Millipede Lab

In Tsukuba, which is about 50 km northeast of Tokyo, we met Tsutomu and his student Ryosuke Kuwahara. I really enjoyed meeting Ryosuke because he’s very enthusiastic about natural history and takes wonderful photographs of xystodesmid millipedes (here’s a link to his website). That evening, we visited Seiryuji Temple in Tsuchiuria and collected millipedes in a nearby forest. When we walked into the forest, a Luna moth flew in front of us and landed on a tree.

The xystodesmid millipede Parafontaria tonominea

Often, finding millipedes is sporadic, and we spent about 20 minutes before we encountered the first xystodesmid, Parafontaria tonominea (Attems, 1899). That evening, we also found Parafontaria ishiii Shinohara, 1986 and Xystodesmus martensii (Peters, 1864). (Notably in 1895, Cook established the genus Xystodesmus and the family Xystodesmidae based on X. martensii, a species originally described by Peters in 1864 as Polydesmus martensii). With the moth and millipedes, there were other fascinating organisms in the forest including a gray hand-sized centipede, Scolopendra japonica C.L. Koch, 1878.

The xystodesmid millipede Parafontaria ishiii

It became clear that night collecting was the most productive. For example in Sapporo, we found two millipedes in six hours during the day and 50 millipedes in 15 minutes at night. Consequently, we spent our days planning fieldwork, resting for night-collecting, looking at Tsutomu’s natural history collections, and tending to the live millipedes (and of course eating a good amount of ramen).

Xystodesmus martensii

The second night in Tsukuba, we travelled through a network of rice plantations to Mount Hōkyō-san and collected in the cherry, chinquapin, and bamboo forest that covers the mountain. We hiked a couple miles to a site about 400 meters in elevation that was just below the peak of Hōkyō-san. At 8PM, P. ishiii, X. martensii (and another millipede that Ryosuke and Tsutomu will soon share), started to emerge from their subterranean hiding places. An hour later, we encountered more. At the end of the night, we found about one hundred xystodesmid millipedes. We collected about 20 individuals to bring to Virginia Tech for DNA sequencing and molecular phylogenetics. At the moment, we have sequenced about 239 of the family’s 393 species (and nearly all of the taxon’s 62 genera). We’ve done this to estimate the family’s evolutionary history in a phylogeny. The phylogeny of Xystodesmidae is important as a guide for describing species and tracing the evolution of mimicry and features like chemical defenses.

Mount Hōkyō-san, Ibaraki Prefecture

Left to right: Tsutomu Tanabe, Paul Marek, Charity Hall, and Ryosuke Kuwahara

Walking down Mount Hōkyō-san, we enjoyed being entomologists and encountered large Carabus beetles, sawflies, and several huge S. japonica centipedes hanging out on the sides of trees. Ryosuke pointed out a handsome mamushi on the trail near the rice field, Gloydius blomhoffii (H. Boie, 1826), which is species of viper.

The xystodesmid Parafontaria erythrosoma

The third night in Tsukuba, we collected at a lower elevation (89 m) forest near Kashima Shrine in Mito. Again, as a testament to the patchiness of millipede encounters, we found little in the first 30 minutes. But once we encountered a suitable patch, we found lots of millipedes. The first xystodesmid millipede that we found was Parafontaria erythrosoma (Takakuwa, 1942). A handsome millipede, which as its name implies, has a red (“erythro”) body (“soma”). With P. erythrosoma, we found X. martensii again, which is a very widespread species with a considerable amount of variation (Tanabe and Shinohara, 1996). Ryosuke broke open a decaying log and found a species of Hyleoglomeris, which is a glomerid millipede capable of rolling into a ball. We encountered other millipedes including the haplodesmid Eutrichodesmus peculiaris (Murakami, 1966); the polydesmid Epanerchodus lacteus Shinohara, 1958; and a species of the paradoxosomatid Cawjeekelia Golovatch, 1980. Notably, E. peculiaris has maternal care of eggs, which is a rare behavior in millipedes, and an incredibly rare behavior in millipedes of the order Polydesmida (Murakami, 1972).

The haplodesmid millipede Eutrichodesmus peculiaris

We hopped back on the train to the Narita Airport and flew to Kumamoto Prefecture where Tsutomu teaches at Kumamoto University. At the airport, we met Tsutomu’s student Taiki Kato who is studying pill millipedes, including Hyleoglomeris. Taiki plans to reconstruct the evolutionary history of the taxon in Japan (circa 10 species). After eating octopus fritters and these chips, we drove to a Cryptomeria japonica (Japanese cedar) forest northeast of Kumamoto in Haru. The site was next to a pull-off on the road and reminded me of the spots where we collect in Appalachia. We arrived at dusk and once we entered the forest and clicked on our ultraviolet flashlights, we found this absolute behemoth xystodesmid, Riukiaria cornuta (Haga, 1968).

This millipede is the largest (in length) xystodesmid I’ve ever encountered, and I practically lost it with amazement when I first saw it. The Appalachian species Pachydesmus crassicutis (Wood, 1864) is slightly wider but shorter in length than R. cornuta. (The species Rhysodesmus dasypus (Gervais, 1847) is the largest known xystodesmid species at 25 mm wide and 80 mm long). There were a few R. cornuta starting to poke their heads above ground at 8PM; however, at 8:45PM they were everywhere and encountered every few steps. The forest was very dark, so they positively lit up with a bright green glow when illuminated under ultraviolet light.

The giant xystodesmid millipede Riukiaria cornuta

Riukiaria cornuta illuminated with ultraviolet light

I spent the next couple days with Tsutomu at Kumamoto University. I worked in the millipede collection examining specimens and discussing a few collaborative projects with Tsutomu and Taiki. Tsutomu was incredibly wonderful and shared his ethanol-preserved specimens with me, and Taiki helped me find things in the lab, look up localities in Tsutomu’s database, and prepare specimens. Kumamoto University has a fantastic tree-lined campus with a nice guest house and really friendly students.

The last day of collecting, we hopped on a ferry and crossed the Ariake Sea to Mount Unzen in Nagasaki Prefecture. Mount Unzen, known for its violent eruptions, had its last activity between 1991 to 1994 when numerous pyroclastic flows buried houses and spewed toxic gas. When driving on Unzen, I could see steam and smell sulfur in certain areas. Some well-known hot springs, including the aptly named Unzen Hell, sit atop these vents.

Japanese feather millipedes, Brachycybe nodulosa (Verhoeff, 1935)

We collected millipedes at a couple of sites on Mount Unzen. We were looking for the andrognathid millipede Brachycybe nodulosa (Verhoeff, 1935) at the first site, and the xystodesmid Riukiaria marinae (Golovatch, 1978) at the second site. Notably, B. nodulosa is the closest relative of the Appalachian species, Brachycybe lecontii Wood, 1864 (Brewer et al., 2012). It’s even more closely related to B. lecontii than is Brachycybe petasata Loomis, 1936, which is sympatric with B. lecontii in the Southern Appalachian Mountains.

The city of Shimabura with Mount Unzen way in the background

We searched in several Cryptomeria forests without finding B. nodulosa, then eventually found them in the fourth spot. Their rarity is similar to B. lecontii, and often ideal-looking habitats are searched without finding a single individual. Tsutomu is familiar with collecting B. nodulosa and knew to spend less than about 15 minutes at a site then move onto the next one. Tsutomu found the first B. nodulosa on a Cryptomeria log. The microhabitat of B. nodulosa is similar to B. lecontii in that they occur on dead logs that are still intact and firm but with loose bark. In contrast to B. nodulosa, the species B. lecontii seems to like Tulip poplar. I was struck by the orange hue of B. nodulosa that differs from the pink color of B. lecontii.

The cryptodesmid millipede Kiusiunum

Co-occuring with B. nodulosa on the same cedar log, we found the millipedes Kiusiunum Verhoeff, 1942 (Cryptodesmidae) and Kiusiozonium Verhoeff, 1941 (Hirudisomatidae). These three millipedes are dorsoventrally flattened and chemically defended with either alkaloids or alcohols. The odor of Kiusiozonium is similar to the Appalachian millipede Petaserpes that has a camphor (Vicks VapoRub-like) odor.

The hirudisomatid millipede Kiusiozonium

Cryptomeria japonica forest from afar (C. japonica trees in foreground)

Cryptomeria japonica forest closeup

With a couple hours before the ferry departed, we visited the type locality of R. marinae on the south slope of Mount Unzen. The site consisted of a nice dark cedar forest with a thick layer of spongy decaying cedar duff. In a small gully, I found individuals of B. nodulosa, Kiusiunum, and a Hyleoglomeris with a white stripe near its head. In our 30 minutes of collecting, we didn’t find R. marinae. That’s not surprising and if we would have found an individual it would have been lucky given that Japanese xystodesmids are decidedly nocturnal and patchy in distribution.

The pill millipede Hyleoglomeris