Hunting the Silent :

Convergent Evolution in Acoustic

A TALE FROM THE AMAZON

Santiago and I were walking back from a night of collecting specimens in the forest. We had left the research station at sunset, had paddled our low dugout through flooded woodland using the short, pointed-tipped paddles which are good for pushing through mud and shallow swamp, out onto the Amazon river and to the indigenous village of Mocagua, Colombia, where we had met up with our guide and walked through town to a stagnant pond (which had to be crossed in a second canoe), and to the narrow trail which led into the rain forest, and nobody knew how far. For five hours we had wrestled with branches and vines, hiking and searching for crickets, filling the small plastic vials one by one with our quarries, many of which would turn out to be precious representatives of previously unknown species. We had turned over leaves, pried away bark, peered into holes, and sifted through debris, always keeping a wary eye out for those enormous, hand-sized bird spiders which, like crickets, are excellent jumpers.

Crickets had been caught and crickets had escaped, until it became clear that the batteries in our headlamps would not last much longer. Our guide had led us back out of the forest, alive and unharmed. By the time we returned to Mocagua, the generator had long since quit, revealing the stars of the Southern Hemisphere, and casting the platform huts into darkness. An army of dogs announced our return to the village. Some candles were lit in the one building which was used as the general store, the meeting-place, the dance hall, and the saloon: an open-air porch with a thatched roof, without windows or doors. For the third night in a row, we were met by a small group of

Tikuna men who offered us rum and beer at the end of the night. Santiago and I always felt a bit awkward joining them like that, me a conspicuous six feet tall and him still taller, both of us in long sleeves and trousers and smeared with mud from the night’s work -- as the men sat together, freshly washed, shirtless and barefoot and comfortably cool with bottles of warm beer in the Amazon night. Our guide took a seat next to them.

The Tikunas of that region are accustomed to field biologists, but seem incredulous and entertained at the lengths we will go to, to study the details of nature.

Conversing in Spanish, they like to recite our own story back to us, in the form of a question, to make sure they’ve got it right:

“You came to Colombia -- on an airplane -- and you traveled down the Big River all the way here, just to catch some crickets?”

“That’s right,” I say, foreseeing their next question:

“Aren’t there any crickets where you come from?”

“Sure, but not like these.” I lay a few vials out on the table, between the beer bottles and the shot glasses. Chestnut hands reach for the vials and hold them to the candlelight, as our finds are examined and passed around to the others.

“Ah, yes...” says one, “special crickets. Everyone comes here to find special things: special birds, special flowers, special snakes...” The men bob their heads in agreement.

“I don’t care if they are special,” says our guide, “I use them for fishing bait.”

2 “That’s why you find over half of what we catch. You’ve got the eye for ‘em!”

We sit in silence for a while, until one of the men says,

“I know a story about a cricket.”

A cricket had been up late one night drinking rum, and there was a full moon, and he went out to sit on a bridge over the River. He saw his reflection in the water and began admiring himself out loud, so that all the other could hear every word. “Just look at me! Why, I’m the finest creature in the forest,” he said. “I’m more beautiful than Hummingbird, I sing sweeter than Tree Frog, and what’s more, for my size, I am ten times stronger than Jaguar. They ought to make me the king around here, since there’s no one better suited to it.” He went on like this for some time on the bridge, drinking rum and admiring his reflection in the water and congratulating himself on his good qualities, until a particular monkey who liked to cause trouble ran off to tell Jaguar, who was the true king of the forest. Jaguar, being a social fellow, sought out the cricket and sat beside him on the bridge in the moonlight. “Hello, Jaguar,” said the cricket. “Good evening, Cricket,” said Jaguar. “Say, Cricket, I hear you’re up here drinking rum, gazing at your reflection in the river, enjoying how great and wonderful you are, and announcing to everyone that you are stronger and prettier than me and should be elected King of the Forest. Is that so?” “Yes, sir” said the cricket, who quickly added, “but drinking always blurs my vision.”

ACOUSTIC COMMUNICATION IN CRICKETS

Crickets in nature are not known to drink rum and converse with predatory mammals, but the one in this joke is an apt caricature of a real cricket anywhere in the world. In nature, male crickets of most species stay up all night advertising themselves to females through audible chirps produced by rubbing the front pair of wings together Ð a behavior known as stridulation -- and like the cricket in the story, they also fall silent

3 when danger is near. Although female crickets lack the structures for stridulation, adults of both sexes have two auditory organs, one on each front leg, consisting of a perforation and a tympanal membrane, mechanically similar to the ears of mammals. Acoustically, a cricket’s song is characterized by a particular pulse rate, frequency, and amplitude, voiced in a specific combination which is unique to each species, helping to ensure that males and females pair up with the right kind of mate. Listening closely on a summer night will reveal not one type of cricket song, but a multitude of variants, representing the different calling species in the area at the time.

The system, however, is much more complicated than a simple game of call and response. Through several decades of behavioral experimentation, biologists have plainly shown that female crickets are extremely picky, and tend to go for conspecific males with the ‘best’ songs1. The parameters which constitute the best songs seem to be arbitrary and vary from species to species, but females have their criteria and there is little that males can do about it. Successive generations of competitive males and choosy females are believed to drive the evolution of males to make their songs as alluring as possible Ð a process Darwin called sexual selection2. Males with the most attractive songs will be more likely to mate and will therefore leave the greatest number of progeny.

Since there is a partial genetic basis to calling song3, many of the offspring will inherit the ability to sing like their fathers. A number of researchers argue that females don’t care as much about the ‘best’ songs as they do about the information songs contain which can

1 Reviewed in Otte, 1992 2 Sexual selection differs from natural selection in that the affected traits do not necessarily enhance the health, strength, or lifespan of the organism, but rather enhance its reproductive output due to the increased frequency of mating. 3 Webb & Roff 1992, Shaw 1996

4 help them select the best mates. Such cases are known in several species of frogs, where females prefer calls which are indicative of males with large bodies. Evolutionary theory suggests that if a male can produce the song which most closely matches the criteria of conspecific females, his talent is interpreted as an index of his biological fitness, or capacity to produce large numbers of strong and fertile offspring. In anthropomorphic terms, each male courts the females of his species through a broadcast of braggadocio -- singing his sweetest song in an attempt to persuade them that he is the finest mate, worthy of being the father of their children and beyond that, the king of the forest (or at least of all conspecific males).

In today’s noisy world of frogs, birds, mammals, and other animals which produce acoustic signals, the chirping of crickets is usually heard as a subtle undertone within nature’s soundscape, when in fact it is one of the most ancient systems of communication by sound in the history of the Earth. Orthopteran insects Ð the lineage to which crickets belong, together with locusts, other grasshoppers, and their sisters, the katydids Ð are first known as fossils from the late Paleozoic, almost 100 million years before the appearance of the dinosaurs. The first Orthopterans were more like modern katydids than modern crickets, but their wing morphology shows that they stridulated using the same mechanism as crickets and katydids today4. Time travelers to the steamy Carboniferous5 swamps so often depicted in museum dioramas would find their surroundings to be eerily quiet -- with no birds singing, frogs croaking, dogs howling, monkeys screaming, or any other sounds at all, save for those made by the ancestors of crickets and other

4 This evidence comes from the examination of fossilized forewings, which clearly show the specialized vein which is the point of friction during song production. 5 The penultimate period in the Paleozoic era, lasting from approximately 360 to 290 million years ago.

5 insects. Now, more than two hundred and fifty million years later, scientists are aware of about 4000 different species of crickets on Earth Ð about the same number as the known species of frogs and toads combined. The true number of cricket species in the world today may well exceed 20,000 – four times as many as the world’s mammal species, and more than double the 9000 or so known species of birds Ð though we may never arrive at a complete census, since crickets are small, drab, and elusive, and precious few biologists in history have made careers out of combing the world’s natural habitats searching for them.

However, there have been those precious few: Saussure, Chopard, Scudder,

Bruner, Rehn, Hebard, Hubbell, and others, field-hardened men whose names never became famous outside entomology, but who were all skilled biologists and passionate cricket-hunters who spent years engaged in the discovery and understanding of thousands of species. Without question, the most prominent working today is Dr. Daniel Otte at the

National Academy of Sciences in Philadelphia, who has discovered and described over a quarter of all known cricket species, and who has spent years traveling the world, collecting and documenting the cricket faunas, making the first drawings and writing the first descriptions of new species, and depositing the first specimens of those species into research collections. It is a strange profession with few predecessors, all of whom have surely had their share of tangled forests, muddy trails, alarmingly huge spiders, and the eternal questions from the locals near their field sites: “You came to (name of country) on a (type of transportation, usually expensive), and you traveled (some absurd distance), all the way here, just to catch some crickets? Aren’t there any crickets where you come from?”

6 A MODEL FOR THE STUDY OF EVOLUTION

While the traditional aims of cricket research have been to study the evolution and diversity of these insects and their courtship songs, my own research addresses a different mystery, one which pertains more to the evolutionary process than to crickets themselves.

When I first began learning about acoustic communication in crickets, I was struck by the fact that hundreds of species do not sing at all. The absence of singing behavior in these species is almost always the result of physiological muteness and deafness Ð the total, physical absence of stridulatory structures on the wings, and the auditory structures with which to hear them. Rather than being a curious footnote of trivia to the natural history of insects, this presents a classic, but serious evolutionary problem. While these losses could be easily explained if these species all belonged to some single, aberrant lineage of crickets Ð indicating a single past occurrence of loss which was simply inherited by all of the descendants -- it is clear that this is not the case. Instead, most non-acoustic species seem to have evolved from fully acoustic species, time and time again along multiple, separate branches of the cricket evolutionary tree6. Dan Otte estimates that independent evolutionary events resulting in the loss of acoustic communication and morphology in crickets have occurred at least one hundred times, worldwide7. Because some of these losses happened long ago, giving rise to entire families and genera of nonacoustic species, one hundred independent losses amounts to far more than one hundred species of deaf and dumb crickets.

6 Although a detailed evolutionary tree (phylogeny) of crickets has yet to be published, the affiliations between non-acoustic species and acoustic relatives are strongly supported by well studied morphological similarities and traditional classification. 7 Otte 1992

7 Evolutionary convergence is widespread in organisms Ð consider the wings of birds, bats, and insects, animals which do not share the same winged ancestor Ð but it is unsettling to observe over one hundred independent evolutionary events which have converged upon the same result in similar, but not closely related organisms.

Evolutionary convergence necessarily invokes the comparative concept of homology, the inherited ‘sameness’ of structure, pattern, or process between two or more species. When several species with some similar trait are presumed to have evolved from the same ancestor which also had the trait, the trait is said to be homologous among the species, since there is no reason to doubt that it was inherited from the ancestral stock. However, in the case of crickets, the absence of acoustic structures cannot be considered homologous, since the fossil record provides good evidence that the first crickets were acoustic insects, and that non-acoustic species could not have all evolved from a single, non-acoustic form.

This is where evolution gets interesting: without shared ancestry to explain similarity, one cannot help but wonder if there is some rhyme or reason to the observation that similar evolutionary outcomes have occurred over a hundred times independently. Is there some secret evolutionary mechanism which can steer unrelated species toward the same predictable outcome? Certainly not; evolution is a random, unpredictable process, which knows no purpose or direction. Could similarities between unrelated species be merely the results of chance? Quite probable, though this explanation is often unsatisfying to the strong intuition that something meaningful is going on in widespread cases of parallel evolution. Might there be something about the organisms themselves, something about their biology, perhaps Ð which makes them more likely to evolve along

8 a particular path, and more likely to undergo convergent evolution, arriving at similar yet independent outcomes? While it is fine to pose sweeping questions about the mechanics of evolution, they are generally too abstruse to be tackled without examining some model entity that the process alters and affects. For the study of convergence, the loss of acoustics in crickets has the potential for being a very useful model.

In the case of crickets, there may be something to the last conjecture above, that there are biological factors which predispose them to evolve along a certain path; or, more correctly, that the constraints of their evolutionary history prevent them from taking any evolutionary path which is biologically impossible for their pedigree. Evolution can only tinker within the inherited biological tools and developmental rules which are already present in an organism, which is why there is no such thing as a human with antlers, a fish with a navel, or a snail with hands. Even the most severe mutation cannot turn a cricket-like creature into an un-cricket-like one, and this principle provides a major clue as to what might be happening when acoustic traits are lost. The simple truth is that a cricket which lacks the organs for singing and hearing is really not un-cricket-like at all: it is merely a baby.

DEVELOPMENT, GROWTH, AND HETEROCHRONY

Unlike holometabolous insects (such as beetles, butterflies, and ants), crickets are hemimetabolous, meaning that the young are born looking very much like miniature versions of the adults. Whereas holometabolous insects hatch as larvae and pupate into radically different adults, hemimetabolous insects go through less dramatic changes, molting their exoskeletons several times and growing larger between molts until the adult

9 stage is achieved. In the juvenile stages of most hemimetabolous insects, adult structures

(such as wings) may be poorly developed or even missing up until the final molt. As it turns out, both of the structures needed for acoustic communication in crickets, specialized forewings and auditory organs, are traits that do not appear until the termination of growth at the final molt8. The hypothesis suggested by this is that non- acoustic species could be species in which development has been arrested, without affecting the ability to proceed to reproductive maturity. If this truncation of growth could be genetically ‘fixed’ and inherited from one generation to the next, the result would be a juvenilized species which is deaf and dumb, even in the adults. Suddenly, those independent, convergent losses would seem much less mysterious, if some simple, common mechanism could account for them all.

Such a mechanism is well known in evolutionary biology, and has even been called one of the most important processes in the evolution of form. Evolutionary changes in the timing of development and growth, a process known as heterochrony, seem to be responsible for morphological change in a diverse number of species in the animal kingdom. Although heterochrony refers to both paedomorphosis (the abbreviation of development) and peramorphosis (the advancement of it), most case studies have focused on species in which the adults are paedomorphic, clearly resembling the juvenile stages of ancestral or related forms.

The classic critter in the study of paedomorphosis is the Mexican axolotl, a giant salamander of the genus Ambystoma which also includes the familiar tiger salamander of

North America. While most amphibians possess gills as larvae and metamorphose into

8 Huber et al., 1989

10 gill-less, terrestrial adults, axolotls retain their gills and an aquatic existence throughout their lifetimes, growing larger and reaching sexual maturity as if metamorphosis were never an issue. However, the developmental potential to lose the gills and become a terrestrial adult is still latent in the axolotl’s genome. On rare occasions, prompted by a critical change in diet or water quality, axolotls will behave like ordinary salamanders, undergoing standard metamorphosis into terrestrial, gill-less forms. In the laboratory, metamorphosis can also be artificially induced by the introduction of thyroxine, a thyroid gland secretion which is a key regulator in amphibian metamorphosis. Although metamorphic axolotls are poorly studied in nature, it is presumed that they produce offspring which revert to the original, gill-retaining, water-dwelling forms.

A more extreme case of paedomorphosis is found in Perennibrachiate newts, smaller salamanders which, like the axolotl, retain gills and an aquatic existence throughout adulthood; but unlike axolotls, are never known to metamorphose into terrestrial forms, even by artificial stimulation. In these species, paedomorphosis has become permanently fixed in the genome, resulting in reproductive adults which resemble the larvae of other salamanders. In 1951, the developmental biologist Walter

Garstang immortalized these evolutionary enigmas in pithy verse:

Ambystoma's a giant newt who rears in swampy waters, As other newts are wont to do, a lot of fishy daughters: These Axolotls, having gills, pursue a life aquatic, But, when they should transform to newts, are naughty and erratic.

They change upon compulsion, if the water grows too foul, For then they have to use their lungs, and go ashore to prowl: But when a lake's attractive, nicely aired, and full of food, They cling to youth perpetual, and rear a tadpole brood.

11 And newts Perennibrachiate have gone from bad to worse: They think aquatic life is bliss, terrestrial a curse. They do not even contemplate a change to suit the weather, But live as tadpoles, breed as tadpoles, tadpoles altogether! 9

Although paedomorphosis has been uncovered and studied in a great number of other organisms from several animal phyla10, a final example worthy of mention is the primate species Homo sapiens. The physical appearance of humans is markedly different from those of the other great apes, due to a number of characteristics including, but not limited to: skulls with flattened faces, the reduction of body hair, the shape of the pelvis, and a high brain weight relative to body size. However, since the early part of the 20th century, biologists have realized that these and dozens of other human physical characteristics, which at first seem to set us apart from the other primates, are actually congruent with the physical characteristics of baby gorillas and chimpanzees -- our closest living relatives.11 A strange notion, to think that a species which is normally considered to be more ‘advanced’ than its kin may in fact be a result of juvenilization!

One simple way that heterochrony could play a role in the evolution of crickets would be if there were some genetically-controlled event in a cricket’s life which

‘switches on’ the development of adult structures at the final molt. Such an event might be biologically easy to impede, and would have major consequences for adult morphology. As it turns out, such a switch actually occurs, a switch in the concentrations of certain morphogenic (form-producing) hormones in the blood of all insects. In most

9 Garstang 1951 10 Reviewed in Matsuda 1987

12 , there exist several hormones which are present in high concentrations during immature stages. Appropriately called “Juvenile Hormones”, they appear to keep insects in a physically juvenile state throughout most of their growth cycles. Enocrinological studies have shown that Juvenile Hormone concentrations plunge dramatically just before the final molt into the adult stage, when strictly adult structures appear. Experimentally, the artificial introduction of Juvenile Hormones, or the removal of the glands which produce them, have been shown to hinder or accelerate the development of adult structures, respectively.12

Juvenile Hormone concentrations have also been found to be extremely responsive to environmental cues, such as temperature and photoperiod, a phenomenon which has been observed repeatedly under controlled laboratory conditions.13 It is widely believed that the sensitivity of morphogenic hormones to environmental conditions is an ancient mechanism which synchronizes the maturation of insects with the onset of summertime, the most favorable season for flight and reproduction. In a habitat where winter-like conditions are maintained Ð a combination of low light and low temperature -- it seems reasonable to expect to find insects which fail to experience a drop in Juvenile Hormone, and the subsequent development of adult structures. Curiously, it is true that almost all cricket species which show the reduction or absence of acoustic structures are species which live in burrows, under bark, in leaf litter, and other places where the levels of temperature and sunlight would be expected to be low. The most extreme type of habitat

11 Reviewed in Gould 1977 12 For crickets, see Zera & Tiebel 1989; Tanaka 1994; Zera & Tanaka 1996; Roff et al. 1997 13 Ibid.

13 in which environmental cues would be totally lost is in caves, where year-round darkness and constant cool temperatures simulate a world where summer never comes.

MADAGASCAR

In 1998, I traveled to Madagascar to investigate what is arguably the most extensive radiation of cave crickets anywhere in the world, the ancient cricket family

Malgasiinae, containing the single genus Malgasia which lives nowhere on earth but

Madagascar and nearby islands. Although several species of cave-dwelling crickets are known from various locales around the globe, the Malgasiines are the only cricket family in which every known species appears paedomorphic. All Malgasiine species are completely devoid of acoustic structures, as well as other structures which are typical of adult morphology, such as hindwings used for flight, and the three ocelli (simple eyes) which are situated between the compound eyes of most crickets. Although only 14 species have been described to date, prior workers in Madagascar14 found that almost every cave harbors a different species of Malgasia, which lives in that particular cave and nowhere else on the island. Without question, the dearth of described species in this family is only a result of the fact that Madagascar is about as far away as one can possibly get from the United States, and it is no easy (or safe) task to locate and explore caves in a developing nation. Intrigued by these poorly studied crickets, I went to

Madagascar to find them, and to see if I could discover any exceptions to the paedomorphosis which seems to be the rule for this family.

14 Primarily Msr. Lucien Chopard, who was an entomologist at the Paris museum in the mid-1900’s.

14 Madagascar is exactly the kind of place which can divert cricket-hunters towards more charismatic organisms; perhaps that is why its cricket fauna, though rich, is almost completely unknown. It would be difficult for anyone with an appreciation for nature to resign themselves to searching for drab little crickets in dark, dreary, caves, on an island where the forests, deserts, mountains, and scrubland abound with beautiful and bizarre creatures which live nowhere else: special creatures such as lemurs, both nocturnal and diurnal; chameleons, with their fabulous colors and independently orbiting eyes; swift geckoes, large and small, iridescent and cryptic; strange insectivores like the tenrec, which is said to be good in a stew; and the giant predatory fossa, more like a mongoose than any other kind of mammal, and the top of the food chain on the island. In a land where stunning wildlife is easier to find and observe than the squirrels in Harvard Yard, I spent much of my time on my belly in the dark, crawling through dirt and bat guano and prime habitat for scorpions, in search of a cricket which makes no sound. The one incident which nearly caused me to hang up my headlamp for good occurred in a particular cave which is on no map (for all I know, I was the first foreigner to enter it). I was squeezing through a long, tight passageway when my guide’s torch, a flaming branch of dead palm leaves, went out, filling the corridor with thick white smoke. At once I lost both my vision and my ability to breathe, and started getting bombarded by bats from the other end of the corridor, in their own panicked attempts to escape the smoke; all of this, in a cave which turned out to be the only one in eleven weeks that did not contain representatives of Malgasia.

In the end, I found no exceptions to the paedomorphosis which is typical of

Malgasiines, but I did discover several new species, some of which are not cave-dwellers

15 at all. The three or four surface-dwelling species of Malgasia I collected may be able to provide a conclusive answer to the history and causes of the loss of acoustics in this family. Alcohol preserved specimens can yield good DNA that can be sequenced, and these sequences can be used to construct a phylogenetic tree, which shows the historical evolutionary relationships between species. If an evolutionary tree of Malgasia shows surface-dwelling species to be ancestral to cave species, then support for environmentally-induced heterochrony is lessened, since Malgasiines would have been paedomorphic before their radiation into caves. However, an evolutionary tree which shows that cave species are ancestral to surface-dwelling species would be consistent with an evolutionary scenario in which adult structures were lost due to the effects of the environment on development. Additionally, the presence of paedomorphic surface- dwellers would affirm that heterochronic changes can become genetically fixed (as in

Perennibrachiate newts), so that no subsequent change in lifestyle can restore the ancestral acoustic form.

Even if the environmental conditions of caves were not responsible for inducing paedomorphosis in Malgasia, there is still a very strong tendency for cave dwelling species around the world to be the most extreme forms of wingless, deaf, and nonacoustic crickets. In the Oecanthines, a family of tree crickets which are loud singers and good flyers, the only completely wingless and nonacoustic species are found in the cave-like lava tubes of Hawaii. Another family, the Phalangopsinae, has several cave-dwelling species which have lost both their flying hindwings and their acoustic structures (one of my goals while in the Colombian Amazon was to collect forest-dwelling representatives of this family). Then there are the multitudes of species which live in cool, dark

16 microhabitats in forests and underground, which show varying degrees of partial loss of wings and acoustic structures. Any of these species may be the evolutionary products of paedomorphosis, though there are other hypotheses which favor adaptive reasons why acoustic communication has been lost in these species Ð for example, the conjecture that sound does not travel well in caves15, or that predators such as bats and birds, which can orient to sound, have favored the evolution of silent forms16. Overall, the current consensus is that singing behavior will be lost when its costs outweigh its benefits, and that the loss of acoustic structures will eventually follow when they are no longer needed; the process of heterochrony has not yet been recognized or even suggested as an active process in the evolution of crickets. In fact, heterochrony is rarely cited in the evolutionary studies of any model systems, even though it is perhaps invoked a bit too often in vertebrate systems. My dissertation research represents an attempt to bring a new evolutionary perspective to some old evolutionary questions.

A SEARCH FOR THE EVIDENCE

When testing a new hypothesis, such as the speculation that heterochrony can facilitate convergent evolution in similar species, a useful approach is to determine what kinds of evidence would be predicted if the hypothesis were true, and then to see whether the evidence exists. If the loss of acoustic structures is caused by early truncation of development Ð i.e., if deaf and dumb species exist due to evolutionary paedomorphosis rather than adaptive reasons -- then there should be other typically adult structures which

15 Wiley & Richards 1978, Paul & Walker 1979 16 Walker & Masaki 1989

17 are also reduced or absent in the species. Two candidate structures have already been mentioned, both of which are found only in adults which have completed the final molt: the hindwings, which are not used in stridulation but can be used for flight; and the ocelli, simple eyes situated on the forehead between the compound eyes. Other traits which may provide information about developmental maturity are traits which go through slight changes throughout the juvenile molts until they arrive at a certain advanced stage in the adult. As crickets develop and grow, the compound eyes tend to add more subunits

(‘facets’); the genitalic organ in males tends to go from a bifurcated form to a single fused structure; and female ovipositors and the antennae in both sexes tend to get longer relative to body size17.

The importance of examining traits which are not involved in acoustic communication lies in the fact that natural selection can simultaneously affect all of the traits associated with some function, causing them to co-evolve as a single unit. For this reason, it is not enough to see whether losses of stridulatory structures are correlated with losses of hearing organs, because the two are functionally interdependent, and the loss of one would eventually imply the loss of the other. However, if acoustic structures covary with functionally unrelated structures, no argument can be made for the co-evolution of functionally interdependent traits, favoring instead a hypothesis that invokes a process which affects the entire organism, such as heterochrony. In truly paedomorphic species, one would expect to see reduced or missing acoustic structures in some combination with one or more of the following: reduced or missing hindwings, reduced or missing ocelli,

17 Snodgrass 1937; Matsuda 1979, 1987

18 fewer subunits in the compound eyes, shorter antennae, shorter female ovipositors, and/or bifurcated male genitalia.

This is where the work of the globetrotting cricket-hunters becomes priceless.

Although most of them were primarily concerned with the discovery and description of new species, they recorded a wealth of detailed anatomical information in the process of doing so Ð information which can be retrieved from a zoological library and used as data.

For any given species, it is relatively easy to find the original published description, which will contain the minutest details of most or all of the traits of interest. To date, I have been able to construct a table for a great number of cricket species (excerpts in

Table 1), which codes the traits listed above as being adult-like or paedomorphic Ð simply by sifting through the original taxonomic literature. The results have been stunningly promising: there are clear, strong patterns of trait correlation which are fully consistent with my hypothesis of evolution by heterochrony.

However, there is another problem when examining the co-evolution of organismic traits, and once again, it is the problem of common ancestry. It would be senseless (and wrong) to conclude, for example, that the wings of bats are significantly correlated with the fact that bats are covered with hair. While it is true that every bat has both wings and body hair, a correlation between these traits is meaningless, since they are simply traits which were inherited from ancestors: wings from ancestral bats, and body hair from ancestral mammals. The presence of one trait does not imply or predict the presence of the other. In the study of evolution, meaningful correlations of traits are correlations which tend to appear repeatedly among distantly related species Ð i.e., in species that could not have inherited the correlated traits from the same ancestors Ð and

19 which cannot be explained by chance alone. The best way to prevent the observation of meaningless correlations of traits is by possessing a phylogeny, a branching, tree-like diagram which estimates the historical evolutionary relationships between species, and traces the paths of ancestry and descent. With a phylogeny, it is clear which species share common ancestors, and which are more distantly related. More importantly, there are a number of statistical tests which, when applied to a phylogeny, can determine whether a particular correlation of interest occurs significantly more frequently than would be expected by common ancestry and chance18.

Phylogenies can be constructed in a number of ways, but the most widely used approach is to encode as much information as possible about the traits of the species of interest, and then run computerized algorithms to search for the trees which are most consistent with the distribution of traits among the species. Because the human mind can encode physical information in different ways based on perception and opinion -- ‘short’ antennae to one researcher may be ‘medium-length’ antennae to another -- it is often preferable to use data from traits which are not open to interpretation. The sequence of nucleotides in an organism’s DNA is one form of information which is unambiguous, yet which evolves at a somewhat constant rate such that closely related species will tend to have more similar DNA sequences to each other than to more distantly related species.

Obtaining DNA sequences from organisms is relatively simple, provided that fresh or well-preserved specimens are available; a potential drawback if one is interested in species which do not live near the research institution, but nothing a plane ticket, a pair of hiking boots, and some specimen vials can’t fix.

18 Felsenstein 1985; Maddison 1990

20 AN ANTICIPATED DEFENSE IN THE KALAHARI

In the spring of 2001, I will be joining a small group of researchers including Dr.

Dan Otte, the world’s greatest living cricket-hunter, on a specimen collecting trip throughout southern Africa. Beginning our trip in Johannesburg, the plan is to rent a 4x4 vehicle and slowly make our way across South Africa, Botswana, and Namibia, searching for crickets and other insects along the way. Otte’s specimens will be deposited in research collections and described in future publications; mine will be homogenized to extract their DNA for a phylogeny. This will be my first time meeting Otte, a man who can identify most any cricket from anywhere in the world in a matter of seconds (chances are one in four that he was the first to discover it in the first place) and who is said to be proficient in the Zulu language. Many of the landmark publications on acoustic communication in crickets are his, publications which include sound oscillograms of songs he has recorded in the field, along with fascinating meditation on how cricket songs evolve19. Although he has not closely studied the convergent losses which interest me, he subscribes to the view that there are adaptive reasons for the loss of stridulation and auditory organs: predators, population structure, behavioral dynamics between the sexes, and the ecology of habitats. I am apprehensive about sharing my views with him, since heterochrony has not yet been proposed as a prominent evolutionary process in crickets, or even in many other insects. However, for the sake of conversation while crossing the Kalahari Desert, I will probably confess these ideas, and more, perhaps in words such as the following:

19 A comprehensive review of the evolution of cricket song can be found in Otte, 1992

21 Forget about silent crickets for now, and consider the ones that sing. We all know that sexual selection is a central theme to the study of song evolution: males evolve to sing a particular song because females won’t mate with them unless they do. But surely, the morphology of the forewings has some control over the song which is produced.

Frequency, amplitude, and rate must all be constrained by the physical characteristics of the structures that produce them. Short, reduced forewings have to be acoustically different from long forewings. Doesn’t there need to be a mechanism which regulates the physical evolution of acoustic structures, even in the presence of selective pressure from females? Even adaptations and sexual selection require biological processes to actualize their results. If the physical evidence supports it, why not heterochrony?

Paedomorphosis doesn’t need to result in the total loss of acoustic traits; truncation of growth could occur at any stage. Would it be heresy to suggest that the tremendous variation in acoustic structures and song parameters in all crickets could be the result of the speeding-up or slowing-down of the developmental process?

Forget about songs for now, and consider species. The choosiness of females keeps species together, but where do those species come from? If heterochrony can produce a new kind of structure and a new kind of song, couldn’t those two things result in a new kind of cricket? A good number of evolutionary biologists have argued that heterochrony may be one of the most important causes of the appearance of new species.20 If insect development is so responsive to environmental conditions, as we know it to be – then couldn’t the spectrum of habitat types cause the appearance of the first

20 Novak 1966; Gould 1977; Matsuda 1987

22 new forms? And if the new forms have new songs, and the females are picky, couldn’t this set the stage for the genesis of a new species?

Forget about species – let’s talk convergence. People are amazed when evolution produces the same results in unrelated organisms. It must be chance or it must be magic, since evolution cannot know what its left hand is doing. But if there is no homology in independent evolutionary events and outcomes, could there be homology in the biological pathways which produce them? All crickets develop and grow following very similar sets of rules, which were undoubtedly inherited from the same ancestral stock. Is it a paradox to consider that non-homologous forms in convergent evolution can be based on a homologous process?

Hush... I hear a cricket. Like any other creature with a voice, crying softly: “Here

I am, here I am,” to anyone who will listen.

23 Taxon A B C D E Taxon A B C D E Gryllotalpinae: Oecanthinae: Scapteriscus 1 0 3 0 0 Oecanthus 0 0 0 0 * Scapt. abbreviatus 2 1 3 2 2 Xabea 0 0 0 0 * Gryllotalpa 2 0 3 0-2 2 Neoxabea 0 0 0 0 * Neocurtilla 2 0 3 1 0 Prognathogryllus 1,2 0 3 2,3 * Myrmecophilinae4 1 4 3 2 Leptogryllus 3 1 4 3 * Malgasiinae 4 1 4 3 2 Thaumatogryllus 3 1 4 3 * Mogoplistinae: Eneopterinae: Arachnocephalus 4 1 4 3 2,3 Tafalisca 1 1 4 1 2 Ornebius 2 0 3 3 2 Paroecanthus 0,1 0 0,2 0,1 0 Ecatoderus 2 0 3 3 2 Amblyrhetus 0 0 0,2 0 1 Cycloptilum 2 0 3 3 2 Salmanites 1 0 2 2,3 * Mogoplistes 4 1 4 3 3 Nisitrus 0 1 2 0 0 Australian genera 2 0 3 3 2 Myara 1 0 2 2,3 * Pteroplistinae: Agnotecous 1 0 2 3 0 Pteroplistini 0,1 0 3 0 1 Diatrypa 0 0 0 0 2 Odontogryllini 4 1 4 3 1 Eurepa 1 0 2 2,3 * Landrevini 2,3 0,1 0,3 3 0,1 Eurepella 1 0 2 2,3 *

Table 1: Some examples of diversity in acoustic and other structures of crickets. Only seven subfamilies and five traits are outlined here -- the actual data matrix includes data for 10 traits, for a diverse sample of 200 genera from all 18 cricket subfamilies. Higher numbers for each trait indicate more reduced (juvenilized) structures. Rows with several high character scores may indicate paedomorphosis in the taxon (e.g. Myrmecophilinae, Malgasiinae, Arachnocephalus, Mogoplistes, Odontogryllini, Leptogryllus, Prognathogryllus, Thaumatogryllus, Tafalisca).

A: Forewings 0 Ð longer than body; 1 Ð slightly shorter than body; 2 - much reduced; 3 - vestigial; 4 - absent. B: Stridulatory morphology 0 - present; 1 - absent. C: Auditory organs 0 - present on both sides of the foreleg; 1 - present on both sides but smaller on the outer side; 2 - present on outer side only; 3 - present on inner side only; 4 - absent. D: Hindwings 0 Ð longer than body, functional for flight; 1 Ð slightly shorter than body, not functional for flight; 2 - vestigial; 3 - absent. E: Ocelli 0 - normal; 1 - reduction or absence of middle ocellus; 2 - all three ocelli reduced; 3 - all three ocelli absent. ( * indicates information on ocelli not recorded in the original genus description; actual specimens will be examined in order to collect any data such as these which cannot be found in the literature.)

24 WORKS CITED

Felsenstein, J. 1985a. Phylogenies and the comparative method. Amer. Nat. 125: 1-15.

Garstang, W. 1951. Larval Forms With Other Zoological Verses. Basil Blackwell, p. 62.

Gould, S.J. 1977. Ontogeny and Phylogeny. Belknap, Harvard University Press, Cambridge, MA.

Huber, F., T. E. Moore, and W. Loher, editors. Cricket Behavior and Neurobiology, Cornell University Press, Ithaca, NY.

Maddison, W.P. 1990. A method for testing the correlated evolution of two binary characters: are gains or losses concentrated on certain branches of a phylogenetic tree? Evolution 44: 539-557.

Matsuda, R. 1979. Abnormal metamorphosis and evolution. In Arthropod Phylogeny, A.P. Gupta, Ed., Van Nostrand Reinhold, New York, NY.

Matsuda, R. 1987. Animal Evolution in Changing Environments with Special Reference to Abnormal Metamorphosis. John Wiley & Sons, New York, NY.

Novak, J. J. 1966. Insect Hormones. Methuen, London, UK.

Otte, D. 1992. Evolution of cricket songs. J. Res. 1: 25-48.

Paul, R. C. and T. J. Walker. 1979. Arboreal singing in a burrowing cricket, Anurogryllus arboreus. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 132: 217-223.

Roff, D.A., Stirling, G., and D.J. Fairbairn. 1997. The evolution of threshold traits: A quantitative genetic analysis of the physiological and life-history correlates of wing dimorphism in the sand cricket. Evolution 51: 1910-1919.

25 Shaw, K.L. 1996. Polygenic Inheritance of a Behavioral Phenotype: Interspecific Genetics of Song in the Hawaiian Cricket Genus Laupala. Evolution 50: 256-266.

Snodgrass, R.E. 1937. The male genitalia of orthopteroid insects. Smiths. Misc. Coll. 96: 1-107.

Tanaka, S. 1994. Endocrine control of ovarian development and flight muscle histolysis in a wing dimorphic cricket Modicogryllus confirmatus. J. Insect Physiol. 40: 383-490.

Walker, T.J. and S. Masaki. 1989. Natural History. In Cricket Behavior and Neurobiology, eds. F. Huber, T. E. Moore, and W. Loher. Cornell University Press, Ithaca, NY.

Webb, K.L. and D.A. Roff. 1992. The quantitative genetics of sound production in Gryllus firmus. Animal Behavior 44: 823-832.

Wiley, R.H. and D.G. Richards. 1978. Physical constraints on acoustic communication in the atmosphere: implications for the evolution of animal vocalizations. Behav. Ecol. Sociobiol. 3: 69-94.

Zera, A.J. and K.C. Tiebel. 1989. Differences in juvenile hormone esterase activity between presumptive macropterous and brachypterous Gryllus rubens: Implications for the hormonal control of wing polymorphism. J. Insect Physiol. 35: 7-17.

Zera, A.J. and S. Tanaka. 1996. The role of juvenile hormone and juvenile hormone esterase in wing morph determination in Modicogryllus comfirmatus. J. Insect Physiol. 42: 909- 915.

26