Notes | Bat sampling techniques

Gilles Demaneuf

11 min readJul 1, 2022

Quick working notes on sampling, testing, sequencing, isolating.
Not yet finished.

1. About bats

1.a Main points

The important elements to remember about bats are that:

They come in many different species, some very small (too small for anal swabs).
They can be insectivore or fructivore.
They use a lot of energy to fly.
They fly out at night to get food.
They go into torpor in the cold months, since they would likely use more energy than they could collect when flying at night in cold weather, which is made worse by the fact that the source of this energy (insects or fruits) itself is largely seasonal.
Their roosting places are typically trees (if the weather is not too cold) or caves (old mines will do too).
They sometimes migrate from winter roosting places to summer locations.
They keep warm in their roosting place by staying close together.
They often mix with other species in the same cave.
They harbour many different viruses which they share easily in their roost or when migrating.
They typically are asymptomatic to such viruses, possibly because of a high body temperature and natural resistance.

1.b When do bats go into torpor in Yunnan?

Simple answer: they hibernate when the temperature at night outside and the limited availability of food (fruits or insects) does not make it worth flying out.

Practical answer: they typically hibernate in Yunnan from end December to mid March. Obviously local conditions may vary, as well as the weather year-to-year.

**Average monthly temperature in Kunming (Yunnan), China (celsius)**

2. Bat Sampling

Bat sampling can take different forms:

blood sampling
fecal sampling via anal swabs
feces sampling via splat sheets
tissue (for cell cultures of for sequencing the bat itself for identification)

Samples may be done to try to collect viruses or for other uses (e.g. getting tissues).
Best season for sampling for viruses is after the baby bats are born, as they do not yet have built their defences, hence ~ September in Yunnan[source SZL].

Dr. Guangjian Zhu from the Eco Health Alliance and researcher Libiao Zhang from the Guangdong Institute of Biological Resources Application collected a large number of samples for the WIV across the country. [source: Ben Hu interview]

The below PREDICT-2 grant details by EHA has a good short paragraph on sampling techniques to be used on caught bats.

[A very good guidebook: https://www.wabnet.org/wp-content/uploads/2019/12/WAB-NET-FIELD-protocols.pdf]

2.a Direct sampling (i.e: caught bats):

One important finding of SZL was the traces of BatCoV in bats feces are typically ephemeral, while blood sampling of the same bats will often show BatCoV antigens.

Blood, saliva, nose and tissue sampling:
The best way to determine the exposure of a specific bat to BatCoVs is through blood sampling looking for antigens. However this is not very practical and does not pinpoint an exact BatCoV.

Another way is through tissue sampling, but that requires killing the bats and is usually avoided if possible.

The most common BatCoV detection techniques thus involves anal swabs of caught bats.

Anal swabs:
Anal swabs are a very common method. In that case nets are positioned at the entrance of the caves in the early morning before the bats come back from their night feeding excursions. This is obviously only possible when the bats are not hibernating (so typically not in Jan-Feb in Yunnan).

Bats are then manually handled, bagged and swabbed, then released — all in the same day.
There are however possible risk of sample contamination via the gloves (used with multiple bats) and even the bats fur. As mentioned above this technique will not work for some very small species of bats.

Practical limitations:

Tissue sampling is obviously destructive and often limited to a few bats, maybe the ones injured during capture — wand which would typically be euthanised anyway.
Anal swabs is not practical for some very small bat species
Some bat species are notoriously difficult to catch.

Cold Chain:

Samples stored in vial boxes on ice in styrofoam cooler, may be place in cryoshipper, until transfer to ultra-cold freezer (-80°C).

Collected diagnostic samples are to be stored temporarily in vial boxes (consecutively numbered) on ice in a styrofoam cooler until sampling for the evening is completed. Filled cryovial boxes are to be placed in the cryoshipper at the end of the sampling night, or repeatedly throughout the night, for short-term storage in the field (< 1 week) and immediately transferred to an ultra-cold (-80°C) freezer for long-term storage in the laboratory.
Western Asia Bat Research Network, Standard Operating Protocols
Each sample box containing 81 cryovial tubes was placed inside an individual plastic ziplock and transported from the field to IPL on dry ice in a polystyrene foam box. At IPL, samples were stored in a specific −80 °C freezer until analysis
’Bat coronaviruses related to SARS-CoV-2 and infectious for human cells’

Positioning a net to capture bats on their way back to a cave. Note the absence of PPE or gloves. From Shi Zheng Li’s presentation

The catch of the day. Bats hanging in cloth bags.

Olival from EcoHealth Alliance releasing bats from one of these typical cloth bags hanging from a rope behind him.. Field lab in Thailand. Source.

The USAID PREDICT team in Thailand takes blood from a fruit bat to screen for viruses. / Montakan Tanchaisawat, USAID. Source.

A fruit bat gets a sip of nourishment after veterinarians take a blood sample in Chonburi, Thailand. / Richard Nyberg and Montakan Tanchaisawat, USAID. Source. Note the typical white bag the bat is in.

Weighing a bat in its bag with a Pesola spring scale

2.b Indirect sampling: Splat Sheets

Means non-invasive, the bats are not handled. Can be used for urine too.

A common approach, in bat-borne disease research, involves the capture of many individual bats repeatedly over time, where bats are sampled (e.g. serum, urine, saliva) and tested for viral presence using serology or PCR techniques. Best case scenario, longitudinal samples are obtained for multiple individuals, enabling both the discovery of new viruses and description of dynamics in individual-level viral prevalence. Individual level longitudinal data are more common for high-fidelity cave-roosting bats which can be recaptured frequently at the same roosting site (Streicker et al. 2012, Towner et al. 2009).
However, these type of longitudinal data are much more difficult to gather from the tree roosting megachiroptera, such as Pteropus and Eidolon genera (Hayman et al. 2012), which are highly-mobile nomadic foragers, making them poor candidates for ecological studies that rely on recapture of individuals.
Therefore, recent research has supplemented the capture of individual bats with a non-invasive sampling technique that uses plastic sheets to collect urine and feces under bat roosts (Baker et al. 2012, Baker et al. 2013, Chua 2003, Chua et al. 2002, 2001, Edson et al. 2015a, Field et al. 2011, 2015, Marsh et al. 2012, Pritchard et al. 2006, Smith et al. 2011, Wacharapluesadee et al. 2010).

Another technique is the use of splat sheets. This technique was pioneered in around 2007 in Thailand. It consists in positioning clean plastic sheets set in frames (or possibly trays) under the bat roosting sites (more easily done during the night when the bats are out). One then retrieved these splat sheets soon after the bats are back in their caves after feeding outside. The feces splats can then be sampled.

It is much more difficult to identify the exact bat species with a splat sheet. All that one can tell is that the splat sheet was under a specific roost, which normally is made of a certain bat species. But that’s all.

Splat sheets are however good for small bat species and for sectional sampling. In particular they can be used for metagenomic analysis (using shotgun sequencing).

A veterinarian scoops up bat droppings to test for viruses in Chonburi, Thailand. / Montakan Tanchaisawat, USAID. Source

2.c Splitting the samples

Often when PREDICT/USAID is involved, samples get split, especially if they are blood samples. One sample goes to the US, the other one to a local country lab (or to China as was requested by EHA for years 4 and 5 of PREDICT-2), as described by David Quanmen in Spillover (click ‘search inside’).

2.d Safety Recommendations

Practically the main concern of sampling teams seems to be rabies. All members of the sampling teams have to be vaccinated against rabies as bites when handling bats are rather common.

For other bat-borne viruses, the guiding safety principle is to take a precautionary approach, wearing PPEs with the assumption that there is some risk.

See ‘SEABCRU Recommendations for Personal Protective Equipment (PPE) and Health Safety while Conducting Bat Research’, from the ‘Southeast Asian Conservation Research Unit’ (all credit goes to DRASTIC Billy Bostickson for these links)

2.e Safety practices

While rabies vaccination of field samplers seems widespread, as far as BatCoVs are concerned, a rather relaxed attitude is often prevalent.

Here is an interview with David Quammen, describing how bats were sometimes caught by Eco Heath Alliance and bagged, without PPE, during a EcoHealth Alliance sampling trip to Guilin, Guangxi province (not Guangdong as the NPR piece wrongly states), where Chmura and his Chinese colleagues went looking for SARS-like viruses in bat caves.

Shi Zheng Li handling bats with students. Note the absence of mask for one of them. Source

Here is a presentation by Shi Zheng Li explaining that most of the time minimal PPEs are used, unless they know that they are human-transmissible BatCoVs on that sampling site. It is not very clear how she can determine that (1) before actually sampling that site and (2) despite possible mutations and migrations after initial sampling!

Safe (left) and not so safe (right and top). From Shi Zheng Li’s presentation

Shi Zheng Li words above are in obvious contradiction to what she then told the WHO team when they visited the WIV for a few brief hours.

2.f Accidents

As it is, accidents do happen. Michel Callahan (who worked for USAID, DARPA and the CTR programs) mentioned some:

The above SEABCRU safety recommendations also mention what could well be the same accident:

This risk was also mentioned by Marc Eloit from Institut Pasteur, who has been very active sampling the BANAL viruses in Laos and found the closest relatives of SARS-CoV-2:

3. Virus Identification

3.a Background:

A sample done to collect viruses could contain any type of virus, but will be typically checked using a not-so-broad PCR test or less specific metagenomic analysis.
If the test comes back positive, that sample is said to be “positive” which is still rather vague, as in positive for a bat-cov or a SARS-like BatCov.

3.b PCR Testing:

In DNA metabarcoding, PCR is conducted using genetic markers (i.e., DNA barcodes) that target specific taxonomic groups, the PCR products are sequenced using NGS, and the data are compared to DNA sequence databases to categorize and annotate the diversity of DNA sequences in a sample (Huson et al., 2016; Pompanon et al., 2012; Taberlet et al., 2012). See https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6050187/

3.c NGS Metagenomic analysis:

RNA virus metagenomics has developed rapidly with the availability of next generation sequencing platforms (NGS). By contrast to PCR testing, metagenomic analysis does not need a targeted primer and will detect any virus in the sample, including new ones. When applied to feces of insectivorous bats, the technique will detect bat viruses, but also insect viruses (the bats eat the insects) and plant viruses (the bats eats insects that eats the plants…).

TO DO: Check these:

4. Sequencing:

Some viruses are fully sequenced [typically a consensus sequence], some only partially [the RdRp]
The RdRp is typically used to ‘index’ coronaviruses. When sequencing was still expensive, only the RdRp may have been sequenced.
It is now much easier to sequence the whole lot with NGS techniques (New Generation Sequencing, see https://pubmed.ncbi.nlm.nih.gov/31996413/).
A full sequence is subject to the quality of the sample and the quality of the reads. Quite often a ‘consensus’ sequence is built from the various reads.

5. Isolation & Reverse Genetics:

To do:

Difficulty of isolation
Role of Reverse Genetics

From Ben Hu interview:
The reverse genetics system of SARS-like coronavirus constructed by our team Dr. Zeng Leiping and the virus isolation technique established by Associate Researcher Yang Xing Lou after years of exploration and optimization are important foundations of this research.

See Viral metagenomics, protein structure, and reverse genetics: Key strategies for investigating coronaviruses https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5869085/

6. WIV Collections:

Vocabulary:

Sample: biological sample (blood, feces, etc) — does not necessarily have any virus (may depend on the season, location, etc)
Positive sample: a sample that was came back positive for a certain range of viruses (via PCR or metagenomic analysis)
Isolate: a cultured (live) virus, ‘isolated’ from any other virus. Many times used wrongly as a synonym of strain.
Virus: usually used as meaning a specific virus, main be represented by different strains
Strain: ambiguous. Sometimes used as a semantical distinction for a specific physical instance of a virus, which could be as simple as different cultures (different Petri dishes), or different samples containing the same virus. Sometimes used as a group of very similar viruses.
Recombinant or chimeric virus: an artificially created virus. There are many techniques: In the WIV it was very commonly the BAC infectious clone. Usually starts with a live isolate and then a gene or segment is swapped by another.
TO DO: explain reverse engineering techniques.

The WIV had:

the largest virus collection in Asia and one of the largest in the world, if not the largest [source: https://t.co/i3ZMCGfHBV?amp=1]
collected around 15,000 bat samples (source: https://t.co/i3ZMCGfHBV?amp=1 gives 15,000 (China + Africa) and based on series at ID 9508 in Mar 2016 as per Francisco + ScienceMag) [other source: 19,000]
collected 50,000 tick samples [source: https://t.co/i3ZMCGfHBV?amp=1]
a library of 1,400 types of viruses across all species (animal, insects, etc)[source: https://t.co/i3ZMCGfHBV?amp=1, Francisco 1,353] including
At least 957 CoVs (source: Francisco)
At least 118 MERS related viruses (source: Francisco, + paper)
For a total of 60,000 strains across all species (animal, insects, etc) [source: https://t.co/i3ZMCGfHBV?amp=1]
identified 500 of coronaviruses (source: Daszak interview + ScienceMag), including at least 50 close to SARS
isolated 19 coronaviruses by June2016 — no data since then [source: Francisco]

ScienceMag:

Daszak and Shi’s group have for 8 years been trapping bats in caves around China to sample their feces and blood for viruses. He says they have sampled more than 10,000 bats and 2000 other species. They have found some 500 novel coronaviruses, about 50 of which fall relatively close to the SARS virus on the family tree, including RaTG13

=> May or may not be limited to the WIV work done with EHA — not clear.