Representativeness of faunal studies

In my opinion, the breadth of coverage is more important than the sample size. Well, collect a huge sample sitting in one place and using only one collection method. What, will that be enough?

So that's what I'm saying. How can I determine how fully the fauna is selected (or, in other words, whether the coverage is sufficient)? Is it evenly selected enough to compare different biotopes?

Sometimes such questions arise, namely:
I think about "rare" species, many of which are not at all secretive, but are rarely found and in small quantities. After all, we know that a species cannot exist if it is bred in very small quantities in very small territories. But if we find single copies. in different places, so there are enough of them to develop here and there.
In such cases, for some species, their number is given an idea of the fees in traps. Well, again, what idea do a dozen traps filled with some kind give us? Tipo opnyatiye - very much.. and that's it? And how many have not yet been caught, and how many are still in the state of larvae or eggs? How much will it be next year?
Or, for example, Dorcadion fulvum, which I catch every year. Each year, it appears in different parts of the same field and in different amounts. In some years, it runs out in hundreds, in others it is difficult to scrape together a dozen, and yet this small number for the next year gives a huge generation, and it is already concentrated in a new place..

Or, for example, Dorcadion fulvum, which I catch every year. Each year, it appears in different parts of the same field and in different amounts. In some years, it runs out in hundreds, in others it is difficult to scrape together a dozen, and yet this small number for the next year gives a huge generation, and it is already concentrated in a new place..

Oh, he's a rascal...

Admittedly, I expected more responses. Does no one care about the problem, or is it too insignificant?

This post was edited by kovyl-05.04.2010 18: 12

It's too complicated!
In addition, you said in the first post (about Pesenko's work): "But it's very difficult there," and in order to have a fruitful discussion, we need to move on to something more complex than Pesenko's.
Define concepts... the fauna is not all clear. Next, it will be clear what to disassemble.

It is better to deal with Pesenko (with a pencil in his hands, for clarity) and find a critical review (on his book). It was in one of the magazines, but I couldn't find it on the spot (although I made a copy once).

It would be useful to go back and re-read the discussion in the topic about faunalism
http://molbiol.ru/forums/index.php?showtopic=228815&st=150
and in particular, the detailed post of K. Makarov:
"...About faunistics
I envy botanists. They have floristics-a respectable field of knowledge, clearly separated from phytocenology, geobotany, etc. (I suspect that botanists have a different opinion – but from the outside it seems so). In zoology, the concept of faunistics is relatively vague, and its key concept – "fauna" - is interpreted widely-from "a historically formed set of taxa with similar parameters of their ranges" to lists of species and indications of individual finds (and there are also "zooths", etc.). Ambiguity is a common phenomenon in biology (for example, "a large number of taxa with similar parameters of their ranges").forest " means both a biotopic characteristic, and zonal and belt) and this does not bother, as long as the understanding of such a term is unambiguously determined by the context. Unfortunately, this circumstance was ignored in the formulation of the voting question. I gave my vote for faunalism in the traditional sense, but what about other colleagues?
Faunistics, claiming to be a science, should study the meaningful connection of phenomena; the discovery of certain species in a certain territory is only a small part of this science. Rude (but hopefully permissive) exaggerating, we can say that registration is the fate of the device (in physics it is easier, in biology it is more difficult), and the task of a person/scientist/specialist is data analysis. It is strange that the discussion very quickly went the way of discussing the importance of" registrations " - who argues, devices are needed. But I would be surprised at a specialist who deliberately limits himself to the "instrument" level.

Returning to the ideas about fauna, I have to envy the botanists again. They have a well-developed conceptual framework that makes it possible to distinguish ideas about regional flora, specific and local flora, flora of an elementary area, and so on. In zoology, the term "concrete fauna" was introduced by Yu.I. Chernov (1975), but in the course of its application, various interpretations and ambiguities in distinguishing specific and local faunas quickly accumulated. And this is despite the fact that the term has a very limited application – special works on local ground beetle faunas can be counted literally five. For the sake of clarity, it is reasonable to continue the analogy with floristics, distinguishing between a specific fauna (as an elementary and unique lower unit of the faunal system) and a local fauna-as a faunal sample (i.e., a list of species of a certain area – 10 or 100 square kilometers, etc.). Such lists, provided they are relatively complete, are very useful, even if They are represented by a trivial set of taxa. But: we are talking about a list of species of a small, if possible, naturally allocated territory. The questionable scientific value of lists for administrative allotments (especially large ones) is a frequently expressed opinion. Of course, there are some benefits from such lists – for example, for regulating the use of natural resources, etc. But their meaningful interpretation is complicated by the artificiality of the allocation, and the variability of administrative borders is a source of numerous misunderstandings.
example. Serebryanoprudsky district formerly belonged to the Tula province/district / region and only since 1942 has been part of the current Moscow region. The fauna of this southern area significantly complements the lists of animals of the Moscow region. For example, in 2002 N. B. Nikitsky collected Carabus sibiricus haeres in this area (Nikitsky, 2003). An outside observer who records changes in regional lists may interpret these data as a result of the expansion of the species due to climate change or a reduction in the area of cultivated land, etc. It is hard to expect a specialist in global climate warming to be caustic about details (for the Tula region, this taxon is known), and once "in the clip", the example will wander from publication to publication.
The attitude to the lists of local faunas, generally speaking, is also far from unambiguous. At one time, I was thinking of publishing a list of ground beetles in the vicinity of the Istra reservoir. But at first perfectionism interfered (I overlooked it there, I didn't collect it here), then doubts accumulated (every year, the list changes, and you don't immediately understand why: either the climate has changed, or you have learned to catch) and only later realized that the idea is fundamentally flawed. Long-term collections do not specify the characteristics of local fauna, but "blur" it with random finds. Intensive (all types of traps + manual collection) study during one season (preferably all year round) would be correct. Then we get a relatively complete list, clearly linked to the current climatic and biotic situation, and which seems to be quite suitable for the role of a unit of faunal analysis. In 2006/7, at the initiative of A. V. Matalin, we conducted such work in the Elton region – the results were very interesting and instructive.
Faunae and finds
A common variant of faunal studies, the novelty of which seems to be indisputable, is the marking of species outside their known range. There are, however, two considerations in this regard.
First, we do not always represent areas correctly (i.e., our representations are mostly inaccurate). Usually, the range is described on the basis of a number of marks and some ideas about the biology (!) of the species: The taxon is related to coniferous trees, their locations X and Y are known, both lie in the taiga zone. Therefore, we will consider the taiga zone between X and Y to be the range, unless otherwise specified. In this state of affairs, the significance of a new find is not easy to assess, especially if the biology of the species is poorly understood.
Second, we usually underestimate the migratory capabilities of insects. Recently, it was shown (Feng et al., 2007) that even relatively weak "flyers" - ground beetles Harpalus, choosing a flight altitude with a suitable wind, can fly hundreds of km per night. I remember very well how in my youth I used to catch the light in a Pharmacy. With touching care, he took the lamp far out into the sand, plugged it into the weather booth – and spent half the night trying to see the psammobionts among the thousands of water beetles and beetles that were falling on the sheet. Everyone can tell a lot of such stories.
example. In this aspect, one of the results of our work in the Elton region is significant. From about 50 thousand copies. ground beetles collected during the season, a little more than 10 thousand accounted for the share of Harpalus rufipes (95% were caught by soil traps). However, no oviparous females and only a few sexually mature males were found among these individuals (generative phases of this species are usually considered as well as others). Consequently, all the mentioned 10 thousand were migrants who moved from point A to point B mainly on foot (the distance between the extreme reference lines exceeded 30 km). How long did each individual stay in this area? Did you eat? Could it compete for resources? I.e., can we consider these migrants to be members of the community/local ground beetle fauna of the Elton region? Not only does a simple mark not provide an answer to these questions, it doesn't even allow you to put them – but there are still questions.
What does the statement "type C-us found at point D" mean, then? Yes, anything. Its interpretation depends, in particular, on the knowledge of bionomics, and here in many cases (probably, it is more correct to say-in most cases) there is nothing to brag about.
These considerations, in general, are trivial and lie on the surface. They regularly occur to specialists and develop a professionally cautious (in full compliance with Occam's rule) attitude to various kinds of finds. To compensate for the multiplication of finds, it was even proposed to eliminate the exact labels – which, perhaps, is superfluous.
Since ideas about the extent of animal ranges are often extrapolated, it is appropriate to ask the question: what is more important-finding a species at a new point or finding out its preferences, ensuring the correct interpretation of marks on the map?
Faunistics and bionomy
When discussing the composition of any fauna, we explicitly or implicitly take into account its history, environmental compliance with the requirements of a particular species, and interspecific interactions. Questions of faunogenesis are often speculative, while other parameters are more accessible to study. Having found out (at least in the form of a so-called expert assessment) the biological features of the species, we can economically describe its range with a minimum of finds (for example, "the forest zone of Eurasia with the exception of the northern taiga"). Putting dots on a map is not an economical process, and an attempt to interpret the resulting picture will inevitably raise the same questions about the bionomy of the species.
example. In connection with the development of biological control methods, the biology of Lebia grandis feeding mainly on preimaginal stages of Leptinotarsa was studied. L. grandis was described from the North. North Carolina in 1830 and was considered a local species. During the 19th century, an expansion of the range of L. grandis was recorded, as it turned out – due to the spread of the food object-Leptinotarsa juncta. Since the 1870s, the ranges of L. juncta and L. decemlineata overlap and the transition of Lebia to feeding on the Colorado potato beetle, although the preferences of the original host have been preserved (Weber et al., 2006). Now, knowing the features of the biology of this Lebia, we can more accurately assess the range and, in part, predict its dynamics. In addition, this knowledge increases the factual base – understanding becomes available not only reliable finds of the species, but also its reliable absence (if there is a host).: here you can think about the role of the photoperiod, the sum of temperatures, predators/competitors, etc.
Obviously, we will not be able to do without accurate findings, but we can only understand them and assess their range based on the bionomy of the species. The process of filling the map space with dots is endless and meaningless; studying the biological features of a species and extrapolating these concepts seems more rational.
I would very much not like my point of view to be perceived as orthodox, denying the significance of finds, exact bindings, etc. This is part of the original faunal data and cannot be dispensed with. In addition, there are good cases where the very fact of finding already means a lot – for example, if we are talking about amphipacific areas or transatlantic disjunctions. However, in my opinion, it is not correct to identify the factual base with faunalism. In this context, the following question is clearly indicated: who does faunistics? Faunalists – who are they?
Faunists and taxonomists
I tried to remember the faunists "from ground beetles". The top five are almost obvious-Goldhouse, Lindroth, Darlington, Kryzhanovsky, Irwin. Although there are hundreds of publications for each of them, the subject range of articles and monographs is largely similar: some early articles by M. B. are devoted to interesting finds (not always), but the bulk of them are more or less private taxonomic studies (+ determinants). Large generalizing faunal works are usually published in adulthood. The general trend is quite clear – a "faunist" takes a long and difficult time to form as a specialist, accumulating and comprehending facts. Naturally, these facts also include findings/marks, but the ideological component is the area of broader extrapolations. This is also evidenced by the fact that all these faunalists (except, perhaps, Lindroth) did not limit themselves to ground beetles, analyzing the distribution of different groups of insects. The most significant example is that of Darlington, who built his Zoogeography almost exclusively on vertebrate examples , since data on the distribution of insects are difficult to interpret.
I am not sure if there are any significant faunal works in the field of entomology written by non-taxonomists. A more or less close example of this is the summary of ground beetles of Holland Turin (2000). But this is rather an exception: the territory is small, has been studied for a long time, and the author's generalizations relate mainly to ecology, not faunalism.
The next level of faunistics is various regional lists, usually compiled by taxonomists, less often by specialists in related disciplines or amateurs. After all, there are quite a few works, and they are also accompanied by major summaries/determinants (these genres can be difficult to distinguish) for regional faunas. This category of work is certainly important and useful, but with regard to ground beetles, this level of study did not give any noticeable generalizations.
Finally, works on local faunas (not to be confused with lists of arbitrary small territories) are still few in number, but a number of generalizations based on them seem promising (for example, the works of L. A. Kropotkin).Penev's guide to local Carabus faunas). In this area, however, the contribution of ground beetle specialists is surprisingly small – progress is provided by individuals who have developed skills in analyzing local faunas on other groups of arthropods.
Simplifying and coarsening the picture, we can say that in the considered works faunistics as an independent science reveals itself twice: at the level of large generalizations (macrofaunistics) and local faunas (microfaunistics). "
derivatives of taxonomy in its various forms. A typical "faunist" is thus an erudite taxonomist.

2 Adamov
> It is better to deal with Pesenko (with a pencil in his hands, for clarity)
Who-n. have you figured it out enough to be able to answer my question?

> Define concepts... the fauna is not all clear.
In this context, I mean a list of species for a specific territory.


2 Proctos
> It would be useful to go back and reread the discussion in the topic about faunistics
Once again read the topic "Is faunistics an important aspect of entomology?". But that's not what it's about. However, something flashed there as well. I'll ask for Makarov's article:

K. V. Makarov, A. V. Matalin (Moscow State Pedagogical University)

Problems of studying the fauna and population of ground beetles (on the example of the carabid fauna of the Elton region)

Well, while it is completely incomprehensible to compare the composition of their fauna with the faunas in other regions. First, the data here and there may be incomplete; second, why do this at all? Compare the middle zone with the "number of faunas of Eastern Europe" (in particular, the north-Kivach, with Karadag). You can, in principle, compare it with Australia… What does this mean?

2 Adamov
> It is better to deal with Pesenko (with a pencil in his hands, for clarity)
Who-n. have you figured it out enough to be able to answer my question?

No, of course not…

It is necessary to understand not for the answer, but in order to understand.

It is impossible to get a satisfactory answer without Pesenko, and probably not with his work.

> For what period? From the origin of the earth? In the last 1000 years..., 100 years, ... 10 years, ... 1 year?

Well, let's say for 10-20 years.


> Species encountered, feeding (part of the cycle, or all of it), reproducing, sometimes appearing to "commit" the above, etc.?

Let's take the met ones for simplicity, although this, of course, is not entirely correct.

It would be good to somehow justify this framework. Otherwise, you will call fauna "this", and other colleagues-something else, and the headlines of articles all start the same " Fauna..."

This post was edited by Adamov - 08.04.2010 19: 07

The rationale is more complicated.
Well, perhaps so:
I consider the time period to be more or less sufficient (with a good collection intensity and knowledge of collection methods) to identify 75-80% of the fauna. In a year, it seems to me, the fauna can not be identified, even if you collect everything that comes across every day. Well, if only belyanok in the country to collect. And that's not a fact. Of course, you can set a bunch of other reservations (the size of the group, the lifestyle of its members, the size of the territory, the natural area, and much more) - and all this will affect. Sometimes radically.

Why the encountered species? Because the role they play in the community cannot be determined in a short period of time. Maybe they have the border of their range here, maybe they are migrants, maybe ... Everything is probabilistic. Accordingly, the smaller the sample size and the time it took to select it , the higher the probability of getting incorrect data. And in 10-20 years, there is already a chance to even get into the waves of life.

All this is my imho, based only on the intuitive experience of ~20 years of fees.

This post was edited by kovyl-08.04.2010 23: 36

Why can't you catch it in one year? Much, of course, can be determined by the group (volume and biological specifics) and the qualitative and quantitative characteristics of the territory.

Therefore, now they are increasingly talking about"local faunas". As a good example article:
Makarov K.V., Matalin A.V., 2009. Ground-beetle communities in the Lake Elton region, southern Russia: a case study of a local fauna (Coleoptera: Carabidae). S.I. Golovatch, O.L. Makarova, A.B. Babenko & L.D. Penev (Eds). Species and Communities in Extreme Environments. Festschrift towards the 75th Anniversary and a Laudatio in Honour of Academician Yuri Ivanovich Chernov. 2009. P.357-384.

(you can download it here: http://www.carabidae.ru/index.php?name=Library&author=76 ).

But I haven't yet figured out what exactly to call a local fauna. I especially did not understand the part of the question concerning the" amount "of territory for the "local fauna". I suspect that this question is solved only for specific cases and in an abstract form is not strictly defined and is not solved (yet).


In the mentioned work, ground beetles were collected for 1 full year. As I understood, there were quite large samples of traps (traps worked continuously) and additionally used manual collection and light traps.
The "role" of a particular species in the community of the study area was determined using specific methods for determining the generative state, age, etc. And appropriate conclusions were drawn.




What groups and territories are you interested in? What is it all about? If you knew that, it would be easier.

The post was edited by Adamov-09.04.2010 00: 27

I read Pesenko's article " Determining the total number of species in the local fauna (or flora)". Everything is fine, everything is clear. Although, in my opinion, it should be called "Some aspects of the study of the behavior of the normal distribution function on the example of biological objects". Or, if Pesenko really wanted to use the terms that he himself used in the title: "An approximate definition of the total number of species ... based on the assumption of representativeness of the sample." The author cites conditions, the non-fulfillment of which casts doubt on the expediency of all manipulations. It is extremely difficult to meet these conditions.

"This pattern is evident only in large randomized collections (or in the survey of fairly large areas), when more than half of the species of local fauna (flora) are taken into account.
...
In conclusion, we should note the main conditions that must be observed when processing faunal (and floristic) materials in the manner described above: 1) the randomized nature of collections (registrations) or, in any case, their planning and implementation in such a way that the data obtained on the abundance of different species of the studied group of organisms correspond as much as possible to their abundance in nature; 2) a sufficiently large volume of materials (for most zoological objects and collections in large territories) that they contain more than half of the species local fauna (or flora)."

To comply with the first one, the fees should not even be randomized, but total, because only they will somehow reflect the real abundance of species, which is rarely done in real faunal works. By the way, Pesenko himself writes about this:

"Let us illustrate the described sequence of calculations by the example of determining the total number of single bee species (Hymenoptera, Apoidea) in the Lower Don fauna based on the materials of the author's 7-year collection of almost 23 thousand individuals obtained under a program close to complete randomization. In various stations and throughout the entire season, all the bee individuals found on the flowers of 215 species of the most common entomophilic plants were caught."

Imho, it somewhat confuses randomized (random) and total fees. Even so, it is difficult to say at what point in time the sample can be considered representative (see condition 2). Or you need an additional mathematical tool for such an analysis.
The second is even more difficult to observe. Who can guarantee that there are already half of the species in its collections? Or again-an additional mathematical apparatus.

In general, it is difficult to understand where the concepts of "local fauna" and "significant territory"intersect. What should be the maximum size of the territory so that its fauna no longer fits into the concept of "local"? Or what should be the minimum size of a territory with local fauna to be "significant"?

When I asked the question that was put out in the first post of the topic, I thought in a slightly different direction. I was interested in what factors determine the addition of new species to my collections. New species are those that I have not previously captured for the territory. Interim results were summed up by year. After analyzing my not very large and not fully processed fees, I saw an interesting picture. The intensity of adding new species depends on the number of days of collecting (that's right, the more days you catch in a year, the more different species you catch, which means that you find more new ones among them). It also depends on the number of species caught (for the same reason). Another thing is that I do not catch as intensively as I should, and year after year it is not necessary-it is thick, then empty. Moreover, this pattern was similar both for the group as a whole and for some large individual families.
This allowed me to formulate the following hypothesis. If the graph of the function of the number of species by year over the period of research is increasing (it is possible with insignificant, statistically insignificant fluctuations), and the graph of the function of new species has an extremum point where the sign of the function changes from " + "to" -", and the graph of the function begins to decrease statistically significantly, then from this point on, you can- then try to analyze the resulting curve and predict when it will end. That is, the researcher catches more and more species per year every year (the experience is growing), and the number of new ones at some point begins to fall steadily (logically, there are a finite number of them on the territory). Based on the angle of inclination of this branch, you can try to make some forecasts. And here, by the way, we can assume with a fairly high degree of confidence that half (or even more) of the species are already collected.
This approach also has its own pitfalls. For example, if you catch only bull-moustaches and motley-moustaches that fly during the day, you can at some point assume that there are no more new species. When you start catching incandescent lights, you quickly realize that this is not the case. Starting to catch on DRL, you are convinced that the incandescent lamp also caught far from everything, etc. In other words, we need to be sure that we use all fishing methods, in all biotopes, in different seasons, etc.
In addition, there may be many extreme points on the graph of the number of new species (as well as on the graph of the number of species per year) due to fluctuations in its behavior. Where a steady decline will begin is unknown. There is no function equation.
Summing up all the above, I will say that this is why I do not really believe that it is better to intensively fish for a year than for several years. Allegedly, this "blurs" the composition of the fauna with random finds. Maybe only for some groups and in very small territories…

And more. What we need here is a mathematician who can help us critically analyze all these guesses.

I will read Makarov K. V., Matalin A.V., 2009. Ground-beetle communities in the Lake Elton region...

About specialization: I am interested in Lepidoptera of the Volgograd region.
And I asked the question in my first post in the context of dissertations in various fields.

> You have answered almost all your questions yourself

No, unfortunately, I only outlined them. After all, there is no mathematical apparatus to answer them. And the people (returning to the first post) famously compare faunas, relying only on empirical experience (flair).


> It is necessary to collect not" on a case – by-case basis " and by various methods, but in the case of butterflies-all year round.

Well, in winter-except in the apartment


> But data on the abundance of species (for all butterflies) is unlikely to count.

If you take into account my hypothesis, then this is not necessary.


> At some point (a certain year of research), there may not be any new species, but since the research is stretched over time, and during this time, for example, the climate may change, and with it the boundaries of the species ' ranges, new species will appear (not found before), and so it can be your whole life.

This, of course, is true. But an asymptotically approaching abcissa curve of new species would already please...


> I think that over time, there will be a fashion for exploring smaller "formal" territories. For example, by dividing the map into a grid with 10*10 km squares. Analyzing the faunas of these squares, they will draw conclusions about the mutual location of local faunas.

I would rely on the biotope when determining the boundaries of a territory for research of local faunas. Because a square of 10X10 or even 1X1 km can capture, for example, the floodplain along which the species came from the Volga Delta and the desert, where there are desert species. How local will it be then?


Something people in the topic do not speak out. Is the forum unpopular? The topic is not interesting?

This post was edited by kovyl - 14.04.2010 18: 01

> You have answered almost all your questions yourself

No, unfortunately, I only outlined them. After all, there is no mathematical apparatus to answer them. And the people (returning to the first post) famously compare faunas, relying only on empirical experience (flair).

What kind of mathematical apparatus is needed (for solving what problems), ... I probably missed something?
In my opinion, bio-logical comprehension is more important now, and only then everything else.

2 Adamov
> What kind of mathematical apparatus is needed (for solving what problems), ... I probably missed something?

I mean, if you put the question the way I asked it - to determine whether or not the fauna is completely selected (or how completely) - there is no solution to this problem yet. Suppose we split a sufficiently large region into 10x10 km squares. On the territory of the Volgograd Region, for example, there will be more than a thousand such squares. How many squares can you really catch during the year if you use Pesenko's method and catch everything quantitatively? Well, if only traps are set everywhere... As you yourself correctly noted:
"But data on the abundance of species (for all butterflies) is unlikely to count."


> Determining the number of species in the local fauna (theoretically) is a bit different from the task of getting a list of species. To solve the first problem, you need to have data on the abundance of species, but to solve the second one, it is not necessary.

It turns out that judging whether the list of species is complete or not is possible only on the basis of data on the abundance of species (and you still need to be sure that half of the fauna species are collected). And what other methods can be used to find out - it is not known.


> I would not touch at all, when discussing "representativeness of samples in faunal studies", the issue of"abundance of species".

I wouldn't touch it either. This question concerns Pesenko.


But I would define "local fauna" (meaning primarily ground beetles) as the fauna of the territory in which it (the fauna) is fundamentally homogeneous per unit of time (for ground beetles – 1 year).

I believe that the ground beetle fauna of such a region as the Volgograd region is fundamentally more or less homogeneous throughout the year. But this hardly gives grounds to consider it local. I suppose, to tell whether the fauna of an arbitrarily selected region is local or not, you need to: 1) catch all (?) species in it and all (?) species outside its borders; 2) if the sets of species differ statistically significantly, we can assume that the fauna of the selected region is local. With this approach, it is possible that moving the boundaries of the selected region in one direction or another does not change the locality status of the fauna of its region. I.e., it either remains local or is still not so. Then you can: 1) repeat the cycle with moving borders; 2) narrow the scope of the analyzed group.


> You can approximate it in 1-2 years, or even in 100 years. The first one makes more sense, since it is a full-fledged "snapshot" (similar to a photo), and the second one is also a snapshot, but a moving object with a long shutter speed (the difference is only in the "shutter speed". And if "full-fledged images" will give something, for example, when comparing for different years (decades), then it is doubtful to compare" long-exposure images " with anything productive.

I think that in principle the fauna does not change in 100 years. Well, a couple of dozen species will increase or decrease, or a little more. This will be much less even than the "collection error" (i.e., the number of species that were simply not detected).


I would like to emphasize once again that I was particularly interested in this question, since people compare the taxonomic composition of faunas in completely different regions, without being able to prove either the representativeness of their list or the list of the region being compared. Why they make this comparison is also not entirely clear to me, but this is a separate question.
On the one hand, if approximately the same number of species are caught as in neighboring regions, we can say that the sample is representative and reflects the composition. But on the other hand, if there and there catch only 20 species? It seems to be the same, but it is clear that this is not all. Somehow everything is at the level of empiricism, there is no mathematics. And the comparison is carried out using complex methods, clusters are built...

This post was edited by kovyl - 19.04.2010 21: 41

 

2 Adamov
> Knowing the volume (approximately) of the local fauna, I do not cause any contradictions. For example, for local ground beetle faunas in the Rostov region, I would limit the volume to 220-250 species.

But this is an empirical experience that does not rely on any particular theory. a mathematical base? I (and any experienced amateur) can also roughly name the number of species in my group in my region. But this is also an empirical experience, because there is no proof-based mathematical framework. The fact that there are some works on various parts of the RO is fine. But how do you know how complete the lists they contain are? Do you have an evidence base? Yes, similar values (number of species) obtained in different samples (different regions of the PO) by different researchers suggest the total number of species in the PO. But only to assume.
You see, statistics mainly deal with the variation of a certain value in the objects of research. Just take the right number of these objects and data analysis is already a matter of technology. Here, the problem is somewhat different – the number of objects themselves is unknown. It's a pity that mathematicians don't come here, M. B., they would correct me. At least, those mathematicians whom I addressed gave such an explanation and said that in this form the problem is unsolvable. But it turns out (again, I return to the topic of my dissertations) that the faunas of regions (this can be considered a sign of an object that has a value) are compared on samples whose representativeness has not been proven. This leads to the question of the validity of applying all those methods, the masterly possession of which is so brilliantly shown by dissertants. A colossus with feet of clay…


Indeed, the average volume of local faunas is an empirical value. There are very few works on the ground beetle fauna of small areas of RHO. They are characterized by a large sample size, a duration of observations of several years (up to 10), and in addition, they are located in different parts of the region. As a result, the volume of these faunas falls within the specified interval. For the region as a whole, the figure is approximately 60% higher… By the way, this indicates a "fundamental heterogeneity" of the regional fauna.
If there are no such well-studied territories, then there is nothing to rely on. You just need to take and use all the methods to study the selected territory in detail . In the course of the study, it is possible to determine the "effective sample size" (trap samples, for example), as well as the number and quality of the studied sites, and after at least 2 years, it will be possible to judge the "required duration of observations". When all these "criteria" speak in favor of the approaching completion of the species list, you can use the Pesenko method.

You see, you've also started to think that you can't pick out the fauna in one year.



> A long-exposure photo of a moving object is also not fundamentally different from a short-exposure photo of the same object. But what's the point?

Well, that's something else entirely. Meaning, by the way, can also be.

This post was edited by kovyl - 21.04.2010 17: 20

  

in my opinion, in this case, it shows something, only ex. There is - there is, no-means either no, or not found yet.

How can you calculate how many beetles are running around in a certain place? As already written, in one year of the same fulvums, hundreds, in another a few things, in a couple of years again hundreds.. I.e. either larvae develop for a long time, or they have periods, or they are sometimes secretive.. How should this be interpreted? And then how to count their number? And what are the periods of mass brood and how many fertilized females do they occur? Mass broods whether there are strict periods or not, and why in some years the number is so small.. And this is only for one type, and there are hundreds of them..

In this case, again, the testimony is given only in copies. and observations of populations over many years.
Again, this year, only 1 fullwums were found in one place where there were hundreds last year. Hundreds were found in another place, and several in a third. And what does that tell us? Where is the guarantee that next year there won't be hundreds on the first and third sites,but only a few finds on the second? In this case, nothing - just go next year and see..

2 Adamov
> I can't even imagine, in any way, even very abstractly, this very "mathematical proof base" in relation to the problem under discussion. What is this "evidence base"?

For example, you say that the ground beetle fauna of the local fauna of the Rostov region is 220-250 species. In support of this hypothesis, you make the argument: "There are very few studies on ground beetle fauna in small areas of RHO. They are characterized by a large sample size, a duration of observations of several years (up to 10), and are also located in different parts of the region." The question immediately arises – is this large sample size (for example, 50 thousand copies) sufficient? Proof? Is a follow-up period of 10 years sufficient? Proof? That's what I'm talking about. What are the sufficiency criteria?
Because, for example, if these works contained other data (the sample size is 20 thousand copies), how would you judge the sufficiency of the data provided? Would you consider this sample sufficient? What if the observation period was 5 years?


> In my opinion, in this case, not mathematics, but biology is primary.

No one is arguing. Although biology, in general, has nothing to do with it. Biology in this case consists in collecting material and defining it.


> How can one construct some kind of "mathematical proof" of representativeness (or math. criterion), without relying on empirical data? Nothing! And how can this source data be trusted when "creating" such a mat? criteria?
Naked mathematics can't prove anything.

Mathematical proofs are not based on empirical data. In terms of proofs, mathematics deals with axioms (statements that do not require proof) and theorems (statements that require proof).


> You just need to take and use all the methods to study the selected territory in detail . In the course of the study, it is possible to determine the "effective sample size" (trap samples, for example), as well as the number and quality of the studied sites, and after at least 2 years, it will be possible to judge the "required duration of observations". When all these "criteria" speak in favor of the approaching completion of the species list, you can use the Pesenko method.

So I want to know how you will determine the effective sample size, judge the required duration of observations, etc.? Based on "I think"," it seems to me " or " calculations showed that..."?


> In order to judge the average volume of local fauna, several years of research are required, if none has been conducted before. This is why a longer study is needed at several points to determine that there are not 20 or 500 or even 300 species in the local fauna, but for example 220 and 5, 6, ... 10 years of research bring only 0-2 species, without significantly affecting this figure, which will still be considered average.

You will not decide, but only think about life.


> Having data on the average volume, it is possible to determine the number of species using the Pesenko method using 1-year research data in a new territory – this will be the criterion of representativeness.

Yes, if previous studies were conducted, say, at the corners of a 10x10 km square located in a homogeneous area, you can assume that there will be the same number of species in the center of this square. But this is true only for small territories. For regions – it is no longer available. I cannot take the arithmetic mean for the Saratov and Astrakhan regions as a basis, knowing perfectly well that the conditions in them are very different and, accordingly, different complexes of species are formed. Here, indeed, biology is primary.


You can generally approach this problem from purely biological premises. If we are talking about fauna, then (1) we mean the habitat of a set of species, (2) the number of which must not be lower than a certain level at which the species is preserved. If this is the case, then nothing prevents us from recording the presence of each species from the entire population (fauna) in a short period of time (individually for the group). The problem is only in the selection and use of methods that should correspond to the biological / ecological characteristics of the "group".

Is there nothing stopping you? Yes, but only in the case of very small territories. The problem will also be that we will doubt whether all the views are fixed. Any method has an error rate. Or do we assume a priori that we will catch all the species using the method we use?

This post was edited by kovyl - 26.04.2010 20: 50

  

One story about faunal research.
There is such a galaxy of butterfly farmers who conduct research on "science". They specifically take something into account. they go to count something, catch it in a certain lane, so many meters away from the researcher. Then they build special graphs, use smart words, and insert the calculated values into smart formulas. And they get a certain result.
For example, for one of the sections of the Russian Altai, 140 diurnal species were given, and I gave 156 for one ridge (included in that section). And a total of 165 for the selection. What it says.
Only about one thing - the experience and desire to work of some and the desire to write a scientific publication for others.
In faunal research, the only thing you need to do is to want to study the region and know the collection methods. And plow.

2 Yakovlev
Another story about faunal research.
I'm opening a dissertation on lepidoptera of the Lower Volga region. The author caught 40 thousand copies.
I'm opening another dissertation on the lepidoptera of the Samara Luka. The author caught 40 thousand copies.
I am opening my third dissertation on the lepidoptera of the Upper Oka basin. The author caught 40 thousand copies.
I am opening my fourth dissertation on lepidoptera of the Yaroslavl region. The author caught 40 thousand copies.
Despite the fact that they were different people who caught different numbers of years in different territories, it's an interesting coincidence, isn't it? Maybe they all said "enough", catching 40 thousand copies.? Is this some kind of magic number?
I have no doubt that these people caught and caught a lot. But still strange.
That's the first thing.

And secondly, I didn't understand, is this a call to conduct research as God wills? In this case, only the list of species will be representative in the work (you can doubt its reliability as much as you want, but this will not make it worse). But to make all sorts of comparisons, to deduce coefficients - this is really sorry. First confirm the validity of this analysis.

> For example, for one of the sections of the Russian Altai, 140 diurnal species were given, and I gave 156 for one ridge (included in that section). And a total of 165 for the selection. What it says.

That's exactly what they're talking about here. However, your case is not so revealing. Not that much of a difference. And here is another one that better illustrates how to approach the "detection" of fauna with caution. The report on lepidoptera Volgo-Uralensis (part on moth) for the Volgograd region contains about 40 species. It would seem that there is a job, there are numbers in it (this is me Adamov). However, the article published after a short period of time already lists about 180 species of this family for the same territory. Well, if the number 40 did not have time to use for any "analysis", and then you can make far-reaching (but not in the right direction) conclusions.

Carefully re-read what is discussed in the topic.

This post was edited by kovyl - 27.04.2010 23: 40

it seems to me that the most important thing in faunal research is to know things that are common to different insects - for example, the necessary population size in order for it to be stable, i.e. the minimum number of individuals per certain biotope for a certain species so that it can maintain existence. Because you won't be able to calculate the full number of copies anyway.. This is hardly possible at all!

One species, even a very large one, can be searched to the point of freaking out and not found, considering it super rare, or even absent.
It's like we have with cerambix cerdo, which we wanted to catch from childhood, but could not. And it is included in the red Book and there are not many finds.. And for some, this kind of shit, just people know how to catch them.. Beetles in the crowns are found, what can you do..
Or as with Gnorimus variabilis-there was 1 find in the entire history, it was considered that the species was wildly rare.. No one caught it so easily in the crown traps - it falls there very well!

And what about the various territories where crowds of entomologists catch and catch everything year after year, and every year new populations, species, interesting finds! And years pass, and interesting insects do not stop being found! And how can we use an obscure form of mathematical apparatus to track the population size of at least one species? If dozens of entomologists can't tell for sure, catching it every year in one place.. What can 1 person give?
Everything is really very complicated here, or is it just me
What is the minimum ekz for the view? Having calculated approximately this, we would be able to give an answer to how many at least they should be in one place or another - and all we need to learn here is to look for them.. Although there are also very complex species, for example Cornumutilla sp. - a huge range, a huge food base, but very few people catch, and they catch only a few! And no one will tell you how to catch it! Well, there are certainly some instructions, but these are not guarantees, but only-look for the tipo..

I think that the most important thing in faunistics is a properly and fully compiled list of species for the territory.
40 thousand butterflies - you made me laugh. All the more so, I understand who you're talking about. Apparently, this is a mystery.
Oh, who would collect 40 thousand butterflies on the Central Siberian plateau. And then Oka, Luka, Lefty... Let's go to Siberia. 80% of the territory of Russia is almost idle
About the Volga-Uralensis - any healthy person understands that 40 species for the Volgograd region is the 5th part of the masimum fauna.

2 Yakovlev
> I think that in faunistics, the most important thing is a correctly and fully compiled list of species for the territory.

So that's what I'm saying. And in dissertations - the same thing. There is a list, a good list - what else do you need? No, let's analyze how many families, genera (!), etc. are there (and why, for example, not tribes, etc.?) and even compare them with other regions. Well, it is clear that there is a different picture, so what? I have already suggested comparing it with Australia. But if you have taken up the analysis, then see this thread.


> Oh, who would collect 40 thousand butterflies on the Central Siberian plateau. And then Oka, Luka, Lefty... Let's go to Siberia. 80% of Russia's territory is almost idle

If everyone rushes to Siberia, they would still think that there are 40 species of moth in the Volgograd region (or even less). In the European part, there is also enough to do.


> About the Volga-Uralensis - it is clear to any healthy person that 40 species for the Volgograd region is the 5th part of the masimum fauna.

First, "to any healthy person" is not an argument. Secondly, how did you determine that "40 species for the Volgograd region is the 5th part of the fauna maximum"? Calculations - to the studio! And then we are fighting here, discussing, and maybe everything is already ready?

This post was edited by kovyl - 28.04.2010 05: 31

 
First, "to any healthy person" is not an argument. Secondly, how did you determine that "40 species for the Volgograd region is the 5th part of the fauna maximum"? Calculations - to the studio! And then we are fighting here, discussing, and maybe everything is already ready?

You can almost say for sure the number of diaries in the Volga region..
if we talk about moths, then there are a lot more of them there, even only by a few spades.

2 Ripper
> You can almost say for sure the number of diaries in the Volga region..
if we talk about moths, then there are a lot more of them there, even only by a few spades.

And "goraazdo" is exactly how much? And on the basis of what such statements?


2 Adamov
> What should be considered a "sample" in general, in relation to the question of studying fauna (mainly meaning local ... it's easier that way)?
I can say that not exactly the number of instances.

Of course not. In this case, the number of species will be considered as a sample.


> The scheme is approximately as follows: 1) select the territory to be explored (possibly even an arbitrary one); 2)select ("separate") areas in the given territory. land plots (biotopes, etc.)... this is also a sample, since you subjectively select some areas, and you do not study the entire set of selected classes (if we are talking about really feasible work, by the way, Pesenko also said this in his book). This sample (of biotopes) should also be representative; 3) further, in my opinion, Pesenko discusses the issue of placing samples on these " biotopes "and gives preference to randomized"schemes" to varying degrees. But I don't agree with the last point. In my opinion, it doesn't make sense to "randomize"faunal collections.
I don't remember whether Pesenko has it, but as a fourth point I would note the representativeness provided by the range of collection methods and the volume of samples (number of traps). The range of methods is determined mainly by the objects of research. Sample volume – by the sensitivity of the methods.
The 5th point, I think, can be called repetition (which, by the way, will mainly determine the number of specimens caught), in a broad sense (in space and time). But this is understandable, the more the better, like everything in this business.

Everything is so. The only thing I would like to clarify is: "in this territory, they are selected ("isolated") land plots (biotopes, etc.)... this is also a sample" – this is not a sample, but the conditions for forming a sample (in our case, the number of species).


Now let's figure out how these 4 aspects of representativeness can be taken into account when "creating" a mathematical criterion. …

Yes, I agree on all points. But here's one small detail. We assume that the researcher will select everything correctly, find it, and so on.


> If species live on a territory and you have studied this territory in its main "differences", using all methods with sample volumes exceeding a certain sensitivity threshold, then why can't you record all the species?

Because a sufficient amount of samples is unknown. I say again that we can only assume that all (or almost all) types are fixed.


>Can I use an example from entomology?

Example of a proof, axiom, or theorem? I didn't get it right. I'm talking about mathematics, and I'm being asked to give an example from entomology.


> To be precise," I think"... "I thought"... "I decided"... "by the way, some people do this here, but I'll probably do it a little differently"

Yes, but the whole question is on what basis you say this. Based on your own feelings (experience, etc.) or mathematical calculations. And then you can say that the notorious Jaccard coefficients, etc., are somehow calculated incorrectly: "I decided that it is possible to calculate it differently" (there is a very good "Jumble" on this topic – "Arithmetic").


> Yes no... I'll make up my mind. I can guess if I take a number of values for adjacent territories from the literature and draw a conclusion for the required one.
From my premer, " 220 " can be used in the Pesenko method (determining the number of species). There, after all, an abundance of more than half of the local fauna is required. That's half of it, about 110 species. If I determine the abundance for 150 species, for example, I can determine the volume of fauna using the Pesenko method. And then focus on this theoretical figure.

Here is already Tov. Yakovlev gave an example with pyadenitsy of the Volga region. It is good that they did not take the first digit, otherwise they would have got into a puddle with it.
Then, I said that it is doubtful to compare the fauna of the region with remote regions (for example, the middle zone is compared with the north of the ECH, Karadag, etc.)


> Do I have the right to assume that in the east and south, there will also be an average of about 230 species? I believe so.

Sure. You have the right to assume anything. This is why there are proofs of theorems that assume something. But subjectivism in their proof is completely inappropriate.

This post was edited by kovyl - 28.04.2010 16: 02

much more.. Well, I assume that only moths, about 500 species, and maybe more. But since I do not know the moth fauna, my words are rather empty

But there are people who have been catching them for many years, making lists, entering label data. Thus, the overall picture is already formed, and for the rest you need to continue to catch.

This is a hopeless argument, I think. Yes, mathematics rules, but not in this case. Because of the inability to standardize everything.
If someone is well versed in mathematics, remember biometrics. There is also information about the sample size. Something comes to mind about the magic number 30, which is less than-and not a sample, but so... But this is calculated for the ideal case, when everything comes across with equal probability. It is clear that this is not the case in faunistics. The harder and more thorough you worked, the closer you got to the ideal.
And about intuition and common sense - so there is basically an unconscious comparison with other more or less similar (in climate, size, and other factors) territories.
By the way, the number of species can't possibly be a sample in this case. Rather, the sample is something like the "number of collection acts", and what is there to understand by this..

... ... ...
Oh, who would collect 40 thousand butterflies on the Central Siberian plateau...
.... ...

И?

No, I can't, I don't know much about the Volga Region. . But, it seems, the locals should be zant, and why don't you know? 0_o Little catch, read, know other hunters, as well as the fauna of neighboring areas? Or is there something in the Volga region that makes it stand out so much from other regions that there are species that no one knows about?

In Moldova, a few people caught butterflies (more or less), and even more so not so many and carefully.. But I can tell you an approximate number of types of diaries.. There were a lot more people fishing in the Volga region and a lot more thoroughly, weren't there?

2 Vorona
No, the magic number 30 has nothing to do with it. And it's not about equal probability.

> And about intuition and common sense - so there is basically an unconscious comparison with other more or less similar (in climate, size, and other factors) territories.

So what do the middle lane, Kivach and Karadag have in common? Maybe just the square... Who will have a region that is comparable in area to Madagascar-compare. Discover a lot of interesting things.


> By the way, the number of species can't possibly be a sample in this case.

Perhaps, instead of the term "sample", we should use "general population".


Ow, mathematicians!

The number 30, of course, has nothing to do with it. I just gave an example of the existence of a mathematically valid minimum sample size. but. This is all applicable when conditions can be standardized: here's a field (sorry, it's easier for me with tops) or a forest. The forest is better. We number all the trees, take a table of random numbers - and go ahead, type a sample. According to this table. Here you can say: you didn't get 30 (or better - 100) objects - your sample is not representative.
How about faunal research? Well, you will have a number, and what to do with it? And the "general population" can hardly be applied. This may be the number of species that are actually present on your territory, but then the sample is the number of species that a particular researcher obtained. And the sample is not equal to the general population.
Is the number of species in the middle zone, in Kivach and in Karadag the same?
It reminded me of a story. Practice in entomology with physiologists in Ladoga. Somehow it just so happened that in the forest and in the meadow they collected the same number of families, or orders in general (but we were not much steamed with the definition, so we definitely did not know the species). Molchel, who was assigned to prepare a report on this particular issue, makes a bold statement. that the number of ecological niches in the forest and in the meadow is the same Stifled moans of the team "I-di-oooot!" No one checked each other

and we usually make selections-mowing 100 strokes of the net.
There are sometimes interesting species, and it may be normal to find some species, but there are always a few beetles there, and even fewer good ones..
This, of course, is absolutely not indicative, but it is so accepted, they say it shows something.
And once I was sent to mow down butterflies-they say catch a hundred butterflies in one place or another, then we will determine and see who is how many.. So what! What will it show?! Well, potmo do the percentage of species, and the sense!? You can't apply a mathematical approach here - it won't give you anything! In this case, only the seeker will be able to get closer to the truth, evaluating what he saw, where he saw, in what numbers..
But there is a catch - they send a person, and they can't trust his authority, knowing little about him, so they will not be able to evaluate the fauna on the basis of it is not known what way this person evaluates the place. In this case, it would work better - if you sent 2 people - who do not understand anything, who will catch everything in a row, and at least a little versed in butterflies and knows where and what to catch, then it will be possible to compare the pure percentage of mowing everything in a row with the data that the fumbling person will get. Especially if the first one will catch by time, and not by the number of copies.. For example, give both 1 day. That would have shown more. the fact is that the second one would hunt more precisely for interesting species, and would complement the picture in terms of species, and the first one would be more quantitative. Then you can compare neighboring fauna, if the same 2 people will do all the dirty work and will apply the same approach everywhere. That we can determine, for example, by the number of species or the abundance of animals - that one place is better populated, there are more or less species there, etc. Of course, you need to compare similar areas. And, in principle, we can compare different places - we will find that there are different numbers of species and different densities of these species, then we will ask why. Then it will be necessary to conduct other studies, for example, what affects the number of species and their number.