Fieldwork: a game of chance, a test of will

As we are driving back from the Altai toward Novosibirsk, I’m thinking about how lucky we were to have found two of my important target spider species, Pellenes gobiensis and Pellenes stepposus. In both cases what seemed like bad luck turned into good luck.

We’d stopped along the Buguzun River because we suspected Pellenes gobiensis might live there, but as the rain fell, I thought we’d missed our chance to find them. I was surprised when Galina came to my tent to say, “I found a Pellenes!”. Galina had figured out how to find them in the rain, nestled in little camouflaged retreats under rocks out on the steppe. Here is a female:

Pellenes gobiensis, female. The dark markings under her eyes, “tears”, are seen in females of various other species of Pellenes and Habronattus.

Here’s the twist of luck: rain not only didn’t prevent us from finding this species, but it likely helped us. Our subsequent searches turned up many specimens, but all were adult females in nests with babies. This means that they likely wouldn’t have come out onto the rocks even if it had been sunny. And, if it had been sunny, Galina and I probably would have simply walked along, scanning the ground. Not seeing anything, we might have assumed it was the wrong habitat, and not bothered to turn over rocks, which is normally a low-success activity. The rain changed that: if we were going to look, we had no choice but to take the laborious route, and it worked.

In the case of Pellenes stepposus it wasn’t rain, but a delay, that led to their discovery. On the Kuray steppe we had pretty much finished our search for Pellenes and aelurillines, and were ready to move on to another site. But, one of the tasks of this expedition is to set up pitfall traps for an ecological study, and that task was taking longer than planned. I wasn’t involved in setting up the traps, but it meant I couldn’t leave yet. I looked around, exhausted, took a deep breath, and looked for something new to try. I saw some Artemesia along a stream and thought I might as well start looking on those, not expecting to find Pellenes (but perhaps some other interesting jumping spiders). Lo and behold, specimens of Pellenes stepposus started falling on my sheet.

Pellenes stepposus, female.

(By the way, P. stepposus is probably not actually a Pellenes, but most likely belongs in another genus like Sibianor. That doesn’t diminish its importance to me, as Sibianor specimens are useful as a more distant point of comparison with Pellenes and Habronattus.)

I think of both of these successes as due to a twist of luck, though you could argue in both cases the “bad luck” simply provoked us to work harder. I am reminded of the old saying: “I am a great believer in luck, and I find the harder I work, the more I have of it.”

Raising spiders on a diet of human

Rain brings most jumping spider hunting to a halt, at least in temperate areas. Perhaps they need the sun to bring their body temperatures up for their needed activities (hunting, developing, mating); perhaps the water itself is a threat. The spiders don’t come out, and it’s awfully difficult to find them in their little retreats. I imagine them sitting by the fire drinking hot chocolate. As I write this (14 July) I am, like the spiders, sitting in my little retreat (tent) by the Buguzun river in the Altai, waiting for the rain to stop and the sun to come out. This is a good time to feed spiders.

I am waiting for the mosquitoes to bite. In fact, I have my tent flaps open, to help them find me. You see, I need to feed the spiders that we hope to rear to adulthood, the most important of which for me are the subadult males of Pellenes logunovi we got at Tekelyu. The problem is that the mosquitoes are rather skinny, and I’d like the P. logunovi to have a big meal. But, mosquitoes can be quickly fattened up from the richest source of animal material nearby, namely, me.

When a mosquito bites me, I wait for it to start to fill up with my blood, then I gently put a vial over top of it. When it’s done, it tries to fly away but it’s trapped by the vial. Here’s one finished its meal of me, about to become a meal of a Pellenes logunovi:

Mosquito fat with my blood.

What a nice juicy snack for a spider! A few minutes later, the spider was feasting on my blood:

Spider eating mosquito and me.

It’s somehow pleasing to be lower on the food chain than my spiders, at least occasionally.

By the way, there are jumping spiders that appear naturally specialized on this sort of diet… Evarcha culicivora.

Aelurillina

I’ve looked forward to this Siberia trip for many reasons, one of which is the chance to learn about aelurilline jumping spiders from the world expert. Aelurillines have been the primary focus of study by my host in Siberia, Galina Azarkina, for almost 20 years. In North America we see them rarely, as only one native species lives there: the striped Phlegra hentzi, whose males have a pale blue face. It lives alongside the much more common and diverse Habronattus, on open sunny ground with dry leaf litter. In Eurasia and Africa, however, aelurillines are diverse and common, occupying habitats very similar to those of my familiar Habronattus back home.

In just a few days of collecting I’ve now seen 7 species of aelurillines: Phlegra fasciata and two Asianellus at Karasuk, three species of Aelurillus at Kuray, and an Asianellus at Kokorya. I must say, aelurillines are charming. Here is the most dramatic, a beautiful male Aelurillus from the Kuray steppe:

Striped Aelurillus species from Kuray, male.

He has especially long, spiny legs:

Striped Aelurillus species from Kuray, male.

I look forward to seeing more aelurilline diversity as we continue our explorations here in Siberia.

Sunbathing spiders

Pellenes, the target of my spider field work in Siberia, is somewhat of an exception among jumping spiders for its choice of habitat. Most jumping spider species wander about on plants, from grasses to trees, using their special abilities to see and jump to navigate and hunt in this 3 dimensional habitat. Not all inhabit the cool greenery, however. Some, including Pellenes and Galina’s aelurillines, hunt regularly on sun-baked ground. Here is Pellenes peninsularis sunbathing on a rock outcrop in Ontario, and an Aelurillus on the Kuray steppe of Siberia:

Pellenes peninsularis (left) and Aelurillus sp. (right) sunbathing on rocks.

They don’t sunbathe on rocks to get a tan. It probably helps raise the spider’s body temperature to develop, hunt, escape and court more quickly, but it may come with the risk of heat stress, dessication and visibility to predators.

Only a few evolutionary lineages of jumping spiders have braved a lifestyle on hot sunny ground. Five have done it and diversified in a big way. The sitticines of South America did it, and did it with enough success to spread all the way to Eurasia and diversify into many species there — although the Eurasian lineage has shifted to many other habitats, not just open ground. The harmochirines of Africa and Eurasia returned the favour, starting on sunny ground in the Old World before some “Pellenes” went to the Americas to diversify there as the paradise spiders, Habronattus. The peacock spiders of Australia (Maratus) similarly live on sunny ground, as do Yllenus and relatives of Eurasia, and the aelurillines of Africa and Eurasia.

How have these five lineages come to be similar, as they have adapted this new habitat? Apart from the females typically being quite cryptic — the same colours as the rocks they are on, for example — we don’t yet know the answer to that question. Here are some suspicions:

1. They are probably more heat tolerant than their relatives.
2. At least some members of each group have unusually long jumping legs (3rd legs in the case of Pellenes and relatives; 4th legs in the case of Sitticus and Yllenus).
3. Three of these groups bear male courtship ornaments that are spectacularly detailed and colourful (Maratus, Habronattus, and Stenaelurillus) — perhaps encouraged by easier vision in the high-light environment.

Given that Galina and I both study sunny-ground lovers, it’s not surprising that we make a good collecting team. In fact, the synergy that helped us get Pellenes logunovi has worked frequently: each of us has played a key role in the others’ successes on this trip. She was the first to find Pellenes logunovi and later, P. gobiensis (more on that species in another post). I was the first to find several of her species: two Aelurillus at Kuray and an Asianellus at Kokorya.

Diverse eyes: the nonlinear effects of collaboration

You might think that having two scientists makes discovery go twice as fast as one, that three makes it three times as fast, and so on. But, if there is a diversity of talents and inclinations, it’s far better than that. At Tekelyu Valley, our diverse eyes led us to the most important target of my trip to Siberia: Pellenes logunovi.

After our chilly first night camping on the alpine steppe of Tekelyu Valley, we greeted sunrise with anticipation. Once it was warm enough to hope the spiders would be out, Galina set off in one direction, I in another, to seek Pellenes logunovi. I had a particular image of the habitat I was looking for, short bushes with patches of rock between. At a nice south-facing slope I found two male Pellenes, to my great excitement. The spiders seemed too small, but otherwise from what I could see, they might be P. logunovi. I headed back to camp and took a look at one under a microscope — no, wrong Pellenes (P. limbatus). Still important scientifically, but not my dreamed-off P. logunovi. So I set out again to a slightly different habitat. Another male Pellenes, but P. sibiricus. By this time it’s after 11:00 am and getting hot. I’m getting discouraged, because soon it will be too hot and the spiders will be taking their siestas. It seemed we might miss our chance here to find P. logunovi.

And then Galina radioed — she’d found a female she thought might be P. logunovi. She was on the other side of the hill. Off I went on the long trek, and when I finally reached her, I saw that she was on a dry grassy hillside with small rocks, a rather different habitat from the ones where I’d been. She handed me the specimen and instantly I was 90% sure it was P. logunovi (and later, I confirmed it under the microscope). I was so excited.

More important than anything else, she had found the type of micro-habitat that Pellenes logunovi likes. A place like Tekelyu has dozens of different types of ground — with dry grass, or green grass, or moss, or Artemesia, or rocks, or wetter soil, or on a slope, or south-facing, or near pine trees, or some combination. A spider species might prefer a precise combination for its micro-habitat. The diversity of micro-habitats at Tekelyu was too great to survey completely in the time we had, so we had to focus on some best guesses. My inclinations led me to the wrong combination; Galina’s to the right combination.

On Galina’s hillside we didn’t find as many specimens as needed, but that wasn’t as important as knowing that they were at Tekelyu, and learning the habitat. If she hadn’t discovered them, we might have decided to leave that afternoon to look for another site to camp, and might have failed to find the species at all. Instead, we stayed through the afternoon and discovered that they were ridiculously common on rocks of a grassy gravel bar of the river below our camp. We now have enough specimens to do the research we want to do.

The exploration phase of science depends on diverse ideas and creativity, which can unlock hidden doors to progress. Having two searchers had a compound effect on our spider gathering: it doubled our chances of finding their habitat, and once that was found, we could put double effort into gathering specimens. Because both of us could follow leads that either one of us found, our productivity as a team was well more than twice what either would achieve solo. Galina and I have made a great team for several other species as well, though those stories are for later.

As I explained in a previous post, P. logunovi was my highest priority because it seems unique among Eurasian Pellenes in its similarities to the common group of species native to North America. That was judged based on pickled specimens in museums. Now, seeing them alive, I see how much they look like North American species. Here are photos, of P. logunovi at left, and the American P. longimanus at right.

Pellenes logunovi from Siberia and Pellenes longimanus from Texas.

We didn’t find any adult males of P. logunovi (it seems they aren’t in season), but we have many almost-adult (“subadult”) males that we will try to raise to adulthood. After only a few days, my 5 week long trip to Siberia is already a success.

Tekelyu

As I write this (12 July) I am in one of the most beautiful places I’ve ever been, an expansive valley full of solitude, held by ever-silent hills painted pale green and brown by the short alpine steppe vegetation. There are only a few trees, on the slope opposite me. In the distance to my right are horses silhouetted against the bright steppe valley beyond, still shadowed by the mountain behind me as the sun rises. I imagine that their ancestors might have grazed here thousands of years ago, perhaps even before being domesticated. I can see a small settlement with two yurts down-valley, near the gravel-lined river wandering through the broad valley floor. We are in the valley of the Tekelyu River north of Saylyugem Massif, near Russia’s border with Mongolia, at about 2200 m elevation.

View from Tekelyu Valley

We arrived here after a heroic drive by Ivan and navigation by Galina, by roads, almost-roads, and never-roads across the countryside, fording streams and shaking ourselves up. We came here to find one of the known localities of our target jumping spider species Pellenes logunovi, near Saylyugem Massif. A bit confused about which valley to follow, we waved down a man and two children on horseback, whose words encouraged us to simply settle down in this valley. Not exactly at the target locality, but close, and we hoped the habitat would be good.

I will say more in subsequent posts about the spiders we found here, but for the moment I will soak in the sweeping hills, the steppe, the chill night air, the sheep and horses passing up and down the valley like breezes. I may enter and leave this valley quickly, like a breeze, but I’ll always remember Tekelyu.

Karasuk

I have many memories of our three day trip to Karasuk, near the Kazakhstan border, but an atypically small proportion of them are about spiders. This is partly because there were so many new experiences for me: small Russian towns, roadside кофе shops, colourful markets, the steppes, Soviet monuments, kvass, foraged mushrooms, and a new pancake recipe. Contrasts are everywhere, from Ladas and rusting Soviet machinery to Mercedes and fancy new farm equipment. But another reason that most of my memories are non-arachnological is, in fact, that there weren’t that many spiders.

We’d had four species as our primary targets: one species of Asianellus for Galina, and three species of Pellenes for me. All are known to occur in the area. After the first hour on the steppes, I was worried; the habitat seemed empty. In fact, in two full days of collecting around Karasuk, we found only a single Pellenes specimen and 2 juvenile specimens of the Asianellus.

I’d expected that we might be late in the season, and so perhaps the July population consisted mostly of adult females hiding with their babies. Even still, I would have expected to see more Pellenes, at least juveniles or stragglers. It was a bit spooky. But, the winter and spring were unusual climactically (the former warm but fluctuating; the latter wet and late), and so perhaps this was as an unusual year. The habitat looked just right for Pellenes.

We did find some other useful jumping spiders, however, including two species of Sitticus for a chromosome and phylogeny project. We also found Phlegra fasciata, whose males are smouldering embers and females are racing-striped. Handsome, and they do move quickly.

Phlegra fasciata male (left) and female (right), from the Karasuk steppe

Courtship ornaments as adversarial patches

When I read a few months back about machine language research into tricks that would deceive image-recognition systems, I immediately thought about spider courtship. The research looked into “adversarial patches” (summary; original), little pieces of an image that could be inserted (imagine a sticker placed on an otherwise identifiable thing) that would make machine learning algorithms go awry, misidentifying the overall image. The example used was an inserted patch that would make an algorithm mistake a banana for a toaster. If done subtly, the adversarial patch wouldn’t even be noticed by we humans, but the image recognition system would seize some special concocted detail and become delusional, thinking the image was of a toaster.

So what does this have to do with spider courtship? Evolutionary biologists have long wondered why males (usually) in many animal species have elaborate or exaggerated features that they display to the female in courtship. One long-popular theory suggests that these features indicate the male’s good genes: “I have such good genes that I can afford to divert my resources to conspicuous frippery” (picture an impractical muscle car). Females that choose those males give an advantage to their children’s genes, which will be thereby be paired with good genes, getting a boost in future generations. Recently, though, there’s been a lot of talk about an antagonistic arms race between males and females for control of reproduction. Perhaps these courtship features don’t represent high quality of the male, but simply take advantage of a susceptibility in the female. Just like an adversarial patch.

Suppose a female has evolved to recognize a potentially threatening wasp (= toaster) and to recognize pestering males (= banana). If the male could have a courtship ornament that makes her think, at least momentarily, that he’s a wasp, she might freeze in place rather than reject him instantly in exasperation. Freezing her there might give her brain time to notice his redeeming qualities. If you look at any of the images of dancing peacock spiders (Maratus), you’ll see how many of them look like big insect heads. Many species of paradise spiders (Habronattus), which I talked about in my last post, have strange knees that look a bit like small insect heads. Here are the knees of various species of Habronattus:

Third leg of male Habronattus captiosus, H. cuspidatus, and H. viridipes. The male wiggles them when doing his courtship dance to the female.

These knees may not look convincing as insects, but search for images of “peacock spider” and you’ll see insect-like displays. If it seems ridiculous that males would have body parts mimicking insects, it actually represents a respected idea in the field, called “sensory exploitation”, in which the male mimics something the female is already attuned to such as prey or predator. So, the idea of adversarial courtship ornaments already exists in the literature.

Where machine learning research comes in is that it might help us understand the evolution of these ornaments. Learned machines may not have reached the subtlety of a human brain, but they might have reached that of a spider’s. Findings in machine learning research may illuminate how tiny spider brains handle the complex images their sharp eyes supply from the messy world. For instance, could susceptibilities in our current machine learning methods help us understand the susceptible cognitive weaknesses that affect the spiders and their evolution?

So, here are some questions that biologists studying visually-mediated courtship might ask of the field of machine learning (“ML”):

1. Do the successful features of adversarial patches help us understand why spiders have such complex ornaments? One of the outstanding puzzles in sexual selection theory is to explain highly complex courtship. If each detail signals a male’s quality, so many quality indications would come at a high cost (cost is part of the theory). If instead the ornaments are mimicking a female’s sensory target, perhaps the complexity is needed simply to get the mimicry right. Or, could there be some direct advantage of complexity? Do ML image classifiers get distracted by edges, and the way to make an adversarial patch is to fill it full of many edges, with contrasting colours, and possibly a higher level pattern (like a sense of stripes)?
2. Do the details tend to matter, i.e. do small changes in an adversarial patch render it ineffective? If any slight change in detail loses effectiveness, then that could explain uniformity of ornaments in a species.
3. For any trained ML image recognizer, is there a large series of very different patch forms that would succeed at being adversarial? This, combined with details-do-matter, would imply an uncanny valley: slight changes from a good patch fail, but big changes might succeed, by hitting a different style of patch. This could explain the peculiar observation that some Habronattus have very uniform ornaments within a species, and yet seem to be susceptible to hybridization with rather different-looking species. Thinking of the sensory bias theory of sexual selection, perhaps the shared ancestral susceptibility isn’t a single peak in an optimality landscape, but a set of peaks, with different species settling on different ones — all species are susceptible to all of these many different peaks, but there’ll be a penalty for any male with a slight change from the current peak on which a population sits.
4. How does susceptibility to adversarial patches scale with neural net size and with input size? Does the ratio matter; i.e. are Habronattus and Maratus susceptible because their eyes or their environment give them too much information to process — many pixels being input but only small brains?
5. Does the classifier’s number of functionally-different targets among which it has to distinguish (bananas, toasters, cars, puppies, etc.) have an effect on its susceptibility? If the female spider’s brain has to distinguish ants from wasps from prey from sticks from males, will be she more susceptible than females in environments without ants or wasps? (This might combine with the previous question to be about dimensionality. In other words, is there some equation like susceptibility = (pixels * targets)/neurons?)
6. Does susceptibility rise or fall if there is a stronger requirement to generalize (i.e. visually very different objects are all generalized as “prey”; very different objects needing to elicit the response “sit still”)? This could add a different sort of dimensionality.
7. Does re-training an ML image classifier to avoid susceptibility to an adversarial patch tend to open up a new vulnerability? In a small neural net with only so many “neurons” to go around, can you get a whack-a-mole scenario where retraining to avoid being fooled by adversarial patch A opens up susceptibility B, retraining against which opens up susceptibility C, and so on? This could generate male-female arms race leading to constant change, or perhaps ever accumulating complexity.

It may be that the answers to these questions are already in the machine learning image recognition literature. If so, let’s talk!

Paradise spiders

I’d like to propose a common name for my — I hate to say this — favourite jumping spiders, Habronattus. Truly, it’s like saying which child I like best, except that the other jumping spiders won’t complain. I’ve travelled all over the world, but never fallen for any jumping spiders as I did, when I was young, for the beautiful Habronattus of North America. They are the paradise spiders.

I’ve long described Habronattus as being like little birds-of-paradise, with glorious courtship ornaments — colourful and intricate — and complex dances. Like the birds-of-paradise, the species of Habronattus have strikingly diverse ornaments and dances. While the internet-famous peacock spiders (Maratus) of Australia focus attention more on their rear ends, just like peacocks, different species of Habronattus have ornaments in many different places: on the underside of their first pair of legs, or on the front or top of their first legs, or second legs, or third legs, or their faces, or palps, or top of the head, or top of the abdomen, or underside of the body. In my post on reliving my youth in Siberia, I showed the bright green first legs and reshaped third legs of Habronattus viridipes. There are species with faces like gold leaf, ice blue legs, swollen cheeks, concave pink-green legs, big eyebrows. Here is a small sample of other ornamented Habronattus, in drawings I made long ago of some Canadian species:

Paradise spiders.  Ornaments of different styles on different body parts in Habronattus. Missing are the most complex species, with ornaments on the face, first and third legs, because they’re covered in other posts. Some colours not obvious here: the jaws of H. altanus glisten metallic violet, not visible here because the drawing is faded. The hairs at the tip of the silky-fringed leg of H. hirsutus are dark blue. The abdomen of H. decorus has a metallic pink sheen.

Let’s continue this contrast of Maratus as peacocks versus Habronattus as birds-of-paradise. (The Maratus fans might object to some of this, but I ask, how many of you have been in love with Maratus since 1971?) Now that we are learning more about Maratus diversity, we can see how amazingly complex is their courtship. Habronattus, though way behind Maratus in intensity of colours, arguably exceeds Maratus in complexity and diversity of motions and body parts emphasized. When Damian Elias and colleagues described the courtship of Habronattus coecatus, we wrote of it as like a symphony, with movements, themes, motifs. We noted at least 18 distinct courtship ornaments scattered over the body. We divided the courtship into many stages within which were different movements of the first legs, palps, third legs, and body, along with complex songs. It was astounding to us that a 5 mm long creature used such complex communicatory behaviour.

However you balance Maratus and Habronattus, they are both stunning evolutionary radiations of wondrous creatures. Habronattus, as far as I can see, is missing a common name. “Spiders-of-paradise” is a mouthful, and I don’t want to have to type so many hyphens, and so I’ll suggest the more compact form, “paradise spiders”. That’s my proposal. May there be a teenager somewhere in North America currently beholding a paradise spider, entranced.

Shostakovich’s breakfast

As I write this I’m flying over Mongolia on s7 airlines, listening to Shostakovich’s 5th symphony, having just finished the spectacular consequence, pictured below, of my Beijing airport fiasco. Do not, repeat, do not expect that a two hour layover in Beijing is enough to make the connection to Novosibirsk if you booked the legs separately and need to transfer the luggage yourself. The lost connection forced me to stay the night in Beijing (which was a treat) and fly the next day (today) to Novosibirsk, upgraded to business class.

Breakfast, business class, s7 Siberian airlines.

Notice the white tablecloth and cute salt/pepper shakers. Maybe this is standard for business class, but then business class isn’t standard for me. The flavours were exquisite; fresh dill on the scrambled eggs; a delicate caprese-esque salad; a curious but good chicken sausage. There was a second, dessert course as well. This was breakfast. Thank goodness it was at my jet-lag lunchtime; I don’t know that I could have downed it for breakfast. I hadn’t expected to blog about breakfasts on airplanes on my Siberian expedition, but then, experiences happen. Fieldwork is as much about the people, places, and surprises as it is about the research.

s7 Siberian airlines has been a wonder, by the way, from its hip green and purple colours to its website with the best usability I’ve encountered. It feels like an airline designed by Apple (except for the green and purple).

Shostakovich’s 5th is one of the few pieces of 20th century classical music I adore. Yes, it has its moments of over-the-top pomp, but there are some sublime transitions that just twist my heart. I’m listening to it for its Russianness. Getting in the mood, although Shostakovich’s mood might be a few decades out of date.