Habitat fragmentation and its genetic consequences are a critically important issue in evaluating the evolutionary penalties of human habitat modification. Here, we examine the genetic structure and diversity in naturally subdivided and artificially fragmented populations of the endangered tidewater goby (Eucyclogobius newberryi), a small fish restricted to discrete coastal lagoons and estuaries in California, USA. We use five naturally fragmented coastal populations from a 300- km spatial scale as a standard to assess migration and drift relative to eight artificially fragmented bay populations from a 30- km spatial scale. Using nine microsatellite loci in 621 individuals, and a 522-base fragment of mitochondrial DNA control region from 103 individuals, we found striking differences in the relative influences of migration and drift on genetic variation at these two scales. Overall, the artificially fragmented populations exhibited a consistent pattern of higher genetic differentiation and significantly lower genetic diversity relative to the naturally fragmented populations. Thus, even in a species characterized by habitat isolation and subdivision, further artificial fragmentation appears to result in substantial population genetic consequences and may not be sustainable.

A previously unrecognized microsporidian (Kabatana newberryi n. sp.) is described from the musculature of Eucyclogobius newberryi (Gobiidae) in Big Lagoon, Humboldt County, California. Spores are ovoid, ranging in size from 2.8 +/- 0.3 microm in total length and 1.9 +/- 0.4 microm in width (measurements of 30 spores made by calculation from micrograph). The polar filament has 9-10 coils in 1-2 rows. Development occurs in direct contact with host muscle cell cytoplasm, without xenoma or sporophorous vesicle. Phylogenetic analysis of the new species and of 35 other microsporidians known to infect fish using 1115 base pairs of aligned 16S rRNA gene indicate the new species is most closely related to Kabatana takedai. However, the new species differs by 11% sequence divergence from K. takedai. Divergence in morphology and genetic data allow for diagnosis from all other fish-infecting microsporidia and supports recognition of a new species of microsporidian, Kabatana newberryi n. sp., presently known only from a suspected specific host, the endangered tidewater goby Eucyclogobius newberryi.

This study documents predation by the endangered tidewater goby, Eucyclogobius newberryi, upon the invasive New Zealand mudsnail, Potamopyrgus antipodarum, in Big Lagoon, California, USA. To estimate the prevalence of NZ mudsnails in the diet of tidewater goby, the gastric contents of 411 individuals, collected monthly from April 2009 to August 2010, were examined. NZ mudsnails were found in the digestive tract of tidewater goby that ranged in size from 14 to 52 mm total length, corresponding to post-settlement and nearly maximal sizes of this species. Unlike other native species which are unable to extract nutrition from these snails, tidewater goby fully digest this hard-shelled prey, as evidenced by the presence of shell fragments and complete absence of intact shells in the hind gut. The number of ingested NZ mudsnail ranged from 1 to 27 (mean 4.4), and ranged in length from 0.39 to 4.0 mm. The average size of ingested snails increased with fish length (r 2 = 0.42, P \ 0.001). NZ mudsnails were found in over 80% of individuals during the summer and fall of 2009, when the estimated population size of tidewater goby in Big Lagoon was greater than three million. This study documents the first instance of a native and endangered species that preys upon and utilizes the NZ mudsnail as a food source, and suggests that tidewater goby can exert substantial predation pressure upon NZ mudsnails and take advantage of these readily available novel prey items.

Extinction and colonization dynamics are critical to understanding the evolution and conservation of metapopulations. However, traditional field studies of extinction-colonization are potentially fraught with detection bias and have rarely been validated. Here, we provide a comparison of molecular and field-based approaches for assessment of the extinction-colonization dynamics of tidewater goby (Eucyclogobius newberryi) in northern California. Our analysis of temporal genetic variation across 14 northern California tidewater goby populations failed to recover genetic change expected with extinction-colonization cycles. Similarly, analysis of site occupancy data from field studies (94 sites) indicated that extinction and colonization are very infrequent for our study populations. Comparison of the approaches indicated field data were subject to imperfect detection, and falsely implied extinction-colonization cycles in several instances. For northern California populations of tidewater goby, we interpret the strong genetic differentiation between populations and high degree of within-site temporal stability as consistent with a model of drift in the absence of migration, at least over the past 20-30 years. Our findings show that tidewater goby exhibit different population structures across their geographic range (extinction-colonization dynamics in the south vs. drift in isolation in the north). For northern populations, natural dispersal is too infrequent to be considered a viable approach for recolonizing extirpated populations, suggesting that species recovery will likely depend on artificial translocation in this region. More broadly, this work illustrates that temporal genetic analysis can be used in combination with field data to strengthen inference of extinction-colonization dynamics or as a stand-alone tool when field data are lacking.