Uncommon Descent Serving The Intelligent Design Community

At Sci-News: Moths Produce Ultrasonic Defensive Sounds to Fend Off Bat Predators

Categories
Evolution
Intelligent Design
Share
Facebook
Twitter/X
LinkedIn
Flipboard
Print
Email

Scientists from Boise State University and elsewhere have tested 252 genera from most families of large-bodied moths. Their results show that ultrasound-producing moths are far more widespread than previously thought, adding three new sound-producing organs, eight new subfamilies and potentially thousands of species to the roster.

A molecular phylogeny of Lepidoptera indicating antipredator ultrasound production across the order. Image credit: Barber et al., doi: 10.1073/pnas.2117485119.

Bats pierce the shadows with ultrasonic pulses that enable them to construct an auditory map of their surroundings, which is bad news for moths, one of their favorite foods.

However, not all moths are defenseless prey. Some emit ultrasonic signals of their own that startle bats into breaking off pursuit.

Many moths that contain bitter toxins avoid capture altogether by producing distinct ultrasounds that alert bats to their foul taste. Others conceal themselves in a shroud of sonar-jamming static that makes them hard to find with bat echolocation.

While effective, these types of auditory defense mechanisms in moths are considered relatively rare, known only in tiger moths, hawk moths and a single species of geometrid moth.

“It’s not just tiger moths and hawk moths that are doing this,” said Dr. Akito Kawahara, a researcher at the Florida Museum of Natural History.

“There are tons of moths that create ultrasonic sounds, and we hardly know anything about them.”

In the same way that non-toxic butterflies mimic the colors and wing patterns of less savory species, moths that lack the benefit of built-in toxins can copy the pitch and timbre of genuinely unappetizing relatives.

These ultrasonic warning systems seem so useful for evading bats that they’ve evolved independently in moths on multiple separate occasions.

In each case, moths transformed a different part of their bodies into finely tuned organic instruments.

[I’ve put these quotes from the article in bold to highlight the juxtaposition of “evolved independently” and “finely tuned organic instruments.” Fine-tuning is, of course, often associated with intelligent design, rather than unguided natural processes.]

See the full article in Sci-News.

Comments
Alan Fox:
Care to comment on gene duplication?
I will. New genes require new binding sites. Care to comment on how you determined that blind and mindless processes can just produce new binding sites as needed? Not only that, a gene duplication just gives you another copy of something that already existed. It is question-begging to say that gene duplication is a blind and mindless process. Especially gene duplication, followed by a new binding site, followed by function-changing mutations. To a function that just happened to have value and not get in the way of existing systems and functions.ET
August 15, 2022
August
08
Aug
15
15
2022
06:48 PM
6
06
48
PM
PST
Alan Fox:
I pretty much despair of being understood by anyone here, and I find most of what you write about biology nonsensical.
Wow. You never support what you post. Your personal notion of "NICHE designs" still remains in your imagination. As for biology, you haven't shown any knowledge there, either. All you have ever done is refuse to understand ID, all the while refusing to present any scientific alternative.ET
August 15, 2022
August
08
Aug
15
15
2022
06:43 PM
6
06
43
PM
PST
The example with coins is a bad analogy for functional information because the end of flipping coins is nothing functional(it's just an imaginary construct of randomness ) that have no relevance . The example with coins is a bad analogy for functional information because there is an intelligent agent that flipps the coin and materialists try to prove the opposite (that there is no intelligent agent). This example is a self defeater for materialism. The example with coins is a bad analogy for functional information because the chemical reactions to form a functional protein is never binary . A closer to reality example is to replace the binary coin with dice with 1000 sides .Lieutenant Commander Data
August 15, 2022
August
08
Aug
15
15
2022
03:18 PM
3
03
18
PM
PST
On tossing coins, a reminder: https://uncommondescent.com/mathematics/distribution-of-tossed-coins-a-reminder/kairosfocus
August 15, 2022
August
08
Aug
15
15
2022
02:05 PM
2
02
05
PM
PST
Kairosfocus: First, FSCO/I is OBSERVED not calculated or assumed There was no specified complexity in the last example I looked at. Taking a 500 coin string and 500 bit threshold, 500 coins in a typical gibberish pattern will carry 500 b its of info capacity but no specificity or function and would score zero functionally specific bits. Applied to the threshold, that would give – 500 bits, that many short of threshold. But that's not what I got when I applied your version of Dr Dembski's metric based on what you wrote. I got -398.something. I asked you over and over and over again to apply your version of Dr Dembski's metric to a given example. You refused to do so. I did, based on what you wrote on this thread (comment 276) and I got a completely erroneous answer for a clear case of 500 bits of no information. If I applied your version of his metric incorrectly well then you should point that out. You didn't say that I did that. Remember that Dr Dembski meant for his metric to be able to be applied in all cases, even ones with a negative outcome. So, who got it wrong? Dr Dembski? Me? Or you? If Dr Dembski's metric was so wrong why did you spend time reconfiguring it? Since you did do that then why does your version give such bogus values? We both know the expected result from tossing 500 coins, from experiment and from the binomial distribution. Strange that your version of Dr Dembski's metric didn't give that result.JVL
August 15, 2022
August
08
Aug
15
15
2022
02:01 PM
2
02
01
PM
PST
AF, doubling down on a strawman does not transmute it into truth. We both know the expected result from tossing 500 coins, from experiment and from the binomial distribution. This exchange is telling us what long time objectors to the design inference are reduced to arguing. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
01:56 PM
1
01
56
PM
PST
AF at 552, No. You are desperate to avoid this getting into the schools. That's all. The right wingers will track you down. Force you to watch reruns of Father Knows Best and Ozzie & Harriet. Oh yes. Everything will go back to normal. Back to the 1950s.relatd
August 15, 2022
August
08
Aug
15
15
2022
01:25 PM
1
01
25
PM
PST
...a typical gibberish pattern...
KF assumes his conclusion yet again.Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
12:56 PM
12
12
56
PM
PST
Meanwhile, the words I just wrote were specific, complex and in a specific order so anyone – even you – can read them and understand them correctly, just like living things, but you choose to ignore that. And will continue to ignore that. Prove me wrong.
Well, I think most exchanges here happen in a fog of miscomprehension. I pretty much despair of being understood by anyone here, and I find most of what you write about biology nonsensical. On the other hand, you have written stuff (not on biology) that made sense to me.Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
12:52 PM
12
12
52
PM
PST
JVL, projection. First, FSCO/I is OBSERVED not calculated or assumed, what we estimate is value in bits. Taking a 500 coin string and 500 bit threshold, 500 coins in a typical gibberish pattern will carry 500 b its of info capacity but no specificity or function and would score zero functionally specific bits. Applied to the threshold, that would give - 500 bits, that many short of threshold. 500 bits with 72 ascii characters would effectively be at 0, equipoise, 1,000 bits with 143 ascii characters would be 500 bits beyond. All of this was addressed hundreds of comments ago. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
12:46 PM
12
12
46
PM
PST
Kairosfocus: coin flipping, especially of “fair” coins to the tune of 500, will overwhelmingly produce gibberish clustering fairly near to a 50-50 distribution. As any particular string is as good as any other, we have non specific values and most likely non functional gibberish. Correct. So why was it when I tried to apply your version of Dr Dembski's metric I got an indication of design? 500 coins have 500 bits of information carrying capacity, before any flipping or setting is done. Using ASCII code as reference, 500 coins exhibiting 72 or so characters in English would be FSCO/I, a strong sign of design. This has been discussed ever since. You're just babbling now. What you really need to do is to address the fact that a fair application of your version of Dr Dembski's metric gave an indication of design when clearly none existed. You've been given chance after chance after chance to explain your modification of Dr Dembski's metric. You have steadfastly refused to do so. I tried, to the best of my ability based on what you actually wrote, to apply your criteria to a given, clear example. And I got a clearly bogus results. If I were you I'd revisit my interpretation of Dr Dembski's metric.JVL
August 15, 2022
August
08
Aug
15
15
2022
12:22 PM
12
12
22
PM
PST
My correction did not come through. "But, it's pretty clear that any K2 gives a result > 1. Therefore design."JVL
August 15, 2022
August
08
Aug
15
15
2022
10:51 AM
10
10
51
AM
PST
JVL, coin flipping, especially of "fair" coins to the tune of 500, will overwhelmingly produce gibberish clustering fairly near to a 50-50 distribution. As any particular string is as good as any other, we have non specific values and most likely non functional gibberish. 500 coins have 500 bits of information carrying capacity, before any flipping or setting is done. Using ASCII code as reference, 500 coins exhibiting 72 or so characters in English would be FSCO/I, a strong sign of design. This has been discussed ever since. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
10:48 AM
10
10
48
AM
PST
AF, projection again:
How about any of you that think “Intelligent Design” is more than a cover for a right-wing religious agenda tell me something positive or useful about it?
1: "Right-wing religious agenda" is little more than saying Christo-fascist, would be theocrat, i.e. Nazi. 2: This is based on a triple error: [a] that the national SOCIALIST German Worker's Party was of the "right," [b] that the left-centre-right political spectrum has any real utility (it is severely outdated), [c] that the Christian faith has any particular affinity for theocratic, nazi totalitarianism. Go read the Barmen Declaration as a start point. 3: Those who are obsessed with political agendas and see government as messiah project this to others. The correct Christian attitude to this, is to see political messianism as anti-christ . . . both, counterfeit and against count . . . idolatry. Follytricks and bureaurats [that's deliberate] are not the source of salvation. 4: It is obvious that there is a calculated slander that design thought in science is a political-religious push against true science, which is of course generally socialist-progressivist and embeds a priori evolutionary materialistic scientism, while monopolising or dominating serious knowledge. (The cases in point are many.) 5: In fact, design thought and inferences go back to Plato et al, and have never gone away. Where, Newton's general scholium to Principia makes that as plain as could be. 6: From the 1940s to 50's on, with the elucidation of the genetic code and with the recognition of cosmological fine tuning, design thought, inferences, schools and a broader movement have had a resurgence, not as a reaction to leftism or some plot to capture civilisation, but because codes and algorithms in the heart of the cell reflect language and goal-orientation, while fine tuning points to extracosmic design. 7: In the case of FSCO/I, it is manifest that you have had increasingly shrill bluster and accusation, but notably little substance. 8: As for believing in God, that is the general view of a majority of humanity, and ethical theism can be substantiated in detail, leading of course to highlighting the significance of moral government of our rational responsible freedom, an antecedent of science. 9: So, the slander fails. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
10:39 AM
10
10
39
AM
PST
Now, based on what Kairosfocus wrote in post 276 above: X = I(T) - 398 - K2. He says I(T) is the same as -log2(P(T|H)) so I think that makes it 500. He says K2 has a "natural upper limit" of about 100 bits but doesn't specify it any further. So, now we have: X = 500 - 398 - K2 = 102 - K2 So, what have we got? IF K2 is at its natural upper limit (100) then X = 102 - 100 = 2 > 1 so, by Dr Dembski's criterion, design. BUT, it's pretty clear that any K2 1. Does this make sense based on what we know of the (assumed) source of the sequence of Hs and Ts? A completely random sequence is showing specified complexity? That there is design? This is why I asked Kairosfocus REPEATEDLY to evaluate his own version of the metric because I didn't think it was giving sensible answers. I will leave it to him to respond to these results.JVL
August 15, 2022
August
08
Aug
15
15
2022
10:28 AM
10
10
28
AM
PST
"... we do not conclude that there is any specified complexity and therefore no design. Which is what we expected." We? The vast Left-Wing Conspiracy?relatd
August 15, 2022
August
08
Aug
15
15
2022
09:55 AM
9
09
55
AM
PST
I'm going to break apart Dr Dembski's metric to make it easier to compare with Kairosfocus's version. So . . . X = -log2(10^120•pS(T)•P(T|H)) = -log2(10^120) - log2(pS(T)) - log2(P(T|H)) Based on previous analysis: -log2(10^120) is approx = -398.63137 P(T|H) is clearly 1/2^500 = 2^-500 so -log2(P(T|H)) = 500 I will argue that a purely random sequence of Hs and Ts has a pS(T) of 2^500 since basically the only way to describe the sequence is just list it. Or, at least, that there are no sequences with a more complicated/lengthy description Therefore: -log2(pS(T)) = -log2(2^500) = -500 So, what have we got? X = -398.63137 - 500 + 500 = -398.63137 which is less than Dr Dembski's threshold of one so we do not conclude that there is any specified complexity and therefore no design. Which is what we expected.JVL
August 15, 2022
August
08
Aug
15
15
2022
09:52 AM
9
09
52
AM
PST
Kairosfocus: there is always a place for an analysis, but there is also a place for a bound. They are not mutually exclusive. Anyway, I'm going to try another simple example, evaluate Dr Dembski's metric for that example and then do my best to evaluate Kairosfocus's version for the same example. He is welcome to head me off at the pass and do it himself. The example I'd like to use is: 500 fair flips of a fair coin. So, we're assuming that the outcome is actually random, i.e. no bias, no design. This is just to test the metrics, it's not an argument about the origin of sequences generated by coin flips although I'm sure some of you will want to digress in that direction. We expect Dr Dembski's metric to show NO specified information and therefore no design because . . . there isn't any. Because of the time of day my internet is quite dodgy and I've already lost several versions of this post so I'm going to break up my progress into smaller posts. Sorry about that but I'm tired of retyping everything.JVL
August 15, 2022
August
08
Aug
15
15
2022
09:38 AM
9
09
38
AM
PST
AF at 536, "... more than a cover for a right-wing religious agenda..." Oh yes. The worship of Politics. You know all you need to know. The right-wingers are out to get you. Meanwhile, the words I just wrote were specific, complex and in a specific order so anyone - even you - can read them and understand them correctly, just like living things, but you choose to ignore that. And will continue to ignore that. Prove me wrong.relatd
August 15, 2022
August
08
Aug
15
15
2022
08:35 AM
8
08
35
AM
PST
JVL, there is always a place for an analysis, but there is also a place for a bound. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
08:33 AM
8
08
33
AM
PST
AF, confession by projection. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
08:32 AM
8
08
32
AM
PST
Kairosfocus: the WmAD expression invites executing the neg log operation thus reduction to information beyond a threshold. The value for the threshold can be bounded based on search challenge on sol system and observed cosmos scale. This is practically useful, and decisive. It is also far more general than solutions for particular cases. Why put in pS(T) and P(T|H) then which address a particular pattern in a particular sample space if he didn't want to try and make particular arguments in particular cases? Why did you replace parts of his formulation with K2 and I(T) which definitely sounds like it's dependent on T therefore NOT general. Can you define I(T) please. It looks like a function: what are its range and domain? Are you just not interested in possibly reducing the threshold for some situations? Are you not interested in handling cases which have a lot fewer than 500 bits? Would you not like to be able to confidently say 'design has been detected' in such situations? You may not like it, but that is and has always been, my answer. An answer that feeds into a central, decisive result. So, does that mean you can't or just won't evaluate K2 and I(T)? If they're meaningless why not make the whole thing even simpler and combine them together? Did I evaluate Dr Dembski's metric correctly for the examples I chose? Surely you can answer that question. In short, we do not need to have a specific calculation of targets So Dr Dembski shouldn't have bothered with his new metric? You think it was a waste of time.JVL
August 15, 2022
August
08
Aug
15
15
2022
05:55 AM
5
05
55
AM
PST
Which leads directly to the blind, needle in haystack search challenge.
..which is a bogus challenge as there is no way to know how many needles (sequences with functionality) are there to be found nor is an exhaustive search necessary: finding something that works well enough is, well, enough. Care to comment on gene duplication? Happens a lot, apparently, even up to whole genome duplication. Sets up a scenario where genes are surplus and can change over time without deleterious results. No need to search for new needles, old needles can vary over time and sometimes drop on to new functionality.Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
05:03 AM
5
05
03
AM
PST
PS, you already had an answer from the outset, the WmAD expression invites executing the neg log operation thus reduction to information beyond a threshold. The value for the threshold can be bounded based on search challenge on sol system and observed cosmos scale. This is practically useful, and decisive. It is also far more general than solutions for particular cases. You may not like it, but that is and has always been, my answer. An answer that feeds into a central, decisive result. PPS, Rendering phi as p, WmAD wrote, "define pS as . . . the number of patterns for which [agent] S’s semiotic description of them is at least as simple as S’s semiotic description of [a pattern or target zone] T." He symbolised pS(T), and I understand him to be conceiving of other similar targets amenable to a detachable, simple specification, and in the Caputo case he discussed outcomes on a Bell curve beyond a cutoff. More generally we can see him as addressing the dust of similar target zones in a configuration space. Which leads directly to the blind, needle in haystack search challenge. Which in turn invites the sort of bound estimation I gave above and going back over a decade. In short, we do not need to have a specific calculation of targets similar to the observed, once we realise that FSCO/I reduces to a dust in the config space due to need for multiple, well matched, properly oriented and organised coupled parts to achieve function. This comes out in even the contrast between gibberish/random text strings [random tars in organic reactions], simple repetitive patterns [similar to crystal unit cells] and organised functional patterns [e.g. proteins].kairosfocus
August 15, 2022
August
08
Aug
15
15
2022
04:52 AM
4
04
52
AM
PST
LCD Sauce for the goose? How about any of you that think "Intelligent Design" is more than a cover for a right-wing religious agenda tell me something positive or useful about it?Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
04:51 AM
4
04
51
AM
PST
:lol: "If there is no scale with how many decimals I want then weight do not exist."Lieutenant Commander Data
August 15, 2022
August
08
Aug
15
15
2022
04:47 AM
4
04
47
AM
PST
...coded algorithms in D/RNA...
Again, and this is the actual important point for me, you are utterly confused over the difference between complexity and function. There is no way, now and for the foreseeable future, that you or anyone else can predict the functionality of an unknown nucleotide sequence (or the potential product it is a template for), simply by doing simplistic counts and transformations of numbers. The only way to establish functionality is to test for it. Or show me I'm wrong.Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
04:46 AM
4
04
46
AM
PST
Let the observation stand, in comments where you object to FSCO/I you are compelled to produce FSCO/I in the form of complex text in ascii coded English.
I am not disputing your ability to compute some figure that you can call FSCO/I if you wish from a sentence in English. What I want to know is what is the point? What will the figure indicate? Will your method distinguish a sentence that will mean something to the reader compared to a sentence carefully designed to carry only nonsense (Colorless green ideas sleep furiously, f'rinstance).Alan Fox
August 15, 2022
August
08
Aug
15
15
2022
04:39 AM
4
04
39
AM
PST
JVL, what is central and decisive is just that. The core design inference question has a good, affirmative answer. Yes, there are reliable signs of design. FSCO/I is one of them. This relates to coded algorithms in D/RNA and so to OoL and Oo Body plans, including our own. That, is highly significant and, rightly, is focal. It is also detachable from whatever WmAD may have that you think is error. Such, is secondary. KFkairosfocus
August 15, 2022
August
08
Aug
15
15
2022
04:38 AM
4
04
38
AM
PST
Kairosfocus: doubling down on side tracks. The central question is answered You are so weird. I WAS NOT questioning the validity. of having a specified complexity detection threshold. I was trying to see how Dr Dembski's 2005 version works, does it give different thresholds under certain circumstances and does it uphold any design detection claim for a situation with fewer than 500 bits. I was also wanting to compare Dr Dembski's original formulation with your reworked version. So I did the mathematics for Dr Dembski's version, it was pretty simple really but I did pick two very simple examples for a start. Then I asked you to do the mathematics for your version and you have never, ever been able to evaluate your version for the specific examples. After it was clear you wouldn't (or couldn't) evaluate your own version I asked if you accept my work with Dr Dembski's version. Which you also wouldn't (or couldn't) do. NO WHERE in all that did I express a doubt about the whole point of looking for specified complexity as an indication of design. That's a different conversation but we're not having that now. I think I may have found a case of FEWER than 500 bits which should please you. But you either don't understand what I've done or what I've said or you just can't change your tune. I will try again: Can you evaluate your metric for the case of flipping a fair coin 400 times and getting 400 tails, i.e. can you tell us what K2 and I(T) are for that case. Yes or no? Do you accept that the mathematical work I've done is correct and that Dr Dembski's metric seems to indicate a threshold of 401 bits for the above example in order for a valid claim of specified complexity to be made? Yes or no? I'm going to keep asking 'til you answer those specific questions.JVL
August 15, 2022
August
08
Aug
15
15
2022
04:19 AM
4
04
19
AM
PST
1 3 4 5 6 7 23

Leave a Reply