematodes, eds. A. Ciancio and
K.G. Mukerji, Dortrecht, The Netherlands: Springer, 197-210.
10. Gray, J and Lissmann H.W. (1964) The locomotion of nematodes. Journal of
Experimental Biology, 41, 135- 154.
11. Greng, W., Corsman, P., Berry, C., Feng, Z and. Schafer, W.R. (2004)
Automatic tracking, feature extracyion and classification of C. elegans
phenotypes. IEEE Transaction on Biomedical Engineering, 51, 1811- 1820.
12. Hapca, M.S., Budha, P., Crawford, J.W and. Young, I.M. (2007) Movement of
the nematode, Phasmarhabditis hermaphrodita, in a structurally heterogeneous
environment. Nematology, 9, 731-738.
13. Hussey, R.S and. Barker, K.R. (1973) A comparison of methods of collecting
inocula of Meloidogyne spp. including a new technique. Plant Disease Reporter,
57, 1025-1028.
14. Imbriani, J.L and Mankau, R. (1977) Ultrastructure of the nematode pathogen
Bacillus penetrans. Journal of Invertebrate Pathology, 30, 337-347.
15. Mankau, R. (1975) Bacillus penetrans n. comb. causing a virulent disease of
plant-parasitic nematodes. Journal of Invertebrate Pathology, 26, 333-339.
16. Mankau, R. (1980) Biological control of Meloidogyne populations by Bacillus
penetrans in West Africa. Journal of Nematology, 12, 230.
569
17. Mankau, R and Imbriani, J.L. (1975) The life-cycle of an endoparasite in some
tylenchid nematodes. Nematologica, 21, 89-94.
18. Niebur, E and Erdos, P. (1991) Theory of the locomotion of nematodes.
Biophysics Journal, 60, 1132-1146.
19. Rao, M.S., Gowen, S.R., Pembroke, B and Reddy, P.P. (1997) Relationship of
Pasteuria penetrans spore encumberance on juveniles of Meloidogyne incognita
and their infection in adults. Nematologia Mediterranea, 25, 129-131.
20. Sayre, R.M and Wergin, W.P. (1977) Bacterial parasite of a plat nematode:
morphology and ultrastructure. Journal of Bacteriology, 129, 1091-1101.
21. Stirling, G.R. (1991) Biological control of plant-parasitic nematodes. Progress,
Problems and Prospects, Wallingford: CAB International.
22. Storer, T.I., Usinger, R.L., Stebbins, R.C and Nybakken, J.W. (1979) General
Zoology, 6th edition. New York: McGraw-Hill Book Company.
23. Thorne, G. (1940) Dubosqia penetrans n. sp. (Sporozoa, Microsporidia,
Nosematidae) a parasite of the nematode Pratylenchus pratensis (De Man)
Filipjev. Proceeding of the Helminthological Society of Washington, 7, 51-53.
24. Tytell, E. D. (2004) The hydrodynamics of eel swimming II. Effect of swimming
speed. Journal of Experimental Biology, 207, 3265-3279.
25. Wallace, H.R. (1958) Movement of eelworms. The influence of pore size and
moisture content of the soil on the migration of larvae of the beet eelworm,
Heterodera schachtii Schmidt. Annals of Applied Biology, 46, 74-85.
26. Wallace, H.R. (1959) The movement of eelworms in water films. Annals of
Applied Biology, 47, 366-370.
570