130 
THE CORAL TRIANGLE: HEARST BIODIVERSITY EXPEDITION 
With respect to its taxonomy and identification, Gymnodoris is a particularly challenging 
genus of nudibranchs. Plagued by inadequate descriptions and little variation of external color pat¬ 
terns, the true diversity of this group has remained poorly documented in the literature. 
We have found that molecular techniques have been useflil in identifying putative undescribed 
species and have further highlighted the difficultly of reconciling old descriptions with the true 
diversity of this group of voracious predators. While there are approximately 30 described species 
of Gymnodoris, several of these have descriptions that are inadequate for matching species to them, 
and most of these species names likely lack associated type material. Few Gymnodoris species 
have been described in recent years. Most recently, Hamatani and Osumi (2003) described 
Gymnodoris inariensis, a small, blackish, gill-less species from Japan, and Swennen (1996) 
described Gymnodoris pattani, a translucent greenish species from Thailand. Here we describe 
three new taxa and compare them to known species of Gymnodoris. Because two of these new taxa 
were difficult to tell apart in the field, in addition to looking at morphological details, we sequenced 
specimens for a fragment of the cytochrome oxidase subunit I (COI) gene and calculated uncor¬ 
rected p-distances to see if this was consistent with the differences seen in morphology. 
The material was deposited in the Department of Invertebrate Zoology and Geology of the 
California Academy of Sciences, San Francisco (CASIZ), and sequences for the COI gene were 
uploaded to GenBank. 
Methods 
Morphological Examination. — We subsampled many of the specimens for DNA prior to 
morphological analysis or measurement, resulting in lengths that are smaller than the intact pre¬ 
served specimens. 
We dissected specunens by making an incision with forceps near the foot. Reproductive stmc- 
tures were drawn initially using a drawing tube attached to a dissecting microscope and subse¬ 
quently scanned, traced and rendered digitally using Adobe Illustrator CS4 or CS6 and Photoshop 
CS4 or CS6. 
To examine the radulae, we placed the buccal masses of specimens in a 10% NaOH solution 
until the tissue surrounding the radulae had dissolved sufficiently. We subsequently placed these 
radulae in deionized water to remove residual salt and mounted them on SEM stubs using double 
sided adhesive or no adhesive at all. The radulae were imaged using a scanning electron micro¬ 
scope (LEO 1450 VP). To assess whether the penis of specimens was armed with spines, the dis¬ 
tal portion of the vas deferens was excised and cleared and stained by first dehydrating in a series 
of increasing concentrations of ethanol, then placed in acid fuchsin stain, then placed in xylene, 
removed and mounted on a glass microscope slide. The slides were photographed using a Spot 
FLEX Mosaic Model 15.2 digital camera mounted on a Leica DMRB microscope. Scale bars were 
applied in SPOT Software 5.0.25. Adobe Photoshop CS4 was used to get a rough approximation 
of the length and width of penial spines, by measuring and averaging the lengths of four different 
visible spines andwidths of four visible spine bases using the scale bar for measurement calibra¬ 
tion. 
Molecular Methods. — In this paper we were interested in comparing COI sequences for 
two taxa that had a similar external appearance that were difficult to tell apart in the field. We sub¬ 
sampled several specimens for DNA analysis by taking a small piece of tissue from the foot or tip 
of the “tail.” We were careful to avoid subsampling digestive tissue since some gymnodorids are 
known to eat other nudibranchs including congeners and conspecifics, in some cases. 
Polymerase chain reaction (PCR) was used to amplify a partial COI gene fragment using the 
standard “Folmer” COI primers (Folmer et al. 1994). Each 25 pL reaction contained: 15.0 pL ster- 
