Comparison of active and passive sampling
techniques to collect Aedes larvae
T.F. Arevalo Valderrama
Universiteit van Amsterdam
Amsterdam 2016
A thesis submitted for the degree of Bachelor of Science (BSc) Supervisor: prof.dr.ir. P.F.M. Verdonschot _______________________
Abstract
Mosquitos cause a great nuisance for the population of Griendtsveen, Limburg. The village is located in an area that has undergone some hydrological changes, causing the conditions to be ideal for the proliferation of mosquitos. A previous study by Verdonschot et al. (2015) showed that 80% of the nuisance causing mosquitos consisted out of one species; Aedes cinereus. To combat this nuisance additional research is needed. Intervention at the larval stage of the mosquitos in preferred, as it is easier to combat mosquitos when they’re not fully developed and can’t disperse by flight yet. By doing so the amount of mosquitos that could potentially reach the village and cause nuisance is reduced. To do this the collection of larvae is essential. The current method of collection is dipping. However, this method is not capable of making accurate population estimates (Boyd 1930; Goodwin and Eyles 1942; Service 1971 mentioned in Silver, 2008). That is why this research tested different possibilities to collect larvae of A. cinereus with passive collection methods and whether these methods would produce more reliable population estimates than dipping (active method). This research showed that from all the passive methods tested the Floating trap appears the most promising and could offer the most accurate population estimates. This method could possibly replace the dipping method as leading method to make population estimations with.
Acknowledgements
First of all, I would like to express my gratitude to my supervisor prof.dr.ir. P.F.M.
Verdonschot for his help, guidance and for providing me with all the necessary facilities in order to carry out this research.
Besides my supervisor I would like to thank my daily supervisor; ir. T.B.M. Dekkers for her continuous support, guidance and insightful comments on a daily basis.
I would also like to thank dhr.dr. P. Roessingh, my coordinator, for his help and support in performing the statistical analysis for this research.
Table of contents
ABSTRACT ... II
ACKNOWLEDGEMENTS ... III
LIST OF ILLUSTRATIONS ... V
LISTS OF TABLES ... V
INTRODUCTION ... 1
MATERIAL & METHODS ... 2
LOCATION ... 2
COLLECTION METHODS ... 3
The Dipper – active method ... 3
The Vietrap – passive method ... 4
The Floating trap – passive method ... 5
The AFRIMS trap – passive method ... 5
STAGES OF RESEARCH ... 6
Phase 1 ... 6
Phase 2 ... 8
Phase 3 ... 10
DATA-‐‑ANALYSIS ... 12
Phase 2 ... 12
RESULTS ... 13
PHASE 1 ... 13
Results phase 1 ... 13
PHASE 2 ... 13
Missing larvae at recapture ... 13
Results phase 2 ... 14
PHASE 3 ... 16
Results phase 3 ... 16
DISCUSSION ... 17
PHASE 2 ... 17
PHASE 3 ... 19
CONCLUSION ... 20
REFERENCES ... 21
APPENDICES ... 22
APPENDIX A ~ FIELDWORK LOG GRIENDTSVEEN ... 22
APPENDIX B ~ FIELDWORK VALUES PHASE 2 ... 42
APPENDIX C ~ FIELDWORK VALUES PHASE 3 ... 43
List of Illustrations
Figure 1 Map of the designated research area. ... 3
Figure 2 The standard pint dipper ... 4
Figure 4 The Floating trap ... 5
Figure 5 The AFRIMS trap ... 6
Figure 6 The range of A. cinereus around the village Griendtsveen.. ... 7
Figure 7 The different test sites at which phase 2 was performed. ... 8
Figure 8 Containers used during the experiments ... 9
Figure 9 The validation locations ... 11
Figure 10 Assigned research sites ... 13
Figure 11 The population estimations per trap. ... 15
Figure 12 The relationship between caught larvae and population estimates.. ... 16
Lists of Tables
Table 1 The number and rates of recaptured ... 14Table 2 The number of larvae caught per collection method. ... 14
Table 3 Population estimations for each experiment site per trap. ... 15
Table 4 Amount of larvae caught per catchment method ... 17
Table 5 Population estimations for the Floating trap with the new formula. ... 18
Introduction
A great nuisance of biting insect, by mosquitoes in particular, is experienced in the small village of Griendtsveen, Limburg. Griendtsveen is located in an area also referred to as the Pelen, which in past times was used for peat extraction. Nowadays most of the surrounding area is converted into nature reserves: Maria Peel, Grauwveen and the Deurnse Peel (Figure 1). The conversion of this area into nature reserves caused hydrological changes, which on its turn is the probable cause of an increased presence of mosquito populations (Verdonschot, 2015). The number of mosquitoes present in and around Griendtsveen is found to be significantly higher than in other parts of The Netherlands.
A previous study by Verdonschot et al. (2015) showed that 80% of the nuisance causing mosquitoes consisted of the mosquito species Aedes cinereus. Mosquitoes appeared to come from a radius of 1.5-‐‑2 km around Griendtsveen and breed in places that become temporarily flooded in winter and spring. However much is unknown about the ecology of A. cinereus and its occurrence in the area. To better combat the nuisance situation additional research is needed.
To investigate the nuisance caused by A. cinereus it is essential to collect larvae as measures that reduce breeding sites are most effective. Some simple methods can be employed to collect larvae. This can be done by means of passive or active methods. Currently active dipping is the most commonly used method (Silver, 2008; Becker, 2010). In this method a soup ladles, a ‘dipper’, is used to scoop the larvae up from the water body in which they reside. There are a lot of different versions of dippers as the materials from which they consist usually depend on local availability. Unfortunately it is not always clearly recorded which type of dipper is used (Silver, 2008; Becker, 2010). These tools are inexpensive and simple to use. They are however also time consuming to use as the number of dips needed to make an estimate about population dynamics (e.g. number of developmental stages and population size) depends on the size of the surface of the water body in which the larvae reside (Papierok et al., 1975; Croset et al., 1976; Mogi, 1978 mentioned in Becker, 2010). For every m2 of surface area 1-‐‑2 dips should be taken. For larger pools or larger areas with many water bodies this becomes a time-‐‑consuming task. Furthermore, the dipping method only allows for a rough estimation of the aquatic mosquito population (Becker, 2010).
This research will look at the possibilities to collect larvae of A. cinereus with passive collection methods and look if these passive methods produce more reliable population estimations than the dipping method. If all or one of the passive traps offers more reliable population estimations then this could lead to more effective and less time consuming method of larvae collection.
To start the research the habitat of A. cinereus has to be located is found three kinds of traps will be tested on five locations: the Vietrap of Russell & Kay (1999), the Floating trap of Undeen and Becnel (1994) and the Armed Forces Research Institute of Medical Sciences (AFRIMS) trap of Harrison et al. (1982). Results from these traps will be compared to results from the dipping method. The previously named traps are described in greater detail in the
method section. When the most effective trap is determined this trap will be validated in the field.
The expectation is that all of the passive methods prove to be a better way to collect larvae and provide a more reliable way to produce population estimations. However, in the case of this research area the Floating trap is expected to be most effective as most of the pools in which the larvae of A. cinereus reside consist out of shallow temporarily waters. As the Floating trap is the shortest of the three traps this trap is expected to be most suitable for the research area.
Material & Methods
The research was carried out in three different phases. In the first phase the habitat of A. cinereus was located. In the second phase the passive and active collection methods of interest were tested in order to make a comparison between the two. And in the third phase the selected passive trap were validated in the field. Before elaborating on the different phases the research location and the various ways of passive and active methods tested will be discussed.
Location
The research location on the comparison of active and passive sampling techniques to collect Aedes larvae is performed in the nature reserves around Griendtsveen. These nature
reservates are the Deurnse Peel and the Maria Peel. The Maria Peel consists out of tree different complexes, namely the Horster Driehoek, Mariaveen and Driehonderd Bunder. The research was carried out during a period of 8 weeks starting from April the 18th until June the 10th. The area selected to conduct the research is particularly suited as the number of mosquitoes is currently significantly higher than in other parts of the Netherlands providing a good source of mosquito larvae to perform experiments. The research is in particular focused on techniques to sample Aedes larvae, and a big proportion of the mosquitos in the area is composed out of A. cinereus, making the area perfect for this research (Verdonschot et al., 2015).
The increased amount of mosquitos in the surrounding of Griendtsveen is thought to be due to the hydrological changes in the area (Verdonschot et al., 2015). These hydrological changes are the result of directives that promote the development of peat mosses and bog
regeneration in order to restore the natural environment of the area. This has been put into action as a result of the bog recovery project LIFE+ by Staatsbosbeheer (Staatsbosbeheer, 2015).
Marsh mosquitos like A. cinereus tend to reproduce in shallow temporary waters that occur for a longer period of time (several weeks to 6 months), are weakly acidic and rich in organic matter. The temporary nature of these waters causes for the low occurrence of predators. The hydrological changes in the area around Griendtsveen cause a lot of these temporary waters to occur which makes the area perfect for A. cinereus to proliferate.
Mariapeel and Deurnsche Peel
Figure 1 Map of the designated research area. The area consists out of the Pelen nature reserves also referred to as the Deurnsche Peel and the Mariapeel. The Mariapeel consists out of tree different complexes, namely the Horster Driehoek, Mariaveen and Driehonderd Bunder. Source: Basisregistratie Topografie, 2015.
Collection methods
The Dipper – active method
The dipper used for this research is the most standardized version of the dipper, ‘the
standard pint dipper’, see figure 2. It consists out of a white plastic container with a capacity of 350 ml and a diameter of 11 cm that is attached to an extendable aluminium pole (Dixon & Brust, 1972; Lemenager et al., 1986 mentioned in Becker, 2010). The pole helps to reach the water bodies from a distance in order to not disturb the larvae, which would otherwise cause them to dive out of reach.
The Dipper
Figure 2 The standard pint dipper is showed. Source: Becker, 2010.
Trap operation
The dipping method is conducted by pressing the dipper into the water until the edge of the dipper is just below the water surface. By doing so the water, including mosquito larvae, flows into the dipper. This has to be done from such a distance that the larvae don’t get alerted by the shadow or ground vibrations caused by the investigators and at such a rate that the larvae are not able to dive away.
The Vietrap – passive method
The Vietrap consist out of a funnel that is 180 mm deep and has a diameter of 185 mm (Russell & Kay, 1999). The funnel is attached to a plastic screw cap container (figure 3). To assemble the Vietrap a circular hole is cut in the screw cap and the narrower side of the funnel inserted. A
counterbalance was connected to the broad side of the funnel before fixation to the cap. As
counterbalance a piece of chain was used.
Trap operation
To use the trap container is filled with water for two thirds before lowering it onto the water surface with the funnel upward (Russell & Kay, 1999). By gently lowering it further into the water the trap inverts in such a way that air stays trapped in the
trap stays Floating. Mosquito larvae enter the trap when moving towards the surface true the wide side of the funnel and stay trapped within the container.
On retrieval the trap is gently lifted in such a way that the samples stay in the container.
The Floating trap – passive method
The Floating trap is an easily constructed larval trap that consists out of a transparent plastic cup with a convex bottom and a content of 100 mL (Undeen & Becnel, 1994; Silver, 2008). In the centre of the cup a 6-‐‑mm hole is cut out. Three pieces of cork are attached to the outside of the cup so that is stays Floating. The trap stays Floating at the water surface in such a way that the bottom is situated about 5mm below the surface (figure 4).
The Floating trap
Figure 4 The Floating trap. Source: Undeen & Becnel, 1994; Silver, 2008.
The AFRIMS trap – passive method
The AFRIMS trap consists out of a 13-‐‑cm diameter 6-‐‑cm deep kitchenware container with a removable lid (figure 5). A light coloured funnel is attached to the bottom by inserting the small narrow end in a 1cm diameter hole cut in the middle of the bottom of the container. The Funnel has a 10,5 cm diameter and is 8,5 cm deep. The container is wrapped by a 44 cm piece of tube that is joined together by a wooden plug in order to keep it Floating.
Two holes are cut in the snap on lid opposite from each other and rubber stoppers with screw hooks are inserted. Nylon fishing thread is attached to the screw hooks and threated true the funnel where it is attached to a similar rubber stopper.
The AFRIMS trap
Figure 5 The AFRIMS trap by Harrison. Source: Harrison et al., 1982; Silver, 2008.
Trap operation
To operate the AFRIMS trap all rubber stoppers are pulled free and the trap is lowered into the water (Harrison et al., 1982 mentioned in Silver, 2008). When the water reaches the upside of the Floating tube the upper plugs are inserted in order to trap air and keep the trap
floating. To remove the trap the upper plugs are pulled out in such a way that the bottom plug shuts the funnel. The samples can be extracted by removing the lid of the container.
Stages of research
Phase 1
Location
Previous research showed that A. cinereus emerges in an area with a radius of 1,5-‐‑2 km around Griendtsveen (Verdonschot et al., 2015). As the research focuses on the possibilities to catch larvae of A. cinereus this phase will focus on the area 1,5-‐‑2 km surrounding
Griendtsveen. Consequently phase 1 was mainly carried out in the Deurnsche Peel, the Horster Driehoek and the Driehonderd Bunder, as these are the regions of interest within the mentioned action radius of A. cinereus. The regions and the action radius are visible in figure 6.
Action radius of A. cinereus
Figure 6 The range of A. cinereus around the village Griendtsveen. Source: Verdonschot et al., 2015.
Procedure
In phase 1 of the research fieldwork was performed to establish the habitat of A. cinereus. Previous research showed that A. cinereus comes from a zone 1,5-‐‑2 km around Griendtsveen (Verdonschot et al., 2015). During the first few weeks this zone was surveyed for the
occurrence and density of A. cinereus larvae. This first phase was done in two teams of two students. Water bodies in the area were checked for the occurrence and density of A. cinereus by dipping, as this is currently the most commonly used method (Silver, 2008; Becker, 2010). Larvae caught on location were collected in 50 mL plastic bottles containing ethanol with a percentage of 75%. These bottles were labelled and the tested water bodies were marked by GPS so that the corresponding location could afterwards be retrieved. The marked bottles were sent to Alterra in Wageningen for the identification of the mosquito larvae. When sufficient locations were found containing a high proportion of A. cinereus the research moved on to phase 2.
Phase 2
Location
The test locations for phase 2 are all located in the Mariapeel and Deurnsche Peel and were selected on the basis of the results of phase 1. The 5 test locations chosen for phase 2 were selected on basis of the abundance of A. cinereus (Dekkers, personal communication, June 2016). When a large portion of the mosquito larvae found at one of the locations researched in phase 1 consisted out of larvae of A. cinereus the site was selected as test site for phase 2. The sites used for research during phase 2 are shown in figure 7.
Phase 2 test locations
Figure 7 The different test sites at which phase 2 was performed.
Procedure
When the habitat of A. cinereus was determined 5 research locations were chosen to perform phase 2 in which three different passive traps; the Vietrap, the Floating trap and AFRIMS were tested and compared with the dipping method (see figure 7). At each test location 4 containers were placed in order to test the different methods, one container per method. These containers were 35 cm deep and 52,5 wide and could contain 75 L (see figure 8). For each individual experiment 200 larvae were used. So, 800 larvae were used for each test series per location. These larvae were captured at the beginning of phase 2 at the first test location and stored in 500 mL plastic bottles. For storage the bottles were filled with a couple of leaves to provide substrate and were filled until the 400 mL mark with filtered water from the location of capture in such a way that a space remained for oxygen.
Considering the length of the fieldwork and the time needed to capture the amount of larvae needed for the experiments the decision was made to perform pseudo replication. Only the first 800 larvae were freshly captured. After each following experiment the larvae were recaptured as much as possible. The amount of missing larvae was freshly captured at one of the test locations. More on the recapture of the larvae follows in a later section of this
chapter.
Test container
Figure 8 The container in the picture is one of the containers used during the experiments.
For the experiments the test containers were placed in as much shade as possible and partly placed into the water to prevent rapid warming and to mimic the natural habitat of the larvae at the test location as much as possible (Verdonschot, personal communication, April 2016). Next the four containers were each filled with water from the location of testing until the 25cm point was reached. The water added was filtered with a sieve before adding to make sure no mosquito larvae were added. After this approx. 2cm of substrate, 8 times the content of a standard pint dipper (8 scoops) from the nearest pool habitat were added to the containers in order to mimic the natural environment of the larvae at the test location. The composition of the substrate added was estimated and recorded in the form of proportional scale of measurement. The scale ran from 0 to 5. First, a difference was made between fine particular organic matter (Fpom), which was non-‐‑identifiable organic matter and coarse particular organic matter (Cpom), which was identifiable substrate. Secondly the
composition of the Cpom was estimated on an equal scale. So, for example the Cpom was estimated to consist out of 20% oak leaf and 60% birch leaf then oak leaf got 1 point and birch leaf got 4 points.
After the adding of the substrate the container was given a resting period of minimal 12 hours (h) in order to let the substrate settle down. After this resting period the containers were checked using a dipper in order to assure that no larvae entered the container due to the adding of the substrate. If larvae were found the dipping continued until no larvae were found during a series of five consecutive dips.
When the containers were checked 200 larvae were added per container. Next the variables O2 (mg/L), acidity (pH), conductivity (µμS/cm) and temperature (C°) of the four containers and the water body from which the water was extracted to fill the containers were measured using a portable Hach HQ40d two-‐‑channel multi meter. When the values of the variables were noted the GPS coordinates and a picture of the test location were taken. Next the
passive traps were added. The passive traps are usually used for an overnight period of 12 or 24 h (Russell & Kay, 1999; Silver, 2008). Due to the retrieval opportunities in this experiment there was chosen to test the passive traps for an overnight period of 24h. After the passing of this period the passive traps were removed and the 4th test container was sampled by
dipping. As mentioned before 1-‐‑2 dips should be taken for every m2 of surface area. Keeping this in mind dipping once is enough to sample the test container. The numbers of larvae collected per method were recorded and the previously mentioned variables were measured again in the four containers and the water body.
The last part of the procedure consisted out of the recapture of the larvae in the containers as much as possible. This was done by removing and filtering the top layer of the water with the aid of a bucket and a sieve until the water level reached the top of the substrate. When the top of the substrate was reached dipping was applied to extract as many of the remaining larvae as possible. The larvae collected in the sieve and the dipper were put into a white coloured tub from which they were pipetted into 500 mL bottles. And again stored as mentioned at the start of this chapter by adding 400 mL filtered water and some leaves to provide substrate. As it was impossible to recapture exactly 800 larvae the missing number of larvae was freshly caught and evenly distributed over the test containers at the next location. This procedure was replicated at each test location.
Phase 3
Location
The test sites at which phase 3 was conducted where all located in the research area of interest within the Deurnsche Peel, the Horster Driehoek and/or the Driehonderd Bunder. The sites for this phase were also selected on the basis of the results of phase 1 and on information provided by Alterra (Dekkers, personal communication, June 2016). The sites used for research during phase 3 are shown in figure 9.
Phase 3 validation locations
Figure 9 The different locations at which the validation of the passive methods took place.
Procedure
In this phase the traps were validated in the field in order to see if the results obtained under predetermined conditions in phase 2 were comparable to results obtained under field
conditions. To validate the traps 5 different well-‐‑populated test sites were chosen (figure 9). These sites are described in the previous section. When the pools at the sites were selected the variables O2 (mg/L), acidity (pH), conductivity (µμS/cm) and temperature (C°) of the water body was recorded together with the GPS coordinates. When this was done the passive traps were placed and a picture was taken from the test site. After an overnight period of 24h the passive traps were recollected and emptied in a white coloured tub so that the amount of larvae could easily be counted. When the amount of larvae was noted the variables were measured again and afterwards the collected larvae were thrown back into the pool from which they were retrieved. The test site was given a resting period of 24h to recreate the initial conditions and to allow for equal catchment opportunities as much as possible. After the resting period dipping was carried out at exact the same location. The pools selected were in all cases of such a size that dipping once was enough to achieve viable results. The number of caught larvae was recorded and collected in labelled 50 mL plastic bottles containing ethanol with a percentage of 75%. These bottles were also sent to Alterra in Wageningen for the identification of the mosquito larvae. Then the previously mentioned variables of the tested water body were measured and the composition of the substrate was estimated. For the estimation of the substrate the difference between Fpom and Cpom and the composition of Cpom were noted.
This procedure was repeated at all 5 locations in order to make a comparison between the passive and active methods.
Data-‐‑analysis
In this chapter the methods of data-‐‑analysis are discussed. All the data acquired were analysed using the statistical program R Studio, software version 0.98.1091. Phase 1 was left out of consideration as this phase was about the selection of suitable test sites for the other phases, and thus no analysis were used. Phase 3 was not analysed either as the collected results were not sufficient to perform statistical analyses. Because of this only the data-‐‑ analysis of phase 2 will be discussed.
Phase 2
To start the data-‐‑analysis the amounts of larvae caught for the Vietrap, Floating trap and AFRIMS trap were compared with the amounts caught by the Dipping method. To do this the data had to be plotted in a histogram to see in which way the data were distributed. As the data was Poisson distributed the choice fell on a Generalized Linear Mixed-‐‑Effects Model (GLMM). This test was performed with a significance boundary of p < 0,05.
Then the existing dataset was expanded with an additional variable namely the population estimation, which was made with the recorded amounts of caught larvae. These estimates were made for each test that was carried out. The estimates were calculated according the following formula:
P = (v/c)*n
Absolute population estimates (P) of larvae could be made by using simple proportions, such as surface area of water in the larval habitat (v), surface area of the trap used in the method of capture (c) and the number of larvae caught by this method (n) (Verdonschot, personal communication, June 2016).
Then the population estimations were compared between the active and passive methods. As the data was not normally distributed a log transformation was carried out. To
compensate for zero values 1 was added to the log transformation. As the resulting data was normally distributed a regression analysis was performed also with a significance boundary of p < 0,05.
To get an overview of the population estimations per method the estimations were plotted in a boxplot. To finish the data-‐‑analysis the amount of caught larvae was plotted against the population estimation for each trap within a common plot.
Results
In this chapter the results of the fieldwork and the data-‐‑analysis are presented for each of the consecutive phases. The results found during the fieldwork are presented in a more detailed version in appendix A ~ The fieldwork log.
Phase 1
Results phase 1
Phase 1 resulted in the designation and approval of research sites that had a mosquito population with a sufficient large proportion of A. cinereus for carrying out the subsequent phases (Dekkers, personal communication, June 2016) (figure 10).
Assigned research sites
Figure 10 The assigned research sites are indicated by red dots on the map. These locations accommodate a sufficiently large population of A. cinereus for performing phase 2 and 3.
Phase 2
Missing larvae at recapture
Between different test sites it proved to be impossible to recapture all the 800 larvae
previously added to the containers. These missing larvae were unable to be recaptured at the test sites during phase 2. This was mainly due to larvae that hid in the substrate, as it was too time consuming to check the substrate these larvae were left out and replaced. Another reason for larvae gone missing is because some of the larvae emerged and fledged from the test containers.
The number of missing larvae and the failure rate of recapture are visible in table 3.
The numbers that are visible under ‘Missing at recapture’ had to be freshly captured before experiments at the following test site could start. For test site 2 for example 130 fresh larvae had to be captured to replenish the number of test subjects. Location 5 is an exception as this was the last experiment of phase 2. Only a subset of the 800 initial larvae was recaptured for identification at Alterra. The high failure rate at location 2 can be explained by uncooled storage conditions over the weekend. The elevated water temperature had as a consequence that many larvae fully developed into mosquitoes.
Table 1 The number and rates of larvae that were unable to be recaptured after the experiments at each location.
Test site
Absolute population
number
Dropout after test
Failure rate
1
800
130
16,25 %
2
800
250
31,25 %
3
800
146
18,25 %
4
800
200
25 %
5
800
500
62,5 %
The subset that was recaptured at location 5 consisted out of 292 larvae. From these 292 larvae 289 larvae were A. cinereus, 1 was Ochlerotatus sp and 2 were Chaoboridae. This
percentage is sufficiently high to be able to say that the traps have been tested on A. cinereus.
Results phase 2
In phase 2 the different collection methods were tested. These methods were tested under known conditions. The larvae caught per trap during phase 2 are shown in table 2. These numbers were later on used to make population estimations for the larvae in the test
containers. The population estimations are visible in table 3. For all information gathered on the traps see appendix B ~ Fieldwork values for an extensive table with all the values
gathered in this phase.
Larvae per method
Table 2 The number of larvae caught per collection method.
Test Vie trap AFRIMS trap Floating trap Dipping
1 5 0 5 4
2 3 9 4 1
3 2 4 8 3
4 6 2 4 2
By making a histogram in R studio of the number of larvae caught it appeared that the data show a Poisson distribution. By fitting the data in a generalized linear mixed-‐‑effects model with a poisson distribution the Floating trap was the only trap that deviates significantly from the dipping method (p= 0.00661). This also means that this trap captures most larvae under the scientifically accepted test methods for this trap.
Population estimations per method
Table 3 This table shows the population estimations for each experiment site per trap.
Test Vie trap AFRIMS trap Floating trap Dipping
1 52 0
551
91
2 31 248 441 23 3 21 110 882 68 4 62 55 441 46 5 10 193 1433 23 Mean 35,2 121,2 749,6 50,4The Floating trap also has a significantly different population estimation compared to that of the dipping method (p=0.00327). The other traps showed comparable results to the dipping method, the Vietrap has a p-‐‑value of 0.64596 and the AFRIMS trap has a p-‐‑value of 0.88399. These traps are therefore not significantly different from the dipping method. The difference in population estimations between the traps is clearly visible (figure 11). The boxplot shows that the Floating trap also has the largest variance within its population estimation.
Population estimations per trap
Figure 11 The population estimations per trap.
To give an overview of the larvae collected and the resulting population estimation the population estimation was plotted against the amount of larvae caught (figure 12). As is Actual Population
visible the amount of larvae caught for the Floating trap results in significant higher
population estimates. Making the Floating trap stand out and as calculated before significant different from the other methods. Possible causes for these elevated population estimates will be discussed in the discussion.
Larvae against population estimate
Figure 12 The relationship between the caught larvae and the resulting population estimates. The red line belongs to the Floating trap, the orange line to the AFRIMS trap, the green line to the Vietrap and the black line to the Dipping method.
Phase 3
Results phase 3
In this section the results of phase 3 are described. The larvae caught by the passive and active method show that the Floating trap is the only trap tested for validation during this phase (table 4). The Vietrap and the AFRIMS trap were left out of consideration; the reasons for this choice will be discussed in the discussion. As is visible in the table is was not
possible to catch any larvae at most of the test sites. Only 2 out of 5 sites had complete
results, as is visible in the table. Because this number of usable results is too low for statistical tests, it was not possible to perform a validation of the trap. Due to the failure of the
validation there are no further results for this phase. The causes of failure will be pointed out in the discussion. A more extensive table with all variables measured in phase 3 is available in appendix C ~ Fieldwork values Phase 3.
Larvae per method
Table 4 This table contains the amount of larvae caught per catchment method during the third phase.
Test
Floating trap
Dipping
1
0
0
2
2
0
3
0
0
4
3
3
5
3
4
Discussion
This study looked at the possibilities to collect larvae of A. cinereus with passive collection methods and if these passive methods would produce more reliable population estimations than the dipping method. The most important research results are that the Vietrap and the AFRIMS trap caught similar amounts of larvae as the dipping method but that the Floating trap caught significantly more larvae. For the population estimations part of the hypothesis the Vietrap, AFRIMS trap achieved similar results to the dipping method but the Floating trap was significantly different. This confirms the hypothesis, as all methods are able to catch larvae and are just as successful or even better in doing so compared to the dipping method. However, for the population estimate applies that only the Floating trap is significantly different from the dipping method. The Vietrap and the ARIMS trap are not significantly different from the dipping method in predicting a population estimate. Therefore, this research has so far only been able to confirm the first part of the hypothesis namely that the passive traps are able to catch larvae as successful as the dipping method in case of the Vietrap and the AFRIMS trap or are even more successful in doing so in case of the Floating trap.
For the Floating trap, which gives an overestimation of the population it does not necessarily confirm that the trap offers a better or worse way than the dipping method in making
population estimations. The underlying reasons to explain this will be discussed during the discussion part of phase 2.
Phase 2
In phase 2 a population estimate was calculated. There are multiple ways described to do this in Silver (2008) but most of them involve data that is acquired by stirring the water before dipping. The formula that belongs to a dipping method that most resembles the dipping method used in this research is the formula offered by Mori (1989), this formula is mentioned below (Mori, 1989 mentioned in Silver, 2008).
P = (vw/vt)*n
The absolute population estimates (P) of larvae can be made by using simple proportions, such as the volume of the tested water body in the larval habitat (vw), volume of the trap used in the method of capture (vt) and the number of larvae caught by this method (n).
This method however led to severe overestimations by all catchment methods tested. Therefore, it was decided to adjust this method by using the surface area of the tested water body and trap instead of the volume (Verdonschot, personal communication, June 2016). This formula was used to calculate the population estimates for all catchment methods used as no formulas were found that were specified to calculate the population estimates for passive methods.
The adjusted method led to results that were closer to the actual population size for all methods. The only trap that remained having different results was the Floating trap. This is clearly visible in figure 11 and 12 in the results section. The reason behind this is that the formula by Mori (1989) is specifically intended for dipping. The Vietrap and the AFRIMS trap have comparable surface areas but the surface area of the Floating trap is considerably smaller than that of the dipper. This is why the formula leads to a considerable
overestimation of the population if used for the Floating trap. This is why the formula should be calibrated for the Floating trap. To find a correct formula for usage by the Floating trap a larger dataset would be appropriate, thus additional fieldwork is required. If a
suitable formula is constructed field validation is needed to confirm this. A way proposed to calibrate the formula with the research done so far is by dividing the true population (200) by the mean estimated population (750) of the Floating trap and adding this to the current formula. This results in the following formula:
P = (4/15)*(v/c)*n
With the absolute population estimates (P), the surface area of water in the larval habitat (v), the surface area of the trap used in the method of capture (c) and the number of larvae caught by this method (n). The new population for the Floating trap acquired with the new formula are shown in table 5.
Population estimates for the Floating trap
Table 5 This table shows the population estimations for the Floating trap that were acquired with the proposed formula.
Test Floating trap
1
147
2118
3235
4118
5382
Mean200
It could be that if formulas would be calibrated for usage with the Vietrap and the AFRIMS trap this would also lead to more accurate population estimations. But since these traps didn’t catch a significant different number of larvae compared to the dipping method this is less likely. As the Floating trap caught most larvae of all methods under the scientifically accepted method of catchment it could be assumed that this method could provide the most accurate way to make population estimations. Simply due to the reason that larger sample sizes offer a more reliable reflection of the population mean. Another argument why passive methods offer a better way to make population estimation is that they are tested over a period of 24h, whereas the dipping method is only a representative of one point in time.
Even if this paper could not completely prove that passive methods are a better way of providing accurate population estimations the obtained results point in that way. Furthermore, the dipping method has proven to be a poor way to make population
estimations. Several researchers have tried and failed to make good population estimations by relating the numbers of larvae in a dipper of known volume to the numbers in a breeding place of known size (Boyd 1930; Goodwin and Eyles 1942; Service 1971 mentioned in Silver, 2008). Though it is theoretically possible dipping is unlikely to be a good sampling method for obtaining accurate population estimates (Silver, 2008). Therefore its safe to assume that passive methods, and in particular the Floating trap could provide better population estimations and that it is at least worth to further investigate these methods.
Phase 3
Phase 3 was meant to validate the passive methods in the field in order to see if the results obtained under predetermined conditions in phase 2 would resemble the results that would be obtained under field conditions in phase 3 and also to make a field comparison between the passive and active methods. At first all the methods would be subjected to validation in this phase, but during the start of the phase it became clear that from the passive methods only the Floating was suitable for field validation. This was due to the properties of the research location. Most of the pools that were suited and selected as test location were too shallow for the AFRIMS and Vietrap. This is why only the Floating trap has been subjected to all tests.
Unfortunately phase 3 did not work out as planned due to unexpected events. This led to insufficient data being collected, making it not possible to carry out statistical analyses. The failure of phase 3 had two likely causes. The first cause was that the validation period took place in a period of very bad weather. Heavy rainfall caused the small temporarily pools containing mosquito larvae to dilute and to become interconnected. This made it a lot harder to catch anything due to there being more water relatively to mosquito larvae. The second reason was that the period of validation took place in a period with less
mosquito larvae being present. This is probably due to the validation period taking place in between a period in which the larvae previously present completely developed into
mosquitoes and a period in which the second batch of mosquito larvae hatched from their eggs.
Due to a lower amount of mosquito larvae being present and the larvae being present being spread out over a larger submerged area it became complicated to catch anything.
And without sufficient larvae being caught it became impossible to make a comparison between the field results of the passive and active methods and between the results of phase 2 and 3.
Conclusion
The main conclusions that could be drawn from this research are that the Floating trap caught significantly most mosquito larvae. By doing so this trap is assumed to be the best method to make population estimations. However the formula used to make population estimations should be calibrated for this trap. Furthermore, additional research is needed to establish whether this trap performs the same in the field and to calibrate the in the
discussion offered formula more accurately. No significant difference has been found between the Vietrap, the ARIMS trap and the dipping method during this study.
If an accurate formula is formulated and calibrated for the Floating trap in order to make population estimations and the trap would be validated in the field it could offer a new and revolutionary way to make population estimations. As the dipping method is proven not to be successful in making accurate population estimations the Floating trap could in time replace the dipping method as leading method for making population estimations. The Floating trap could offer new insights on the population of A. cinereus and by doing so help to combat the nuisance caused for the population of Griendtsveen. If the accuracy of the Floating trap is further confirmed it could become an integral part in research on Aedes mosquitos worldwide.
The current research has shown that passive methods, in this case the Floating trap in theory offers a more accurate way of making population estimates for the Aedes mosquito. However as field validation was not successful additional research is needed. But so far the Floating trap looks like a promising method to replace the dipping method in making population estimations.
References
Becker, N., Petric, D., Dahl, M.Z.C., Boase, C., Madon, M., & Kaiser, A. (2008). Mosquitoes and Their Control (2nd ed.). New York: Springer, pp. 46-‐‑47.
Russell, B.M., & Kay, B.H. (1999). Calibrated Funnel Trap for Quantifying Mosquito (Diptera: Culicidae) Abundance in Wells. Entomological Society of America, 36(6), 851-‐‑855.
Silver, J.B. (2008). Mosquito Ecology; Field Sampling Methods (3th ed.). New York: Springer, pp. 154-‐‑221.
Staatsbosbeheer (2015). Mariapeel. Retrieved from:
http://www.staatsbosbeheer.nl/over-‐‑staatsbosbeheer/dossiers/life-‐‑nature/life-‐‑ peelvenen/mariapeel, consulted on 16 June 2016.
Appendices
Appendix A ~ Fieldwork Log Griendtsveen
Thomas Arevalo
General information
Methods Passive methods: Vie=VietrapAFRIMS= Armed Forces Research Institute of Medical Sciences trap Float= Floating trap
Active method:
Dip= Standard pint dipper
Location:
Loc= Location (environmental values of location)
Test Containers
On each test site there are 4 containers, one for each method.
The containers are filled with 25cm water from the location each. The water added is filtered with a sieve before adding to make sure no mosquito larvae are added.
Containers are filled with approx. 2cm of substrate (8 times the content of a standard pint dipper (8 scoops))
This will be left for a period of minimal 12h in order to let the substrate settle down. Before the methods are tested the containers are checked using a standard pint dipper in order to make sure no larvae entered the container.
After the minimum of 12h and the checking of the containers 200 larvae are added per container. Then the variables O2 (mg/L), acidity (pH), conductivity (µμS/cm) and temperature (C°) are measured. After this the passive traps are added. These traps will stay in the test containers for an overnight period of 24h before removing the dipping method is tested.
Test subjects
For each test 200 larvae are used, as mentioned before. This means that for every series 800 larvae are used. These larvae are recaptured after the test is finished, so right after the removing of the passive traps. This is done by filtering the top layer of the water until the level that substrate is reached. When the top of the substrate is reached dipping is applied to extract as many larvae as possible that are left. Collected larvae from the filtered water are put into a white coloured tub from which they are pipetted into 500 mL bottles. The larvae are counted while being pipetted into the bottles. For storage the bottles are filled with a couple of leaves to provide substrate and are only filled until the 400 mL mark so that space
develop into mosquitos and fledge from the test containers. The missing number of larvae is freshly caught and evenly distributed over the test containers at the next location.
Substrate
Fpom= Fine particular organic matter, non identifiable organic matter Cpom= Coarse particular organic matter,
Substrate is estimated on the basis of a scale of 5 For each location a scaling is made for: • Scale 1 = Between Fpom and Cpom
• Scale 2 = Cpom (composition of identifiable substrate)
Location
Location is noted with number of location followed by the number of repetition on that location. Number of repetition is a 1 or a 2, 1 is for values before testing of methods and 2 is for values after testing of methods. This is necessary in order to make a difference between the measurements on the same location as comparisons will be made between and within locations.
For example:
Location 2.1 is location 2 with values before the testing of a method.
Each location is photographed after the adding of the passive traps.
Locations
Location 1.1 10-‐‑05-‐‑2016 Coordinates: 51°25’29.583”N 5°55’05.620”E Substrate: Scale 1: Fpom=0 Cpom=5 Scale 2: Cpom: Berkenblad = 4 Pijpenstrootje = 1Values at location 1.1
Table 1. This table shows the values measurable values at location 1.1.
Trap O2 Ph Con temp1 temp2 temp3
Vie 2,67 3,83 127,3 18 16,9 17,3 AFRIMS 2,01 3,81 125,9 17,5 17,1 17,1 Float 2,85 3,75 128,5 17,3 16,7 17,3 Dip 1,2 3,75 127,4 17,7 16,8 17,8 Loc 1-‐1 2,86 3,78 125,8 17,6 16,8 16,7
Mosquito larvae were added at 13:00 Traps were placed at 14:06
Location 1
Figure 13. This picture shows the setup at location 1 while the traps are added.
Location 1.2
11-‐‑05-‐‑2016
Values at location 1.2
Table 2. This table shows the values measurable values at location 1.2.
Trap O2 pH Con temp1 temp2 temp3 Larven
Vie 2,47 3,9 122,3 18,6 16,9 17,9 5 AFRIMS 2,08 3,71 125,3 17,6 16,6 18,1 0 Float 3,41 3,74 127 17,7 16 17,3 5 Dip 1,01 3,72 125,5 18,4 17,4 18,8 4 Loc 1-‐2 2,31 3,75 118,9 18,1 17 17,4
The traps were extracted at 13:21
From the 800 larvae in the container 130 went missing after recapture. 130 new larvae were caught at location 2.
Location 2
Figure 14. This picture shows the setup at location 2 after the traps are extracted.
Location 2.1 12-‐‑05-‐‑2016 Coordinates: 51°26’16.354”N 5°55’58.030”E Substrate: Scale 1: Fpom=3 Cpom=2 Scale 2: Cpom: Berkenblad = 2 Pijpenstrootje = 1 Eikenblad = 1 Russen = 1
Values at location 2.1
Table 3. This table shows the values measurable values at location 2.1.
Trap O2 Ph Con temp1 temp2 temp3
Vie 3,08 4,22 94,3 17,1 16,5 16,8 AFRIMS 3,42 4,31 92,3 16,8 16,1 16,4 Float 3,04 4,31 93,2 16,8 16,3 16,7 Dip 3,23 4,52 88,1 17,2 16,1 16,3 Loc 2-‐1
Mosquito larvae were added at 09:50 Traps were placed at 10:45
The measurements of the oxygen level, acidity, conductivity and temperature have been forgotten at this location.
Location 3
12-‐‑05-‐‑2016
After the finishing the work for location 2.1, location 3 was prepared and set up ready to use for after the weekend. Water and substrate were added. More information on this location is found in the section about location 3.1. The containers were covered with plastic so that no new mosquito eggs could be laid over the weekend.
Preparations at location 3.1
Figure 3. This picture shows the setup at location 3 before any methods are tested. The test containers are covered with plastic in order to keep them ready to use for after the weekend.
Location 2.2
13-‐‑05-‐‑2016
Values at location 2.2
Table 4. This table shows the values measurable values at location 2.2.
Trap O2 pH Con temp1 temp2 temp3 Larven
Vie 3,25 4,53 88,1 16,4 15,3 15,4 3 AFRIMS 3,24 4,34 93,5 15,7 15,4 15,6 9 Float 3,69 4,3 93,1 15,6 15,3 15,5 4 Dip 3,29 4,36 95,3 16,7 15,8 15,9 1 Loc 2-‐2 1,19 4,32 101,8 15,7 15,1 15,1