That's an interesting question, Freek!

From a conservation point-of-view, releasing individuals seems appropriate. On the other hand, removing single individuals from a sutaining population should do little harm - unless the population is too small (or unstable) already to sustain. In this case, however, one might expect this population to go extinct anyway...

For validation of your findings, one would expect you to keep at least some of your specimens as reference - my suggestion would be not to collect every captured individual but a (representative) selection and release the rest...?

As Martin suggested, it also depends on the size of your population(s) and how efficient your traps are (ultimately what proportion of the population you may collect at each sampling). If this proportion is known negligible, then you can probably remove caught individuals without any trouble. On the other hand, removing individuals from a very small population encompasses several risks (catching a high proportion of one sex, one particular age class, if this particular subset is more prone to be captured for instance). If you have no idea of this proportion, and if you have the time and resources to do so, I suggest you first mark a set of individuals in your traps, release them, and repeat the process to get a more accurate estimate of your population size.

Release, unless it is absolutely necessary to collect. Then, you should have a valid reason and a post-collection plan. Your own ignorance of the taxa can't be a reason. Typically "I don't know what this species is, I should ask my friend". What are you going to do in science with this is sample that is not possible without the sample? Good photographs are often enough. Where is the sample going to go once you loose interest in the sample? The whole things is not just a question of numbers, but also of ethics and being humane.

Collecting some individuals from a local population will have no influence on the population itself, unless the population is too small allready too sustain, but these populations will also go extinct the next years without collecting specimens.

If you want to collect the individuals for not double counting them, then you need to collect them all and this have a great impact on your local population size, especially if you keep going on for some years.

Collecting individuals has sence for taxonomical groups where little information is available of for some parts of the world where a lot of taxonomical work stil need to be done. And this is certainly not the case for butterflies or dragonflies in many parts in Europe, especially western and northern Europe.

Generally, release. Which is the goal of specimen identification, what tools are you planning to use? I agree good photographs are usually enough, other characters can be determined on the field. If you have the possibility to use DNA to classify them, than a wing clip or haemolinph samplin is enough.

Unless there are really reasons to believe that the species in question are highly threatened I would always recommend to make a reference collection of voucher specimens for each monitoring project or survey focusing on insects. Ideally this collection should after completion of the project be deposited in a recognized research institution or museum so that it is also available for future comparisons and taxonomic work. For publication in some journals this may be demanded. This is especially important in countries where many species are not yet described and you have to use codes for them. However, it may also be valuable in the better researched regions as taxonomy changes over time and species may be split later. If you are certain of being able to correctly identify species in the field while the individuals are alive, you should be able to release the majority of the trapped individuals later on in the project. If you are concerned about double counting, then maybe you could mark them before releasing them. Thus recaptures would also give you additional information that you could analyze. In some invertebrate groups or species, identifications of life individuals is, however, not always possible. Rather microscopes or even dissections may be required. And lets face it many people do make mistakes when keying species, especially when they are not yet very experience with the group. So the ability to have some vouchers to check later on is valuable. If you are working with large butterflies and are really concerned that taking even a few individuals may have too much of an effect on the population, then photography may be an alternative. However, in most cases I would recommend a reference collection of vouchers.

The replies so far point to a near-complete agreement about best practice which is encouraging. Double-counting is rarely a big problem, so that is not a reason for collecting and not releasing. Reliable documentation is also necessary, and this may require voucher specimens. In my experience, the usual problem is that voucher specimen deposition does not happen, even if a killing trap (i.e. pitfall trap) is used.

I think we ought to think a bit more about other collection methods where large numbers of specimens are collected, usually by an "automatic" method - I mean how to reduce the damage we cause by sampling when the sampling method kills the specimen.

Although my field of work is marine, our way to proceed could be applicable to other areas, such as butterflies. We try to release specimens (e.g. fishes, megafauna) when possible. If we need to identify them more appropriately we take photos. Only if necessary we take a specimen to be identified at the laboratory. In the case of macroalgae, if there is not necessary to determine biomass, we identify them at field. Only in the case of infauna or plankton we take samples to be identified at the laboratory

A good quality photo(s) is is almost as good as having the specimen, so often you do not need to keep it.

In very well known faunas such as your country (Finland) sampling of butterflies might not be necessary (except for a few very similar or even cryptic species) for an experienced lepidopterist, but in tropical countries (and I assume you are talking about those?) collection of voucher specimens (at least one per species) is usually necessary for any serious scientific project and will not cause any harm to their populations. The remaining specimens should be marked not only to avoid double-counting, but also to gain additional knowledge (e.g. on the life-span or mobility).

I agree with the releasing thing. Despite I totally understand M. Zimmer point of view, what happens with unknown conservation status species? Maybe you consider that a couple of the few wouldn't affect the population, but you would never know for sure if it does. Nowadays there are many things you can do to get some DNA samples or so. And a good picture should be enough to complete the description. Probably, we should consider about policies for new species description and the quantity of samples (dead animals for collections) required to establish whether there is a new spcies or not. Well, just a little bit more of what it is discussed in earlier comments. All the Best, ROcio

Unless specimens are properly curated and stored, then they are usually of no value for identification or other studies. I think the ethical approach is to minimise the number of individuals you kill (appropriate to your study); sometimes killing specimens is necessary, but more often specimens could be live trapped and released. 2010 Seldon, D, Beggs, J.R. The efficacy of baited and live capture pitfall traps in collecting large-bodied forest carabids. New Zealand Entomologist 33:30-37

I work with coccinelids and what I do after individuals have been identified, is to release them at the end of the collecting session, in order not to count them more than once. If I have doubts in some specimen I take a photograph or in the last case I collect the individual.

I totally agree with my colleague, Geert De Knijf: collecting or only photographing/observing the specimen entirely depends on the ability to ID the specimen with 100% certainty in the field. This might well work for larger and well studied invertebrates of well studied regions of the world such as butterflies or dragonflies of the Holarctic. When it concerns much poorer known taxonomic groups like e.g. long-legged flies (Diptera, Dolichopodidae), it's paramount to collect at least some (male) specimens, in order to examine them under the microscope. Pictures do not always show the necessary diagnostic characters, and in some cases, only a dissection of the genitalia can provide a certain ID. And even in well studied invertebrates (unless they are extremely rare), I don't think that retaining a small number will have a devastating effect on the population.

I agree with Mr Marc Pollet. For species level identification (specially unknown and small taxa), we should collect the organisms.

Hi, I agree with previous reviews in avoiding the maximum collection of  animals. But because the identification of some species of rodents and bats can be very difficult (group of animals that I work), and the pictures do not help much with the identification, collect of the animals is necessary. Fortunately we have a zoological collection at the University, where they are deposited all animals collected. In this way the collection grows and collected animals can be available for other research.

Thank you all for responding to my question! It’s great to read all these opinions.

Here is my experience: I have been performing trapping of fruit-feeding butterflies monthly for 13 years now at a site in Uganda (see my recent Ecosphere paper). After identifying the first specimen we simply scored the individuals and released them back. The butterflies were so numerous that we didn’t mark them at that time, but a large sub-set was marked at a separate time (see my life span record paper in Experimental Gerontology), and recapture rates are modest so populations will not be affected too much and the forest is 700 km2 so we won’t make any species go extinct with this! The killing is a bit painful: ‘it’s a dirty job but somebody’s got to do it’, Mathematicians have declined to use my data as it is not possible to verify the identification of each and every individual, and yes mistakes have been made! It has been discovered that one of the commonest species is actually two species, but no way to go back and correct that. We also discovered E. kakamega only after two years. Some specimen probably passed for E. uganda before. Moreover someone asked me for a time series of certain species for stable isotope analyses. I wish I could help him with that! Now with sequencing becoming more rapid and cheap, collections are going to have applications that we can now hardly imagine.

Indeed, good curating is crucial and not easy. I worked as a PHD student in a department that later collapsed so any material that I left there in a freezer (probably poorly labeled) disappeared. I left a reference collection at the ZMA which later fused with Naturalis: let’s just hope it’s still traceable. Other stuff is at ABRI in Nairobi, others with Niklas Wahlberg and others for phylogenetics. Now I leave some specimen in Tartu. I don’t keep my own collections.

A further advantage of collecting is that you can hit the ground running: you can start collecting real data before you have become a specialist on all the taxa involved. Pictures can miss essential traits such as genitalia or androconia. You can later take pictures of the set specimen. Taking pictures in the field is very laborious, too, taking away valuable field time.

So I do recommend collecting with little restraint (don’t watch them die, they would die anyway in the jaws of predators, from starvation or disease, and now they contribute to science! Or as a colleague used to say: ‘it’s not an ethical, but an esthetical problem’), but the more easily recognized species can mostly be released if you are 100% sure about their ID, perhaps with marking if possible. So, invest in freezers, and make sure everything is labeled, organized, and someone is in charge for the longer term. When setting, first collect legs for DNA before relaxing the specimen.

Thanks,

Freerk

In my opinion biologists should only collect individuals if they are going to describe a new species and are sure that there are no individuals of that species on some museum yet. Otherwise, good quality photos and live measurements should be enough. There will be the time when no individual is sacrificed, not even to describe species. That will be the time when Biology takes a step further in respecting its essence: life.

I partly disagree with Raquel: for sure photographic records and identification is sufficient in many (but not all!) cases, but certainly not so for describing new species! For species description, molecular data/DNA sequence is insufficient: even microbiologists agree that until the species/strain can be cultured and physiologically/biochemically tested, it should be considered a "Candidatus" rather than a nov. spec. ...

Though in broad agreement with the 'do not collect' message, there are certain cirumstances where correct ID is so critical that a specimen is required. This is only justifiable, however, where the specimen is then lodged in a museum hosting voucher colections, where others can check the ID in the future. I cannot comment on Lepidoptera, other than to say they have suffered disproportionately from historic collecting, and it may well be that museums have excess specimens, but for those of us working with taxonomically difficult groups of Molluscs (Hydrobiids) or Crustacea (Mysids) in lagoon environemnts, where the correct spcies ID is of critical importance for monitoring, collection for vouchers can be justified. But the vouchers MUST be lodged in accredited collections.

Well, its really a good topic for discussion. I strongly opine that, the purpose of sampling and post-sampling plans (curation if smples are collected) must be clear in advance as also advocated by my friends Ajith Kumar and Jacqueline Beggs.

Reasonable collection is not wrong if it may serve a purpose. I know many fellows who have collected the boxes full of insects and then left them for decaying. I also endorse the words of Ajith Kumar that own ignorance of the taxa can't be a reason for making collections

So better be scientific in approach and practice.

At least for small mammals I justify collection of individuals. First of all, snap-trapping kills animal anyway. Without dissection, several characteristics (age, reproduction data) are assessed with error. After all, skulls also are used in variuos scientific purposes.

Some species cannot be identified without skull (or genetic methods). Genetics methods are expensive if used for big amount s of material. New species for science/territory require model specimens.

If species is endangered, then collection should be rigidly justified, of course.

I want to comment in two ways. First the practical part. As several colleages have appointed here, there are groups easier to identify in the field than others. Some are practically imppossible without disecting the animal. Therefore, you must be aware of the difficulties of your group and do what you need to do. Also collecting specimens also must be in function of the aim of your project. It is an ecological, taxonomic, physiologic, etc. work? That must also define the need for collection or not. Anyway, I agree that always must prevail a conservative criteria and collect only the minimum possible. And certainly all the collected specimens must be preserved in well established and recognized academic collections.

Now, a bit in a phyllosophical vein, death is part of life and there are always animals dying for different causes (predation, wheater conditions, disturbances, illness, etc.). Being collected is another one, with the difference that this one must be applied in a humanitarian way and will help us to understand the natural world. As long as we will not put in risk the population by overcollecting, I think in most cases is valid to take a few specimens, always well justified by the research project and bioethical considerations.

Release unless it is extremely difficult to identify by photographs alone, which is dependent on your ability to identify taxa. Dissection of the specimen may be necessary toaccurately identify the specie to aid in studies on biodiversity. But in order to protect local populations, it is important to avoid the option of collecting specimens as much as possible.

I afraid, that "protection of the local populations" in most cases is just over-priced. In the case you work with common species, like beetles, moths, dipterans - no threat if you collect them.

Collection of the species from Red data book should be justified (in most countries, special permits obtained from administrative bodies).

Last question - are collected specimens properly labelled, stored and available for other researchers/studies? If not, then collection is not justified for any purpose.

Dear Freerk, I think it depends on what your sampling is all about. Obviously, if you need to collect them for DNA, butterflies are so fragile and maybe you need a type specimen etc.. I have worked with the Mark-Release Technique with mosquitoes, but it is different. Also, there are different statistical analysis techniques in case you want to capture and release or draw out individuals out of the population.

Cheers,

JNI

This touches on the issue of bioethics especially if you'll be sampling for mammals. However, since what you'll be doing is only for identification then I strongly suggest that you release because there are no invasive procedures to be done and so after doing biometrics, it would be more appropriate if you just take pictures and let the animal live. After all, their main goal is for fitness, transferring their genetic material to other generations, and so by releasing, you give a higher probability of them to reproduce and increase fitness.

I too agree with those who advocate for quality photography!

JNI

Generally, I suggest that it is better to release specimens after they have been observed and their respective characteristics documented. However, in times wherein thorough investigation is needed or when the behavior of the organism is to be tested, collecting organisms could occasionally be used.

Have you considered identification with DNA barcoding? You only need a tiny tissue sample, which for most species would not cause harm. This gives even more definitive results than photos.

I disagree that barcoding will give "even more definitive results" - it will (maybe) at some point, but at current uncertainties overwhelm benefits from this approach in many taxa...

Release specimens alive depends of species and the collecting gears. With fishes, kind of gear, collection time, and depth are important to maintain alive a fish, some species are more sensible to the stress collection, including the quantity in the sample. The barcoding is important, however is first necessary to sample the fish, identify and count the characteristics, and take the picture, with this technique we need to sample all species again, because many of fish species in collections, are preserved in formalin . Usually, many of the samplings are realized in short time, collect in a small area, and at this time, is necessary to sacrifice the fish to obtain many of the information that we need.

Saludos

Jorge

Barcoding would work if and when a proper sequence library has already been built. In the first place, to build the library, iBol requires that a voucher specimen be deposited in a scientific collection.

Thanks, Juan: here we get back to collecting, of course - a voucher specimen requires collecting (at least once)...

Would be helpful to know the purpose of the sampling. Is it just to establish a measure of diversity? Or to actually inventory species and relative abundance? Or something else?

Yes, I also believe it's all about your goals. If you're doing ecological research, as for populations or communities, you'll surely have to keep them all. And if you're dealing with buterflies, maybe you can think about choosing your captures, but if it's a smaller or not so known group, it's impossible. I work with Syrphidae (Diptera), and must collect all individuals for community statistic purposes...

I think that there are at least three considerations that an investigator should evaluate with regards to whether to sacrifice an individual or not. The investigator must be able to confirm their work and the identification must be accurate. We preserve material for three basic reasons. These include vouchers, research, or training.

I work with fish and it is typical that we are taking vouchers for each watershed studied. I know that most of us have preserved our share of specimens, but in some cases the specimens from a routine study were needed a decade later to verify the identification. For example, when we experienced a large fish kill on the West Fork White River our state DNR indicated that they would restock all of the species lost. When they were asked if they would restock non-game species too they answered affirmative. So, a decade later the DNR wanted to look at all of the voucher specimens that I had collected from that watershed. We were able to confirm the identities of all of the non-game species that had been identified based on the museum collections.

So, if a study requires a voucher then 2-3 expired individuals (no more than 5) should be be those used first to satisfy the requirement (Hint: ideally they should not be crushed or have lost all coloration). I think that a project voucher (per crew leader) is sufficient and not necessary for every site. However, what if you encounter rare or imperiled species? Then photographs should be taken if the key characteristics can be adequately photographed. This requires that the crew leader must know what the diagnostic characteristic is for every species that they might encounter. The problem is that it is typically small individuals with either anal fin ray counts or scale counts that are the needed diagnostic trait. This is usually beyond the capability of most field applications, plus by the time all of the handling has occurred the individual may have expired anyway. I personally would not consider a photograph of a rare specimen to be sufficient for definitive identification. Unless a specimen is available for verification this would not provide conclusive evidence that the investigator had collected the species in question. We do not use that information for management purposes so it is always questionable. I personally would rather have a single individual preserved for verification. In most cases the specimen is actually another closely allied species that looks similar. Without the specimen it would be impossible to verify thus causing additional concerns or requirement to conduct another visit to the site.

Second, many of our studies are long-term studies of watersheds. We have preserved a portion of the individuals from each site over the years. The benefit is that we often will look back into the museum and pull specimens from different time periods to identify changes in sex ratio, age and growth, gonad health, and diet. All of these types of studies can be performed with preserved material and it is often provided conclusive opportunities to confirm our suspicions from the past hypotheses with better and more conclusive technological advances. These specimens have provided important indicators of changing environmental quality. Preserved material has other intrinsic and valuable uses. For example, preserved material enables us to study variation of species. We are currently working on fishes from Indiana and Ohio. By having 30-50 individuals in the collection we can often pull enough material from several sites within the same county or local area that we can often put together a large amount of information that would assist us in the conservation of the species of interest.

Thirdly, we are training and retraining professionals and students in identification. I realize that professionals should be able to identify what they are collecting. But, the truth is that many professionals are out of practice or may be unsure of themselves. Some have not seen many rare species or may not have been trained for identification in all cases (i.e., trained on east coast and working in the southeast). Since most professionals will collect their samples during the summer and then process samples in the winter so this requires knowledge of species "fresh coloration" and also "bleached or faded" preserved coloration. Summer sampling and winter processing is the most expeditious use of time. It enables local, state, and federal agency biologists to complete more studies and assess more streams than standing on the bank and identifying every individual. Often, if the "identify all individuals in the field approach" is required -- this will lead to many mis-identifications since flopping fish will not enable a crew leader to count anal fin rays on small individuals. By preserving small individuals and getting the larger specimens back into the water we increase our confidence of the identifications and this is a requirement for a good quality assurance program. I would want to have vouchered material for every crew leader to be sure that they had adequately identified the species. We have heard of professionals misidentifying as much as 80% of the individuals from several studies. To not have preserved material means that large amounts of money are wasted and we are unable to provide a decision on the questions being investigated. Large species and those that are not questioned can be photographed and returned alive to the water, others may require more intensive laboratory analysis to get reliable, accurate, and precise identifications.

For students, I agree with returning everything back to the water alive; however, fresh material is invaluable for students being trained in ichthyology, aquatic ecology, limnology, or fisheries biology. To not have fresh material is a problem since students cannot look at the internet or drawings and be expected to accurately identify live material. Most fish are not perfect and variation does exist. Students need to learn what to look at and this can be done using museum specimens... but specimens with broken fin rays and no coloration is not adequate for training for field processing (and that is the point of training in the field).

My philosophy has been that some of the individuals may need to be sacrificed for study so that others can live. These individuals should be placed into a research or museum collection so that others can benefit from the collections. Unfortunately, too many of our repositories are unfunded and not valued for the wealth of information that can be easily gleaned. By retaining material in a repository a significant amount of money can be saved in the future if the material is maintained and the collections are used for research and not just as vouchers.

I believe collection is inevitable if extensive studies are to be done. But sampling activities should be regulated if possible to prevent unnecessary pressure on the local population of butterflies.

I believe that the entire process depends on the objective of the study. The objective can either be to prepare vouchers/specimen with respect to species composition and presence, or it may be for DNA research. In any of these cases, may be 1 or 2 specimen can be sacrificed, if really unavoidable.

Status of the species to be collected is also very essential. For e.g. certain species which are critically endangered, etc., should not be collected as per the laws and such collections fall under punishable acts.

As far as training about species identification is concerned, I would suggest that voucher specimen from other institutes should be requested rather than sacrificing a new individual. After all vouchers are for such purposes only, aren't they?

Depends on what sort of creature you are dealing with as well as its natural population size. Our aim is not to disturb the natural population afterall.

Several small populations of rare plants disappeared after the collecting by numerous botanists, everybody interested to have this species in own herbarium. I can not support such attitude.

Dear Freerk

knowing your work, In theory you've got to collect...as you've got to be sure that you will not count several times the same individual, but that's rather destructive and may be dangerous for some populations of rare Nymphalidae.

How to avoid this problem.

For identification, no problem...as mentioned by many contributors good pictures (both sides) are most of time sufficient (except may be for some little black Charaxes or difficult Satyrinae).

To avoid to count two times...well finally to mark all your collected individuals before releasing is may be the best way as furthermore you will have an idea (if your traps will stay for a long while in the same zone) of the probability of recollect with traps (and so a kind of rate of dispersion) but also, something which is interesting for you the longevity (at least an evaluation of it) of the species.

Hoping it helps

Cheers from Luc

The first thing is to maintain close contact with a specialist in the taxonomy of the group in question. He will decide for capture or not. In nature there are many species that are similar, their separation is often difficult and a bad identification is likely to produce bad publications and unreliable data. If, however, species are easily recognizable - but this sets the taxonomist -of course it is advisable not to catch too manyspecimens, especially if the species is rare or endangered.

Since the question deals with butterflies, I think I can be of help here: my father used to collect butterflies outside our home for 40 years; he used all kinds of baits and attractants. He took freshly emerged males, mostly, generally avoiding females and battered individuals. Regardless of his collecting, every year, or indeed, every season, there were still the same number of butterflies, if not more. We never thought about it in childhood, but in retrospect, it was because the habitat was well protected. Based partly on his work, I made a list of 243 species of butterflies recorded on our forest estate. Today, repeated forest fires, urbanization and unregulated exploitation of the forest has resulted in an all time low not only in terms of numbers but also in terms of species. Based on this experience, I can categorically state that in the case of butterflies or other invertebrates with several annual generations, collecting a number of individuals regularly does not endanger the population, especially if those individuals are males. What counts is that the habitat should be adequately protected. Naturally, this does not apply in the case of vertebrates.

Therefore, different standards apply to different groups. With regard to the majority of insects, barring N. America and Europe, I urge researchers to collect and improve their collections, for it is on reference collections that research rests. India has about 40% of Lepidoptera represented in national reference collections- naturally we are hamstrung when it comes to describing new taxa, or even trying to identify what we manage to collect. The application of vertebrate conservation strategies to insects has led to disaster before the advent of digital photography, since no one would admit to having captured a 'legally protected species'- there are no endangered insects, only endangered habitats. The unit of conservation is not the individual butterfly, which is but the final, short lived stage of the insect, but the habitat.

I agree with previous answers; context is critical. However, I would quibble slightly with Peter Smetacek over the distinction between vertebrates and invertebrates (elephants versus butterflies, yes; fish versus crays or stag beetles reverse the equation).

If you are doing an ecological study in a taxonomically well documented area, such as in Europe or North America, and photos are sufficient for identification then it may not be necessary to collect vouchers. For some taxa over-collection, or any collection, is a threat to populations and species but from a global biodiversity perspective these are a relatively small proportion of cases. As Peter suggests, it is often zealously overemphasised relative to other threats - including the animals that you accidentally killed en route to the site. If the research is beneficial, particularly for habitat conservation, education or biodiversity policy, then collection may be well justified.

If you are working with taxa that are difficult to identify or in areas where the biota is poorly documented then vouchers may be essential. I work on systematics of small montane vertebrates in Southern Africa, a moderately well documented area. There are still enough poorly sampled areas here, however, that genetic analyses often demand a taxonomic review. Photos, measurements and other field notes are essential but they don't replace a voucher - and multiple validations of identification are needed. Several times we have been caught with our systematic "pants" down - sequencing reveals an divergent population from a site that was difficult and expensive to access, without an accompanying specimen. We have had to go back, re-sample and collect frogs, fish, lizards and chameleons. In some cases we did have vouchers but were missing critical life stages.

In really remote, biodiverse areas of the tropics collections are essential. Work on Malagasy frogs by Miguel Vences and others has shown an extraordinary level of microendemicity, convergent evolution and site to site variation in appearance that could not be untangled without collections. The same could be said of any other tropical biodiversity hotspot and most animals weighing less than 500g within them. In most surveys our capacity to collect and preserve specimens is limited, particularly for larger animals. Documenting a poorly known biota creates interest in preserving it, and habitat clearing and transformation is a much greater threat than collection.

Even in situations where the target taxon has a very narrow distribution it would be very odd for collection of a modest number of individuals to constitute a threat, particularly for species with high fecundity, in near pristine habitats with little or no former collecting. Common sense should rule. Take responsibility for your collections. Make sure that your study is justified with independent peer review of ethical concerns for species where this is warranted (long-lived, slow breeding, small populations). Take adequate but modest collections, generally leaving some observed individuals untouched. Ensure your specimens are well documented, well preserved and available to others, preferably in public repositories. Submit your species occurrence data to biodiversity databases so that your vouchered records get wide circulation.