Less invasive blood sampling in the animal laboratory: clinical chemistry and haematology of blood obtained by the Triatominae bug Dipetalogaster maximus

Abstract

Dipetalogaster maximus (Dipmax), a blood-sucking bug belonging to the family Reduviidae, has been used to obtain blood samples, for example for clinical chemistry and haematology, in a variety of zoo animals and wildlife. Using this bug allows stress-free blood sampling as the bug is able to draw blood without the mammal noticing the bug. In laboratory animal science, the need for blood samples from unstressed animals may arise, especially in animal behaviour research. The use of Dipmax bugs may prove a valuable tool for this purpose. To validate the method, we compared an array of standard blood parameters sampled from New Zealand White rabbits, sampled either by the use of bugs or by the conventional method; puncture of vena auricularis caudalis. The overall hypothesis was that there was no significant difference in clinical chemistry and haematological parameters between the bug method and the conventional method. A total of 17 clinical parameters as well as 12 haematological parameters were measured and compared in New Zealand White rabbits. The results showed that for 13 of these 29 analysed parameters, the bug method and the conventional method did not give significantly different results, and the obtained results were thus directly comparable. For the remaining parameters the obtained results were significantly different. However, all parameters were measurable in the bug samples. The influences of the bug metabolism on these parameters are discussed.

Keywords

Blood sampling clinical chemistry haematology

In laboratory facilities and clinical practice, blood sampling for clinical chemistry and haematology most often involves handling and restraining of the animals. This may impose stress on the animals and hence increase the risk of altering stress-sensitive parameters like glucose, creatine kinase, neutrophil counts and lymphocyte counts. Both in animal experimentation and in the clinic, this effect on the parameters may affect the conclusions drawn on the basis of such measurements.

Von Helversen and Reyer¹ described a method for obtaining blood from animals by the use of Triatominae bugs. Triatominae bugs are blood-sucking insects belonging to the order Hemiptera, family Reduviidae and include among others the species Dipetalogaster maximus, Triatominae infestans and Rhodnius prolixus. They are commonly referred to as kissing bugs, as they have a preference for the region around the mouth when stinging humans. The use of bugs presents an opportunity to draw blood samples without stressing the animals and without risking haematomas and extensive damage to the vessel.² The use of bugs has hereafter been validated for use in doubly labelled water experiments,³ determination of progesterone, testosterone and hydrocortisone concentrations in rabbits,² serological studies detecting rabbit haemorrhagic disease virus antibodies,⁴ determination of corticosterone in terns,⁵ and measurement of a pregnancy hormone in the Iberian lynx.⁶ Validating this stress-free method of blood sampling would allow implementation in both animal experimentation and in the clinic. If stress could be reduced during blood sampling, it would result in improved animal welfare and may also lead to less variation and hence use of fewer animals. Moreover, blood sampling can be done when animals are resting which otherwise cannot be done without implanting catheters. In the clinic the bugs may be used for aggressive animals and most likely they would also provide a way of obtaining blood samples from patients with very low blood pressure which otherwise would be very difficult.

Animals exposed to D. maximus and other Reduviid bugs do not seem to react to the bite and this is probably due to the size of the stinging apparatus of the bugs. The penetrating tip of the proboscis in the D. maximus has a diameter of 20 µm⁷ (0.02 mm), which is far smaller than, for example, a 26 G needle with an outer diameter of 0.46 mm. Moreover, the saliva injected into the wound has a local anaesthetic effect.^2,8,9 Triatominae bugs have five nymph stages (L1–L5) prior to becoming an imago, and full development from egg to adult takes between eight months and a year to complete. They will sting both in each of the five nymph stages and as adults, engorging an increasing amount of blood. Using D. maximus, which is the largest one among the Triatominae bugs, rather large blood samples can be obtained.¹⁰ The adults become as long as 4.5 cm and suck up to 4 mL of blood.⁹ In order to prevent blood from clotting during ingestion of host blood and digestion of fed blood, D. maximus uses the protein dipetalogastin as an anticoagulant.^11,12 Dipetalogastin is a heparin-like substance secreted by the bug, and the living blood cells should not be damaged by the bug.¹³ Hence, the use of D. maximus for blood sampling is indeed possible, and the aim of this study is to investigate whether the use of the bug method will severely affect the parameters often used in clinical chemistry and haematology.

Materials and methods

Animals

For this study a total of 28 D. maximus larval stage 5 (L5) (Leibniz Institute for Zoo and Wildlife Research, Berlin, Germany) were used. They were housed in glass jars in an incubator at an average humidity of 83.2 ± 4% and an average temperature of 28.2 ± 0.2°C. Four bugs were used for each of five rabbits adding up to 20 bugs in total. However, some of the bugs did not sting and were replaced by other bugs after a 5 min period. No bug was used more than once. The bugs were not killed after use.

Five female New Zealand White rabbits aged approximately 11 weeks at arrival (body weight range: 2.4–2.6 kg) were used for this study (Lidköping Kaninfarm, Sweden). These five animals were part of a larger group of rabbits and they were kept in a pigpen housing a total of six rabbits. The pigpen contained wood shavings and straw bedding and the rabbits had access to shelters. The pigpen was cleaned once a week. New straw was added every day. The rabbits were fed Altromin 2120 (Altromin Specialfutter GmbH & Co KG, Lage, Germany) and had access to water ad libitum.

Blood sampling

Blood samples were drawn from all rabbits using both a conventional method, namely blood collection from vena auricularis, and the ‘bug method’. Blood sampling by means of bugs was done immediately after the blood sampling by the conventional method. A total of eight samples were drawn from each animal, four conventional samples and four ‘bug samples’. The rabbits were not fasted prior to blood sampling.

Blood sampling: the conventional method

One hour before blood sampling, the rabbits were captured and a local analgesic (Emla^® 30 g creme AstraZeneca, Albertslund, Denmark) was applied over the vena auricularis caudalis of both ears. The ears were covered with occlusive tape and left for an hour for the analgesic effect to set in. The rabbits were released between application of the Emla creme and blood sampling.

Blood sampling was done from the vena auricularis caudalis with a 30 mm 21 G needle (0.6414 mm in outer diameter). After each extraction the needle was removed, and a new needle was inserted into the vein of the other ear. Blood was extracted from each ear twice in the order left–right–left–right. An amount of 3 mL blood from each ear was extracted per course, giving a total of 12 mL of blood per rabbit.

Handling of the blood sample

From each 3 mL sample, 2 mL of blood was directed into a 2 mL ethylenediaminetetraacetic acid (EDTA) (Vacuette^®; Greiner Bio-One, Kremsmünster, Austria) tube and 1 mL blood was directed into two 4 mL serum (Vacuette^®) tubes. The EDTA tubes were turned gently 5–7 times and placed vertically. The serum tubes were placed horizontally on the table. The serum tubes were centrifuged at 2000 g for 5 min, and the serum was transferred to CryoTubes™ and frozen at −20°C. The EDTA tube was placed at room temperature and was not handled further prior to delivery to laboratory.

Blood sampling: the bug method

A glass jar containing four bugs with an unfolded swab secured with a rubber band over the top of the jar was prepared. During blood sampling, each rabbit was placed on the lap of the technician. The glass jar with the bugs was placed on the back of the neck/on the resting ears of the rabbit. After the bugs had finished extracting blood from the rabbit, the glass was removed. A mean time of approximately 15 min was required for the bugs to ingest a blood meal.² Each bug was restrained by hand and positioned with the head between two fingers. A syringe with a 21 G needle was inserted into the abdomen, blood was withdrawn and transferred to two microcentrifuge tubes (Plastibrand^®; VWR International Ltd, Ireland) in order to make a sample for clinical chemistry and a sample for haematology. In order to fulfil the requirements of the laboratory a sample of a minimum of 500 µL full blood for clinical chemistry and a minimum of 300 µL plasma for haematology was prepared. In case an adequate amount of blood could not be extracted from a bug, one microcentrifuge tube was filled and the blood was used for haematology testing only. Anticoagulants were not added to the test tubes as an anticoagulant (dipetalogastin) was already added by the bug.

Handling of the blood sample

The microcentrifuge tube with blood for haematology testing was centrifuged at 2000 g for 5 min, and plasma was transferred to CryoTubes™ and frozen at −20°C. The microcentrifuge tube with full blood was placed at room temperature and not handled further prior to delivery to laboratory.

Laboratory tests

The biochemistry and haematology analyses were conducted at Veterinary Clinical Pathology, Department of Small Animal Clinical Sciences, Faculty of Life Sciences, University of Copenhagen.

The EDTA tubes collected by the conventional method and the full blood samples collected by the bug method were delivered to the laboratory the same day as the blood was extracted. The serum samples collected by the conventional method and plasma samples collected by the bug method were delivered to the laboratory the day after the last blood samples were extracted.

The biochemistry analyses (Table 1) were performed using Siemens ADVIA^® 1800 Chemistry System (Siemens Healthcare Diagnostics, Deerfield, IL, USA). The haematology analyses (Table 2) were performed using Siemens ADVIA^® 120 Hematology System (Siemens Healthcare Diagnostics).
Table 1
Results for clinical chemistry in blood sampled from either vena auricularis (conventional method) or by the use of kissing bugs (bug sampling)

Conventional method, mean (SD) Bug sampling, mean (SD) P value

Albumin (g/L) 38.7 (2.7) 37.2 (1.2) 0.0895

Glucose (mmol/L) 9.13 (0.85) 9.01 (1.10) 0.6348

Total protein (g/L) 60.2 (6.3) 56.7 (5.4) 0.6757

ALT (U/L) 71.9 (5.6) 79.7 (10.2) 0.8558

Creatinine (µmol/L) 97.2 (4.4) 102.4 (7.9) 0.2037

Amylases (U/L) 261 (17.5) 250 (14.3) 0.0502

Carbamide (mmol/L) 60.2 (0.4) 59.4 (0.6) 0.8005

Magnesium (mmol/L) 1.20 (0.07) 1.17 (0.06) 0.2316

ALP (U/L) 489 (40.1) 440 (17.5) 0.0026

Cholesterol (mmol/L) 2.59 (0.23) 2.24 (0.12) 0.0021

Lipase (U/L) 117 (10.3) 104 (5.1) 0.0060

Phosphate (mmol/L) 2.13 (0.11) 2.02 (0.08) 0.0049

GGT (U/L) 5.5 (1.0) 2.2 (2.5) <0.0001

Calcium (mmol/L) 3.42 (0.12) 3.29 (0.09) 0.0211

Sodium (mmol/L) 152.6 (9.4) 140.8 (3.6) <0.0001

Potassium (mmol/L) 4.87 (0.08) 5.61 (1.06) 0.0025

SD: standard deviation; ALT: alanine aminotransferase; ALP: alkaline phosphatase; GGT: gamma-glutamyl transferase

Significant difference between the two methods

Table 2
Haematology in blood sampled either from vena auricularis (conventional method) or by the use of kissing bugs (bug sampling)

Conventional method, mean Bug sampling, mean P value

Haemoglobin (mmol/L) 7.54 (0.21) 7.75 (0.74) 0.2621

Haematocrit 0.32 (0.07) 0.33 (0.02) 0.7617

Reticulocytes ×10⁹/L 80.4 (4.2) 74.2 (21.7) 0.1937

White blood cell count (WBC) ×10⁹/L 6.1 (0.9) 4.2 (0.7) <0.0001

Red blood cell count (RBC) ×10¹²/L 5.65 (0.17) 4.98 (0.32) <0.0001

Platelets ×10⁹/L 599 (26) 198 (60) <0.0001

Mean corpuscular volume (fL) 59.6 (0.2) 65.63 (5.42) 0.0001

Mean platelet volume (fL) 8.11 (0.24) 13.7 (6.6) 0.0011

Neutrophil granulocytes ×10⁹/L 1.5 (0.3) 0.7 (0.5) <0.0001

Eosinophil granulocytes ×10⁹/L 0.06 (0.01) 0.12 (0.18) 0.2401

Lymphocytes ×10⁹/L 4.1 (0.5) 3.1 (0.6) <0.0001

Basophil granulocytes ×10⁹/L 0.35 (0.09) 0.20 (0.05) <0.0001

Monocytes ×10⁹/L 0.12 (0.06) 0.08 (0.03) 0.0025

SD: standard deviation

Significant difference between the two methods

Statistical analysis

Each parameter was analysed by means of a one-way analysis of variance (one-way ANOVA) model with additional rabbit-specific random effects to account for the fact that repeated measurements were taken from each rabbit and different residual standard deviations for the two methods allow different variation for each method. This model allows assessment of differences between the two methods with reference to both the mean level and the precision of each parameter. A likelihood-ratio test was used to assess differences between the two methods with reference to the residual standard deviations: The difference between −2 Res Log Likelihood values as reported by the statistical software was calculated, and the critical value was found using a χ ² distribution with one degree of freedom. The significance level was set to 5%. Data were analysed using SAS statistical software version 9.1 (SAS Institute Inc, Cary, SC, USA).

Results

In total, 28 bugs were used for blood sampling. Only 17 bugs did sting, and thus blood was only extracted from 17 bugs, giving a total of 11 bugs (39.3%) not stinging or not sucking blood.

The mean value of blood extracted from the bugs was 1.2 mL with a minimum of 0.4 mL and a maximum of 2.6 mL. This gave a total of 13 bug samples for biochemistry testing and 17 bug samples for haematology testing. From rabbit 1 the first two conventional blood samples had coagulated and could not be used for haematology testing. None of the bug samples coagulated.

Clinical chemistry

There was no significant difference between the two sample methods with regard to albumin, glucose, total protein, alanine aminotransferase (ALT), creatinine, amylases, carbamide and magnesium.

Alkaline phosphatase (ALP), cholesterol, lipase, phosphate, gamma-glutamyltransferase (GGT), calcium and sodium were all significantly higher in the conventional blood samples compared with bug blood, whereas potassium was significantly lower in the conventional blood samples (Table 1).

Five of the 13 bug samples for GGT had the value 0 IU/L and were excluded from the data. The conventional GGT samples all had values above 2 IU/L. The sodium value was significantly lower in the bug blood, whereas the potassium value was significantly higher in the bug blood.

Total bilirubin was measured, but since 12 of the 13 bug samples had a value of 0 µmol/L, whereas the values of the conventional method were between 0.1 and 0.4 µmol/L, no comparative statistics were done on this parameter, and hence the parameter was not included in Table 1.

Haematology

No significant differences were found for haemoglobin, haematocrit, reticulocytes and eosinophilic granulocytes. White blood cell count, red blood cell count, platelets, neutrophilic and basophilic granulocyte counts as well as lymphocyte and monocyte counts were all significantly higher in blood samples collected by the conventional method, whereas mean platelet volume and mean corpuscular volume (MCV) were significantly higher in samples collected by the use of bugs. The results are listed in Table 2. Two of the conventional blood samples coagulated and could not be used for haematology. None of the bug blood samples coagulated.

Discussion

The set-up for this study originally included four bugs per rabbit, giving a total of 20 bugs each of which representing a blood sample. In the study conducted, 28 bugs were used as a total of 11 bugs did not sting. Eight of these bugs were replaced and a total of 17 bug blood samples were obtained. Hence in this study the ‘stinging success rate’ was only 60%. The bugs had been fasted for five weeks prior to the study; however, they did not sting in the experiment. One reason could be that the bugs normally feed on resting/sleeping animals when it is dark, and in this study they were placed in a see-through glass jar in a bright room. This means that the conditions for the bug were very different from natural conditions. Furthermore, warm-blooded animals may not only signal a blood meal, but they may also represent a possible predator which the bug will try to avoid.^14,15 This hypothesis is consistent with our observations of bugs moving in the opposite direction of the rabbit and might indicate that the bugs that did not sting perceived the rabbit as a predator. The blood sampling was done in the rabbit housing facility, and as the bugs were housed at approximately 28°C the temperature drop may also have influenced the performance of the bugs.

Two of the conventional blood samples coagulated and could not be used for haematology. EDTA was added to the conventional blood samples in this study, but occasionally rabbit blood clots despite the anticoagulants added.¹⁶ The bug adds dipetalogastin to the ingested blood and clotting did not occur in the bug blood samples. This gives the bug method an advantage over the conventional method, as no bug samples were destroyed.

There was no significant difference between the two sample methods for albumin, glucose, total protein, ALT, creatinine, amylases, carbamide and magnesium. The measured values for ALP were 489 and 440 U/L in the conventional and the bug samples, respectively, which seem to be quite high. Reference values provided in the literature vary (e.g. 40–140 U/L¹⁷ and <120 U/L¹⁸) but are much lower. Since ALP is an age-dependent parameter with relatively high values during the early growth phase due to osteoblastic activity, the young age of these rabbits may have resulted in these high values. Skeletal growth for the New Zealand White rabbit is completed at 28 weeks.¹⁹ As the rabbits used for this study had an approximate age of 17 weeks this could be the explanation for the high ALP values.

Our rabbits had not been fasted prior to the blood sampling and abnormal values were therefore to be expected as cholesterol levels should always be measured in fasted rabbits. In fasted rabbits, the reference value for cholesterol level is 0.9–1.7 mmol/L,¹⁷ which is much lower than our findings that had a mean cholesterol level at 2.59 mmol/L (SD: 0.23) for conventional sampling and 2.24 mmol/L (SD: 0.12) for bug sampling. However, all rabbits in this study were handled the same way for both methods, so not being fasted is believed to have had no influence on the purpose of determining a difference between the two methods. Since we do not know if bug metabolism in the anterior midgut had an impact on cholesterol values, it is suggested not to use the bug method until studies on fasted rabbits have been carried out.

The bug method for lipase had lower values than the conventional method. This could indicate that the bug's metabolism had an influence on the test results.

For phosphate and calcium measures, the likelihood ratio test showed that there was no significant difference between the SD of the two methods. This could indicate that it might be possible to use the bug method despite the significant difference.

Five of the 13 bug samples for GGT had the value 0 IU/L. The conventional samples all had values above 2 IU/L and there was a significant difference between the two methods. This could indicate that the bug’s metabolism had an influence on the test results.

The sodium value was significantly lower in the bug blood, whereas the potassium value was significantly higher in the bug blood. R. prolixus* moves water from the lumen of the anterior midgut into the haemolymph by iso-osmotic transport, and it has been suggested that the movement of water across the epithelium is secondary to ion transport and is linked to an ouabain-sensitive Na⁺/K⁺ ATPase located in the epithelium.²⁰ There are no published data describing the movement of water in D. maximus; however, it is reasonable to hypothesize that it is performed as in R. prolixus. Therefore, the passive movement of Na⁺ across the basal membrane into the epithelial cell driven by Na⁺/K⁺ ATPase in the apical membrane could explain why there was a significant difference in sodium between the two methods with the value of the bug method being less than the value of the conventional method. The sodium transport across the epithelia could indicate that it is not advisable to use the bug method to measure sodium. The basal membrane of insects also contains K⁺/H⁺ antiports which exchange extracellular H⁺ for intracellular K⁺ driven by V-ATPases.^21–23 These antiports could explain why the potassium value in the bug samples was higher than the value of the conventional method. Due to the possible addition of potassium to the blood the bug method may not be suitable for measuring potassium.

In 12 of the 13 bug samples the total bilirubin had the value 0 µmol/L. Due to this, a significant difference between the bug method and the conventional method could not be calculated. The values of the conventional method were between 0.1 and 0.4 µmol/L and this could indicate that the metabolism of the bug had an influence on the results and that the method using bugs should probably not be used for determining the bilirubin content.

The bug method values for red blood cell count, white blood cell count, neutrophil count, lymphocyte count, basophil and monocyte counts, and platelet count were lower than the values of the conventional method. This could indicate that the bug’s metabolism had an influence on the results.

The erythrocytes contain much of the bugs’ protein source in the form of haemoglobin, and in the anterior midgut haemoglobin is made accessible to digestion by lysis of the erythrocytes.²⁴ Destruction of the erythrocytes by lysis could explain why the red blood cell value was lower for the bug method than for the conventional method.

The MCV values of the bug method were higher than the values for the conventional method. An increase in MCV can be produced by cell swelling, and it could be hypothesized that cell swelling of erythrocytes precedes cell lysis in the midgut of the bug. However, as there was no significant difference between the two methods with regard to haematocrit (the packed cell volume) and haemoglobin, these results are not consistent as cell swelling, for example, should have influenced the haematocrit as well as the MCV. Further experiments including manual evaluation of blood smears and assessment of, for example, red cell distribution width and mean corpuscular haemoglobin concentration are needed to assess the exact fate of the erythrocyte in the bug.

Platelets are activated by vessel damage and platelet response to vascular injury by adhesion to the vessel wall after which they adhere to one another and swell. It could be hypothesized that the higher values for platelets could be caused by platelet activation resulting from the insertion of needles into the vena auricularis caudalis. However, all the rabbits were bled from the ears prior to blood sampling using bugs, and it is not possible to state whether such activation would have been seen had the bugs been used prior to the conventional method. To further assess the influence of the bug method on platelet activation/platelet counts, a study randomizing the use of the two methods (conventional sampling followed by bug blood compared with bug blood followed by conventional sampling) should be done. The number of neutrophil granulocytes was significantly lower in the bug samples, which could be hypothesized to be due to adhesion to the interior of the bug midgut.

For the blood parameters with a significant difference between the two methods, it has been postulated that the difference could be due to the metabolism of the bug. However, other factors may contribute as well. Conventional blood samples are collected from the veins. The bug blood samples derive from unknown sources as it is not known whether the blood originates from a vein or an artery. A study in dogs by Lue et al.²⁵ compared arterial and venous blood samples and demonstrated significant differences in white blood cell counts, lymphocytes, erythrocytes and platelet counts, haemoglobin and haematocrit. Hence it was concluded that complete blood counts obtained from canine venous and arterial blood samples may not be comparable. It is reasonable to believe that this difference between venous and arterial blood also apply to rabbit blood. Consequently, this is also a factor to consider prior to the use of the bug method.

The bug method may provide an alternative for both haematology and clinical chemistry with further optimization and validation. It is important to realize that some parameters may be affected by the use of the bug; however, this study and others studies have demonstrated that all tested parameters in haematology, clinical chemistry and endocrinology are measurable using the bug method.^2,4,6,9

The major limitation of the method seems to be the varying success of stinging. Other studies have tried to optimize the method by hiding the bugs in containers/shelters for monkeys and lynxes and hollowed artificial eggs for birds,^6,9,26 but still it is recognized that to ensure the number of blood samples needed, more bugs should initially be used.⁷ Such a practice will also ensure that the required volume is obtained. For larger laboratory animals such as pigs and goats, purpose-made collars or body harnesses with small, attachable, dark, isolated boxes containing the bugs could be produced to improve the stinging rate. The bugs used in this study were fasted for five weeks which should have been sufficient to ensure stinging. However, prior to selecting bugs for blood sampling, the researcher could place his hand close to the bugs and choose only bugs that respond positively to the hand by protruding the proboscis.

Another limitation may be the small blood volume taken by some of the bugs. This fact emphasizes the need for starting up more bugs than the actual number of blood samples needed.

In relation to laboratory animal science the bug method also has limitations in pharmacokinetics, for example, where high demands are made on precision when timing the blood samples and on the inter-sample intervals. Another possible limitation of the use of the bug method in laboratory animal science is the exposure of the test animal to the bug antigens.⁷ This may induce an unwanted, uncontrolled immune response. However, it is equally important to recognize the potential in especially experimental animal behaviour research where the sampling of blood from completely undisturbed animals may be crucial, and in the veterinary clinic, where the bugs may be able to sample from patients who are difficult to handle or patients who either have very small vessels or have vessels that for some other reason are difficult to access. Aside from being painless and hence probably stress-free, blood sampling using bugs does not include the risk of haematomas, and the blood is instantly mixed with a potent and reliable anticoagulant.⁷ The present study has shown that the blood samples obtained by the bug method are of a high quality. Hence the use of bugs for blood sampling seems to provide a feasible and manageable tool in both the laboratory animal facility and the veterinary clinic. However, further validation and improvements should be done to optimize the stinging rate and to evaluate possible immunomodulatory effects of the bug method.

	Conventional method, mean (SD)	Bug sampling, mean (SD)	P value
Albumin (g/L)	38.7 (2.7)	37.2 (1.2)	0.0895
Glucose (mmol/L)	9.13 (0.85)	9.01 (1.10)	0.6348
Total protein (g/L)	60.2 (6.3)	56.7 (5.4)	0.6757
ALT (U/L)	71.9 (5.6)	79.7 (10.2)	0.8558
Creatinine (µmol/L)	97.2 (4.4)	102.4 (7.9)	0.2037
Amylases (U/L)	261 (17.5)	250 (14.3)	0.0502
Carbamide (mmol/L)	60.2 (0.4)	59.4 (0.6)	0.8005
Magnesium (mmol/L)	1.20 (0.07)	1.17 (0.06)	0.2316
*ALP (U/L)	489 (40.1)	440 (17.5)	0.0026
*Cholesterol (mmol/L)	2.59 (0.23)	2.24 (0.12)	0.0021
*Lipase (U/L)	117 (10.3)	104 (5.1)	0.0060
*Phosphate (mmol/L)	2.13 (0.11)	2.02 (0.08)	0.0049
*GGT (U/L)	5.5 (1.0)	2.2 (2.5)	<0.0001
*Calcium (mmol/L)	3.42 (0.12)	3.29 (0.09)	0.0211
*Sodium (mmol/L)	152.6 (9.4)	140.8 (3.6)	<0.0001
*Potassium (mmol/L)	4.87 (0.08)	5.61 (1.06)	0.0025

	Conventional method, mean	Bug sampling, mean	P value
Haemoglobin (mmol/L)	7.54 (0.21)	7.75 (0.74)	0.2621
Haematocrit	0.32 (0.07)	0.33 (0.02)	0.7617
Reticulocytes ×10⁹/L	80.4 (4.2)	74.2 (21.7)	0.1937
*White blood cell count (WBC) ×10⁹/L	6.1 (0.9)	4.2 (0.7)	<0.0001
*Red blood cell count (RBC) ×10¹²/L	5.65 (0.17)	4.98 (0.32)	<0.0001
*Platelets ×10⁹/L	599 (26)	198 (60)	<0.0001
*Mean corpuscular volume (fL)	59.6 (0.2)	65.63 (5.42)	0.0001
*Mean platelet volume (fL)	8.11 (0.24)	13.7 (6.6)	0.0011
*Neutrophil granulocytes ×10⁹/L	1.5 (0.3)	0.7 (0.5)	<0.0001
Eosinophil granulocytes ×10⁹/L	0.06 (0.01)	0.12 (0.18)	0.2401
*Lymphocytes ×10⁹/L	4.1 (0.5)	3.1 (0.6)	<0.0001
*Basophil granulocytes ×10⁹/L	0.35 (0.09)	0.20 (0.05)	<0.0001
*Monocytes ×10⁹/L	0.12 (0.06)	0.08 (0.03)	0.0025

References

Von Helversen

, Reyer

. Nuclear intake and energy expenditure in a flower visiting bat. Oecologia 1984;63:178–84

Voigt

, Fassbender

, Dehnhard

, Validation of a minimally invasive blood-sampling technique for the analysis of hormones in domestic rabbits, Oryctolagus cuniculus (Lagomorpha). Gen Comp Endocrinol 2004;135:100–7

Voigt

, von Helversen

, Michener

, Kunz

. Validation of a non-invasive blood-sampling technique for doubly-labelled water experiments. J Exp Zool A Comp Exp Biol 2003;296A:87–97

Voigt

, Peschel

, Wibbelt

, Frolich

. An alternative, less invasive blood sample collection technique for serologic studies utilizing triatomine bugs (Heteroptera; Insecta). J Wildlife Dis 2006;42:466–9

Arnold

, Oswald

, Voigt

, Taking the stress out of blood collection: comparison of field blood-sampling techniques for analysis of baseline corticosterone. J Avian Biol 2008;39:588–92

Braun

, Frank

, Dehnhard

, Pregnancy diagnosis in urine of Iberian lynx (Lynx pardinus). Theriogenology 2009;71:754–61

Vos

, Muller

, Neubert

, Voigt

. Validation of a less invasive blood sampling technique in rabies serology using reduviid bugs (Triatominae, Hemiptera). J Zoo Wildlife Med 2010;41:63–8

Dan

, Pereira

, Pesquero

, Diotaiuti

, Beirao

PSL

. Action of the saliva of Triatoma infestans (Heteroptera : Reduviidae) on sodium channels. J Med Entomol 1999;36:875–9

Thomsen

, Voigt

. Non-invasive blood sampling from primates using laboratory-bred blood-sucking bugs (Dipetalogaster maximus; Reduviidae, Heteroptera). Primates 2006;47:397–400

10.

Aguirre

, Frede

, Rubiolo

, Canavoso

. Vitellogenesis in the hematophagous Dipetalogaster maxima (Hemiptera: Reduviidae), a vector of Chagas’ disease. J Insect Physiol 2008;54:393–402

11.

Lange

, Keilholz

, Schaub

, Landmann

, Markwardt

, Nowak

. Biochemical characterization of a thrombin inhibitor from the bloodsucking bug Dipetalogaster maximus . Haemostasis 1999;29:204–11

12.

Mende

, Petoukhova

, Koulitchkova

, Dipetalogastin, a potent thrombin inhibitor from the blood-sucking insect Dipetalogaster maximus – cDNA cloning, expression and characterization. Eur J Biochem 1999;266:583–90

13.

Von Helversen

, Volleth

, Nunez

. A new method for obtaining blood from a small mammal without injuring the animal – use of triatomid bugs. Experientia 1986;42:809–10

14.

Pereira

, Penido

, Martins

, Diotaiuti

. Comparative kinetics of bloodmeal intake by Triatoma infestans and Rhodnius prolixus, the two principal vectors of Chagas disease. Med Vet Entomol 1998;12:84–8

15.

Schmitz

, Trenner

, Hofmann

, Bleckmann

. The ability of Rhodnius prolixus (Hemiptera; Reduviidae) to approach a thermal source solely by its infrared radiation. J Insect Physiol 2000;46:745–51

16.

Perryclark

, Meunier

. Vascular access ports for chronic serial infusion and blood-sampling in New Zealand white rabbits. Lab Anim Sci 1991;41:495–7

17.

Evans

. Animal Clinical Chemistry – A Practical Handbook for Toxicologists and Biomedical Researchers. 2nd edn. Portland: CRC Press/Taylor & Francis, 2009

18.

Thrall

, Baker

, Campbell

, Veterinary Hematology and Clinical Chemistry. Oxford: Blackwell Publishing, 2006

19.

Masoud

, Shapiro

, Kent

, Moses

. A longitudinal-study of the growth of the New Zealand White rabbit – cumulative and biweekly incremental growth-rates for body length, body-weight, femoral length, and tibial length. J Orthopaed Res 1986;4:221–31

20.

Farmer

, Maddrell

SHP

, Spring

. Absorption of fluid by the midgut of Rhodnius. J Exp Biol 1981;94:301–16

21.

Wieczorek

, Putzenlechner

, Zeiske

, Klein

. A vacuolar-type proton pump energizes K⁺/H⁺ antiport in an animal plasma-membrane. J Biol Chem 1991;266:15340–7

22.

Lepier

, Azuma

, Harvey

, Wieczorek

. K⁺/H⁺ antiport in the tobacco hornworm midgut – the K⁺-transporting component of the K⁺ pump. J Exp Biol 1994;196:361–73

23.

Azuma

, Harvey

, Wieczorek

. Stoichiometry of K⁺/H⁺ antiport helps to explain extracellular Ph 11 in a model epithelium. FEBS Lett 1995;361:153–56

24.

Deazambuja

, Guimaraes

, Garcia

. Hemolytic factor from the crop of Rhodnius prolixus – evidence and partial characterization. J Insect Physiol 1983;29:833–7

25.

Lue

APV

, Jensen

, Strom

, Kristensen

. Comparative analysis of haematological, haemostatic, and inflammatory parameters in canine venous and arterial blood samples. Vet J 2007;173:664–8

26.

Becker

, Voigt

, Arnold

, Nagel

. A non-invasive technique to bleed incubating birds without trapping: a blood-sucking bug in a hollow egg. J Ornithol 2006;147:115–18