Final Report Umpqua Chub Population Estimate USFWS Grant # E-2-54

By Douglas F. Markle1, Kathleen O’Malley1, Mark Terwilliger1, Pete Baki2, Holly Truemper2, and Dave Simon1

1Department of Fisheries & Wildlife Oregon State University Corvallis, OR 97330

2Oregon Department of Fish & Wildlife 4192 N. Umpqua Hwy. Roseburg, OR 97470

4 January 2011

Summary

Over the past several decades the distribution of Umpqua chub has contracted such that

there are now six fragmented populations. Because there is little information on the biology of

Umpqua chub and concern has been raised about their ability to persist in the presence of

smallmouth bass, we began a collaborative effort to better understand their demographics and

genetics. In 2008, we collected 25 specimens from each of the six populations and tested a mark-

recapture protocol. We also collected an additional 25 specimens from each population in 2010

for potential future analysis.

The mark-recapture protocol was designed to have a minimum detection distance of 0 m

and a maximum detection distance of 700 m. Unfortunately, there were only 4 recaptures, but none of the recaptures had moved over 100 m suggesting that over a 24 hr period in summer,

Umpqua chub may not move very great distances. We did not try to estimate local abundance

with these data. Because they live in an open stream network, future efforts to quantitatively

monitor populations may require additional work to clarify movement patterns and abilities.

Umpqua chub are short-lived with a maximum age in our samples of 7 years. Growth is

relatively rapid the first two years of life, then slows, presumably associated with maturation.

Because of the reduced growth rate, there was considerable overlap in lengths of older fish. A 50

mm fish could be 2-7 yr old. Depending on mortality rates, average Umpqua chub generation

time could be expected to be about 4 yr. Our samples from the six populations differed in size

frequencies so we could not detect if there were differences in growth rates. For example, Smith

River fish were the smallest and youngest, but we do not know if those features are characteristic

of the population or an artifact of sampling.

Our genetic analyses suggest that the six nominal Umpqua chub populations are, in fact,

genetically differentiated from one another. Differences between populations are partly

attributable to distance, such that populations that are farther apart tend to have greater

differences than those closer in space. However, over distances ranging from 105-280 km, Smith

River fish showed the greatest differences from other populations (mean Fst =0.100, range 0.070

– 0.121). Excluding the very low Calapooya Creek and Ollala Creek pairwise comparison,

differences among Elk Cr, Cow Cr, and South Umpqua for distances ranging from 57-241 km

were less than half the values from Smith River (mean Fst =0.045, range 0.028 – 0.069). These

data may suggest that there has been more than one event fragmenting Umpqua chub populations with isolation of Smith River prior to isolation of upstream populations.

Future work should include system-wide monitoring, perhaps every five years. At a minimum, the existence and spatial extent of each of the six populations should be documented.

Ideally, unbiased estimates of abundance and age structure should be collected, but our experience to date suggests that both may be difficult to obtain. The Smith River population deserves additional scrutiny as it is the most differentiated and isolated group and threats to their existence appear to differ as well.

Introduction

Umpqua chub Oregonichthys kalawatseti is a small (max. length about 65mm) cyprinid endemic to the rivers and streams of the Umpqua Basin, Oregon (Markle et al. 1991). It lives in moderate- to no-flow habitats typically associated with vegetation (Pearsons1989) and was formerly a U. S. Fish and Wildlife Service “Category 2” candidate species requiring further information. Due to its restricted distribution and threats from non-native fishes such as smallmouth bass Micropterus dolomeiu (Simon and Markle 1999, Simon 1998, 2008), Umpqua

3 chub is currently a State of Oregon “sensitive – vulnerable” species (ODFW 1995). Smallmouth bass is a piscivorous centrarchid native to eastern North America (Scott and Crossman 1973) and are known to reduce abundance, alter habitat use, and extirpate small fishes (MacRae and

Jackson 2001). Smallmouth bass were introduced into Oregon in 1924 or 1925 (Lampman 1946), and into the South Umpqua River from either adjacent farm ponds after 1964 floods (Simon and

Markle 1999) or from illegal stockings in the 1970’s (Daily 1990). Release after the 1964 flood seems questionable based on our review of ODFW Roseburg annual reports from 1964 – 1976.

Those reports contain no mention of escape of smallmouth bass. Further, creel surveys and a poison treatment of 74 miles of Cow Creek and 45 miles of its tributaries in 1970 found no smallmouth bass. Smallmouth bass were first confirmed in the basin (South Umpqua River) in the 1976 Oregon Game Commission Report.

Surveys in 1987 and 1998 suggested Umpqua chub distribution had become increasingly restricted. Markle et al. (1991) surveyed 38 sites in several water bodies throughout the Umpqua basin in 1987 and found chubs at 12 sites. A subsequent resurvey of the same sites in 1998

(Simon 1998, Simon and Markle 2000) found chubs at only 6 sites, while smallmouth bass increased from 7 sites in 1987 to 19 sites in 1998. Markle et al. (1991), Simon (1998, 2008) and

Simon and Markle (2000) expressed concerns that bass were negatively affecting Umpqua chub by restricting distribution, fragmenting populations, and causing unknown genetic and/or demographic consequences. In 2006 and 2007, Simon (2008) greatly expanded on the 1987 and

1998 surveys to determine relative abundance and distribution patterns of Umpqua chub and smallmouth bass throughout the Umpqua Basin.

Based on these findings, an advisory committee met in Roseburg on 6/26/08 with the goal of gathering more information and, if possible, to keep the species from being listed. The

4 outcome of the meeting was a partnership between Oregon Department of Fish and Wildlife and

Oregon State University to further knowledge of the species. Original goals were to conduct distribution and abundance surveys and collect specimens for genetic and age analysis from the 6 nominal subpopulations identified by Simon (2008). Despite limited funding, we were able to confirm known distribution, qualitatively assess abundance, complete genetic and age analyses on samples collected in 2008, and collect samples in 2010.

Methods

Field collections—Umpqua chub were collected by various methods depending on what was efficient for each site including 3/8” mesh seines, cast nets, and minnow traps. Cast nets used were EZ Throw 750’s with a six foot radius, 3/8” square mesh, and 20’ floating hand line.

Steel minnow traps (Nylon Net Company, MT2 G-40 Minnow and Shrimp Traps) were 16”x 9” with a 7/8” fyke opening and ¼” mesh. Nets were used actively to catch Umpqua chub and release by-catch. Minnow traps were typically set overnight, baited with white bread, and checked within 24 hours. When chubs were visually abundant or water temperatures were high, minnow traps were checked more frequently. In 2008, set and pull times for traps were not recorded, so effort was assumed to be 24 hrs for each set. In 2010, set and pull times for traps were recorded. All fish trapped were recorded to species and non-chub were released.

Minnow traps could be routinely used to collect subadults and adults for genetics and ageing at five of six previously documented population isolates (Simon 2008) between 5 -9 September 2008:

Elk Creek, Calapooya Creek,

Olalla Creek, Cow Creek, and

South Umpqua River (Figure

1A). Samples for Smith River were collected 2 October 2008 with cast nets (Figure 1A).

Twenty five individuals from each nominal population were preserved in 95% ethanol; otoliths and fin clips were removed later in the lab. Minimum distance (m) between nominal populations was estimated based on the 2006-2008 samples. Additional samples were collected in 2010, but, due to lack of funds, aging and genetic analyses were not completed. These samples will be stored at ODFW Roseburg for potential future analysis.

As a prelude to future quantitative abundance estimates, we tested a mark/recapture protocol in late summer of 2008. The mark/recapture field work occurred between September 3-

5, 2008, on a section of Elk Creek just upstream of the confluence with Hardscrabble Creek

(T22S-R6W-S11, 12, 13 and 14). Minnow traps were baited with white bread following the protocol used for Oregon chub (Scheerer 2007) and set on September 3. Fish were marked the

next day, and recaptured on September 5. Ten minnow traps were placed 10 m apart in each of

five 100 m sample units in Elk Creek with each sample unit separated by a 100 m unsampled

buffer. From downstream to upstream, the fin-clip marking protocol for the five units was: Unit

1: no mark, Unit 2: lower caudal, Unit 3: lower and upper caudal, Unit 4: upper caudal, Unit 5: no mark. Thus, a marked fish could have a detection distance of 0-100 m if captured in the same unit in which it was marked, or a minimum detection distance of 100 m (eg. marked in the most upstream trap in Unit 2 and recaptured in the most downstream trap in Unit 3), or a maximum detection distance of 700 m (eg. marked in the most downstream trap in Unit 2 and recaptured in the most upstream trap in Unit 5). Within units 3 and 4, Umpqua chub from one trap in each unit had left pectoral fins clipped to more accurately determine movement less than 100 m within a unit. All vertebrates and macroinvertebrates were recorded and the following additional information recorded at each trap site: unique trap number, UTM coordinates, substrate percentages (soil/organic, sand, gravel, cobble, boulder, and bedrock), vegetation presence, type

and percent cover, mesohabitat (Moore et al. 2010), water velocity, shade, dominant/subdominant riparian vegetation (tree, shrub, grass/forb or bare), wetted width of

channel, trap depth (recorded from water surface to stream substrate and bankslope). On the last

day of sampling twenty-five Umpqua chub were retained for further aging and genetic sampling.

Otolith ageing— We examined 150 juvenile Umpqua chub from the six nominal

populations following the methods in Terwilliger et al. (2010). Briefly, the right lapillus was

removed from all fish using a dissecting microscope and forceps and stored dry for later

processing. Each lapillus was embedded in Buehler Epothin epoxy resin, and a 1-mm oblique

section running posterodistal- anteromedial that included the core was made using a Buehler

isomet low-speed saw equipped with a diamond blade. This section approximates one made in

the sagittal plane, but the oblique angle of the lapillus in the chub’s head made it necessary to

perform the oblique section. Sections were mounted on glass slides with Crystal bond adhesive, sanded with 1200-grit wet/dry sandpaper to remove saw marks and gain proximity to the core, and polished with a synthetic velvet cloth and 0.05 µm alumina powder. The otolith section was flipped during grinding and polishing to create a thin section showing visible increments along the entire diameter of the otolith (see Secor et al. 1992). Otolith sections were examined using a compound microscope with transmitted light under 10X magnification. Ages were assigned from counts of growth increments that were comprised of a wide translucent and narrow opaque band, and all fish were assigned a nominal birthdate of 1st January. Within-reader precision was estimated in terms of absolute percent error as outlined in Beamish and Fournier (1981). Blind counts of growth marks were made three times over the course of several weeks; the median age obtained from these reads was determined to be the final age and was used in any subsequent analyses. Average generation time was estimated from the age range of adults and extrapolation of age at maturity from Oregon chub, O. crameri.

Microsatellite analyses—Total genomic DNA was extracted from fin clip tissue following the Glass Fiber Plate DNA Extraction Protocol (Ivanova et al. 2006). Polymerase chain reactions (PCR) were carried out in 5ul volumes to amplify eleven microsatellite loci utilizing fluorescently labeled primers: Ocr100, Ocr101, Ocr103, Ocr104, Ocr106, Ocr110,

Ocr111, Ocr112, Ocr113, Ocr114, and Ocr115 (Ardren et al. 2007). Reaction conditions were as follows: initial denaturation at 94°C for 3 minutes, followed by 26 cycles of 94°C for 30 seconds, 30 seconds annealing at 58°C, 30 seconds at 72°C, and a final extension at 72°C for

seven minutes. PCR products were separated via polyacrylamide gel electrophoresis on an

ABI3730XL DNA Analyzer and binned according to size with GENEMAPPER®.

We tested populations for conformance to Hardy-Weinberg expectations (HWE) and linkage disequilibrium using the program GENEPOP v 4.0.01 (Raymond and Rousset 1995).

We adjusted the initial critical value of 0.05 using sequential Bonferroni corrections (Rice 1989) to account for multiple comparisons made during these tests. To estimate measures of genetic diversity, including mean number of alleles per locus, and observed and expected heterozygosity, we employed the program GENETIX (Belkhir et al. 2004). To determine the level of genetic variation among populations, we executed exact tests for differences in genic and genotypic frequencies using GENEPOP. In addition, we calculated pairwise Fst values (Weir and

Cockerham 1984) to estimate the level of genetic variation among each population pair and we

used a permutation test with 1000 iterations to assess the statistical significance of these

estimates using GENETIX.

To examine the spatial genetic relationship among the six populations, we constructed a

phylogenetic tree using the analysis package PHYLIP v3.69 (Felsenstein 1993). We estimated

Cavalli-Sforza and Edwards (1967) chord distances between all population pairs in each dataset

using the program GENDIST and generated a neighbor joining (NJ) using the NEIGHBOR

program. To bootstrap the data and estimate statistical support for the topology of this consensus

NJ tree, we used the program SEQBOOT. We displayed the trees with TREEVIEW (Page

1996).

Results

Field collections

Thirteen fish taxa, three herps and one macroinvertebrate were collected (Table 1).

Redside shiners were the most abundant taxon and smallmouth bass were never collected. Our

best evidence of relative abundance is the 2010 hourly catch per unit effort (CPUE) for the five

9 sites that could be sampled with minnow traps, but the site-specific nature of the sampling does not justify extrapolation to larger areas.

In the mark-recapture study we marked 106 fish but only recaptured four. Curiously 3 of

26 (11.5%) pectoral fin-clipped fish were recaptured while only 1 of 80 (1.2%) caudal fin- clipped fish was recaptured. Two pectoral fin-clipped fish were recaptured in the same trap, one pectoral fin-clipped was caught within 10 m, and one caudal fin-clipped fish was caught within the original trap unit (0-100 m). Traps that caught the most Umpqua chub were typically heavily

Table 1. Catch summary for 2008 and 2010 sampling. Smith River samples do not include trap effort because multiple gears (traps, cast nets and seines) had to be used. Umpqua chub catch per unit effort (CPUE) calculated per trap for both years and per hour fished for 2010.

Population and Year Calapooya South Olalla Elk Creek Cow Creek Smith River Creek Umpqua Creek 2008 2010 2008 2010 2008 2010 2008 2010 2008 2010 2008 2010 Number of 50 30 12 49 15 20 22 10 n/a n/a 12 16 traps Umpqua chub 98 57 25 25 77 31 51 25 25 47 64 106 CPUE trap 1.96 1.90 2.08 0.51 5.13 1.55 2.32 2.50 - - 5.33 6.62 CPUE hour 2.59 0.45 17.71 1.16 - - 4.61

Other fish Speckled dace 51 7 15 84 3 0 76 6 0 0 1 2 Umpqua dace 7 0 0 7 0 0 0 0 0 0 0 0 Redside shiner 697 105 169 223 36 47 359 92 59 449 61 28 Umpqua 18 20 0 3 0 0 0 5 3 49 1 2 pikeminnow Tyee sucker 2 1 18 1 0 0 9 0 0 0 0 0 Brown 0 0 0 0 0 0 0 0 2 0 0 0 bullhead Juvenile 0 0 0 1 2 0 3 1 0 0 0 0 steelhead Juvenile coho 0 0 0 4 0 0 1 1 0 0 0 4 Sculpin, 14 0 1 13 5 0 3 2 9 5 6 3 unknown Three spine 61 17 0 0 0 0 0 0 3 0 4 15 stickleback Pumpkinseed 0 4 0 1 0 0 0 0 0 0 0 0 Warmouth 0 0 0 0 0 0 0 0 1 0 0 0

Others Tadpole, 0 0 0 0 185 4 0 0 0 0 0 0

10 unknown Rough 2 0 0 0 0 0 3 2 0 0 0 3 Skinned newt Gopher snake 0 0 0 0 0 0 1 0 0 0 0 0 Crayfish 14 5 26 63 0 0 9 25 0 0 5 20

Table 2. Umpqua Chub Mark-Recapture Trap Location and Habitat Description for Elk Creek near Hardscrabble Creek. Data was recorded 9/3-9/4/08 during clear water clarity starting at 10:50 am at water temperatue 63F.

UTM Location Substrate % Vegetation Shade (%) Trap Water Riparian Wetted Present % Mesohabitat Depth (m) Bankslope (%) # Zone Easting Northing S/O Sand Gravel Cobble Boulder Bedrock Veg Type Velocity Left Right Vegetation* Width (Y/N) Cover

1 10T 470266 4834723 15 30 30 20 5 0 N - 0 LP L 75 60 1/3 18.4 0.53 45/5 2 10T 470273 4834726 5 15 20 30 30 0 N - 0 LP L 90 50 1/3 19 0.57 40/8 3 10T 470282 4834721 0 10 45 40 5 0 N - 0 LP L 75 70 2/1 12.7 0.62 40/5 4 10T 470289 4834717 5 5 0 25 15 50 N - 0 LP L 90 75 1/2 9.4 0.43 45/5 5 10T 470295 4834710 5 0 0 0 5 90 N - 0 LP L 80 50 1/3 19.3 0.57 45/5 6 10T 470300 4834700 5 5 15 0 0 75 Y ALGAE 40 LP L 70 50 1/2 26 0.69 45/15 7 10T 470299 4834683 0 5 0 15 0 80 Y ALGAE 50 LP L 75 65 1/2 28.9 0.73 45/10 8 10T 470314 4834676 10 20 0 0 0 70 N - 0 LP L 90 80 1/2 28.3 0.37 40/15 9 10T 470315 4834660 5 10 10 25 25 25 Y ALGAE 15 LP L 85 85 1/2 28.7 0.6 40/25 10 10T 470326 4834663 5 5 15 30 20 25 Y GRASS 30 LP L 85 85 1/2 28 0.49 40/30 11 10T 470425 4834612 40 60 0 0 0 0 N - 0 LP L 85 85 1/3 20.7 0.7 80/35 12 10T 470431 4834606 60 35 0 0 5 0 N - 0 LP L 85 85 1/2 19.3 0.63 80/45 13 10T 470425 4834597 80 20 0 0 0 0 N - 0 LP L 85 85 1/2 16.6 0.7 80/45 14 10T 470442 4834597 20 20 0 0 0 60 N - 0 LP L 85 85 1/3 16.8 0.8 80/50 15 10T 470456 4834595 0 0 0 0 0 100 N - 0 LP L 85 85 1/3 14.8 0.8 85/50 16 10T 470461 4834587 5 5 0 0 0 90 N - 0 LP L 80 85 1/3/4 14.7 0.63 85/40 17 10T 470470 4834576 0 0 5 10 10 75 Y GRASS 15 LP L 80 70 1/3/4 14.5 0.33 85/10 18 10T 470488 4834552 0 5 25 5 55 10 N - 0 LP L 85 60 1/3 8.9 0.58 85/10 19 10T 470492 4834550 0 5 20 30 40 5 N - 0 SB M 85 60 1/3 13.2 0.74 85/10 20 10T 470507 4834540 0 0 0 0 0 100 N - 0 LP L 85 65 1/3 10.5 0.65 85/25 21 10T 470561 4834489 0 5 0 5 90 0 N - 0 LP L 80 85 1/3 13.2 0.52 35/45 22 10T 470562 4834481 5 20 70 5 0 0 N - 0 LP L 80 85 1/2 16.4 0.37 25/45 23 10T 470567 4834465 10 5 30 30 25 0 N - 0 LP L 80 80 1/3 14.7 0.56 15/80 24 10T 470567 4834432 5 35 20 15 25 0 N - 0 LP L 80 85 1/2 16.6 0.5 10/70 25 10T 470574 4834430 0 10 10 30 50 0 N - 0 LP L 80 85 1/3 14.2 0.43 15/75 26 10T 470584 4834427 5 10 30 30 25 0 N - 0 LP L 60 65 1/2 11.4 0.42 15/ 30/ 27 10T 470594 4834430 0 10 30 50 10 0 N - 0 LP L 55 70 1/3 11.6 0.28 40 25/ 28 10T 470601 4834921 30 20 20 10 20 0 N - 0 LP L 55 70 1/3 15.9 0.58 45 25/ 29 10T 470608 4834413 0 0 25 60 15 0 N - 0 RI M 65 75 1/3 11.3 0.31 15

25/ 30 10T 470619 4834410 0 5 25 50 20 0 N - 0 RI M 75 80 1/3 9.7 0.31 15 10/ 31 10T 470723 4834365 25 45 25 5 0 0 Y GRASS 60 LP L 60 65 1/3 20.4 0.98 25 15/ 32 10T 470739 4834375 0 0 10 10 80 0 N - 0 LP L 55 60 1/3 25.2 1.25 40 15/ 33 10T 470751 4834382 5 70 5 5 5 10 N - 0 LP L 55 75 1/3 19.6 1.4 50 10/ 34 10T 470759 4834393 5 20 0 0 0 75 N - 0 LP L 55 65 1/3 19.2 0.57 25 5/2 35 10T 470772 4834400 15 10 0 10 40 15 N - 0 LP L 45 65 1/3 13.6 0.41 0 5/2 36 10T 470761 4834404 10 10 0 0 0 80 Y GRASS 40 LP L 25 65 1/3 28.1 0.25 0 5/2 37 10T 470803 4834362 0 0 15 15 0 70 N - 0 LP L 40 50 1/3 28 0.26 0 10/ 38 10T 470770 4834366 0 0 5 15 5 75 N - 0 LP L 50 60 1/3 27.1 0.45 20 25/ 39 10T 470774 4834403 0 0 10 10 5 75 N - 0 LP L 45 65 1/3 23.3 0.81 25 20/ 40 10T 470770 4834424 0 30 0 10 60 0 N - 0 LP L 45 60 1/3 22.8 0.51 25 25/ 41 10T 470865 4834531 30 30 0 0 0 40 Y ALGAE 70 LP L 45 65 1/3 25.2 0.49 30 25/ 42 10T 470864 4834540 30 10 0 0 0 60 Y H20 50 LP L 35 50 1/3 22.4 0.51 20 70/ 43 10T 470878 4834548 30 20 0 0 0 50 Y H20 40 LP L 35 40 1/2 21.8 0.37 20 20/ 44 10T 470884 4834559 10 0 0 0 5 85 Y ALGAE 90 LP L 35 40 1/3 18.6 0.4 25 20/ 45 10T 470889 4834572 0 5 15 10 0 70 N - 0 RI M 75 50 1/3 18.5 0.28 25 20/ 46 10T 470893 4834583 0 10 30 40 15 5 N - 0 LP L 85 50 1/3 19.5 0.45 30 15/ 47 10T 470907 4834593 0 0 0 5 5 90 N - 0 RI L 65 70 1/3 15.2 0.37 30 15/ 48 10T 470919 4834602 20 20 0 0 0 60 Y ALGAE 50 LP L 70 65 1/2 18.1 0.4 20 5/1 49 10T 470921 4834605 0 10 5 5 0 80 Y GRASS 50 LP L 65 50 1/2 13.5 0.43 5 10/ 50 10T 470931 4834611 0 5 5 5 0 85 Y GRASS 50 LP L 65 65 1/3 12.5 0.58 20 *Riparian Vegetation: 1= tree, 2=shrub, 3=grass/forb, 4=bare

13 covered by aquatic vegetation or overhanging vegetation (Table 2). Areas without vegetation had very few Umpqua chub, including areas with other types of cover, such as boulders.

Otolith ageing

Otolith sections displayed well-formed alternating opaque and translucent bands that could be counted relatively easily. Sections revealed an opaque core surrounded by an opaque area in which growth checks were present. These checks, which seemed to be more prevalent in otoliths that were overground, made determining the position of the first annulus difficult. After viewing several samples, it was determined that the first annulus occurred at the outer edge of the opaque area surrounding the core. Despite difficulties in determining the first annulus,

Umpqua chub were relatively easy to age; ageing precision (APE) was relatively high at 4.23.

According to Campana (2001), an APE value of approximately 3.65 serves as a reference point for many fishes of moderate longevity and reading complexity. Difficulties in determining the first annulus led to our slightly higher APE. Ages agreed across all three reads for 75% of the individuals, and the spread in assigned ages was never greater than one year for any individual.

Umpqua chub aged in this study ranged from 23-65 mm FL and exhibited a maximum age of 7 years (Figure 2). Growth in length was relatively rapid until age 2 after which growth slowed and overlap in length-at-age was apparent.

There were significant differences in fork length and ages among drainages, with

Umpqua chub from Smith River being significantly smaller and younger and chub from the

South Umpqua at 3-C Rock being significantly larger and older than fish from the other drainages (ANOVA, p < 0.001 for both size and age, Figure 3A &B). Conversely, Umpqua chub from Smith River and 3-C Rock exhibited significantly higher and lower mean growth rates, respectively, compared to the other drainages (ANOVA, p <0.001, Figure 3C). A von

Bertalanffy growth curve for all Umpqua chub exhibits relatively rapid growth-in-length up to age 2, followed by a relatively rapid reduction in growth rate for older ages (Figure 4).

Microsatellite analyses

We found evidence suggesting that locus Ocr115 was not conforming to neutral

expectations, performed a sequence similarity search (discontiguous megablast) in NCBI’s

GenBank, and discovered that the sequence, DQ994902, is highly repetitive and shows a

significant sequence identity to several gene regions throughout the zebrafish (Danio rerio)

genome. Based on these findings, we did not include Ocr115 in further analyses.

The remaining ten microsatellites used to characterize Umpqua chub populations showed

variable levels of polymorphism with the number of alleles per locus ranging from 2 to 28 (mean

= 14.4, standard deviation = 9.969). After sequential Bonferroni corrections, we found no

evidence for linkage disequilibrium among locus pairs and all populations conformed to HWE.

Estimates of genetic diversity (A: mean # allele per locus, He: expected heterozygosity, and Ho: observed heterozygosity) were lowest in the Cow Creek population (A = 6.2, He = 0.578, and Ho

= 0.535; Table 3), with nine of the ten loci having an excess of homozygotes. Among the other

five populations, the Olalla Creek population showed the highest mean number of alleles (A =

9.5; Table 3) while expected and observed heterozygosities were greatest in the South Umpqua

River population (A = 8.4, He = 0.602 and Ho = 0.606; Table 3).

The overall level of genetic variation among populations (Fst) was 0.061 (95% C.I. = 0.04

- 0.08). Pairwise estimates of Fst were highly significant (P < 0.05) and ranged from 0.01 for

Olalla Creek and Calapooya Creek populations to 0.12 for Elk Creek and Smith River

populations (Table 4). Similarly, results from the exact tests for genic and genotypic

differentiation show that all population pairs are significantly differentiated with P values <

0.001 (Table 5).

Table 3. Estimates of genetic diversity based on ten microsatellite loci for Umpqua chub populations sampled for this study.

Population A He Ho Elk Creek 8.1 0.546 0.572

Calapooya Creek 9.4 0.567 0.580 Olalla Creek 9.5 0.570 0.590 Cow Creek 6.2 0.578 0.535

S Umpqua River 8.4 0.603 0.606

Smith River 6.9 0.564 0.576

Table 4. Pairwise estimates of genetic variation (Fst) among Umpqua chub populations sampled for this study. estimates are based on ten microsatellite loci. All values are significant with P ≤ 0.01.

Calapooya Creek Olalla Creek Cow Creek S Umpqua River Smith River

Elk Creek 0.042 0.040 0.069 0.051 0.120

Calapooya Creek 0.010 0.053 0.030 0.070

Olalla Creek 0.055 0.028 0.083

Cow Creek 0.040 0.121

S Umpqua River 0.105

Table 5. Genic and Genotypic P-values for each population pair across all loci. Genic Genotypic Population Pair df Chi2 P-value Chi2 P-value Elk Creek and Calapooya Creek 16 Infinity Highly sign. Infinity Highly sign. Elk Creek and Olalla Creek 18 Infinity Highly sign. Infinity Highly sign. Elk Creek and Cow Creek 18 Infinity Highly sign. Infinity Highly sign. Elk Creek and S Umpqua River 20 Infinity Highly sign. Infinity Highly sign. Elk Creek and Smith River 18 Infinity Highly sign. Infinity Highly sign. Calapooya Creek and Olalla Creek 18 42.759 0.0009 Infinity Highly sign. Calapooya Creek and Cow Creek 18 Infinity Highly sign. Infinity Highly sign. Calapooya Creek and S Umpqua River 20 82.24 0.00 Infinity Highly sign. Calapooya Creek and Smith River 18 Infinity Highly sign. Infinity Highly sign. Olalla Creek and Cow Creek 16 Infinity Highly sign. Infinity Highly sign. Olalla Creek and S Umpqua River 20 Infinity Highly sign. Infinity Highly sign. Olalla Creek and Smith River 18 Infinity Highly sign. Infinity Highly sign. Cow Creek and S Umpqua River 20 Infinity Highly sign. Infinity Highly sign. Cow Creek and Smith River 18 Infinity Highly sign. Infinity Highly sign. S Umpqua River and Smith River 20 Infinity Highly sign. Infinity Highly sign.

Pairwise estimates of Fst were significantly and positively related to distance in meters

2 (D) between sites (Fst = 0.021184 + 2.39111E-7*D; P=0.03, R adjusted=25.2%). However, for individual comparisons, the relationship was only significant for Calapooya Creek (Smith River,

P=0.96; Elk Creek, P=0.49; Calapooya Creek, P=0.01; Olalla Creek, P=0.21; Cow Creek,

P=0.08; and South Umpqua River, P=0.10). The coefficient of variation (CV) in pairwise Fst was also much lower in Smith River (22.7%) than in other populations (51.4 – 64.1%). Excluding

Smith River, the CV in pairwise Fst of the other populations fell into two groups: high variability in the central Umpqua tributaries - Calapooya Creek (54.5%) and Ollalla Creek (57.2%), and low variability elsewhere – Elk Creek (26.2%), Cow Creek (30.2%), and South Umpqua River

(28.4%). The greater variability in Calapooya and Ollalla creeks appears to be due to their low shared pairwise Fst about a third of any other pairwise comparison (Table 4).

The neighbor joining tree also suggests that the six nominal populations are genetically differentiated (Figure 5). The internal branches, however, are short with weak bootstrap support.

Again, the Smith River population and to a lesser extent, the Cow Creek population, appear to have diverged considerably more from the others.

Discussion

Field collections

Umpqua chub were either the first, second or third most abundant fish taxon in our samples, usually trailing redside shiners or speckled dace (Table 1). Because our sampling was focused on areas of high density as determined by Simon (2008) and we used methods

previously used with success on Oregon chub (Scherrer 2007), their relative abundance is likely

site-specific and gear-specific. As noted by Scherrer (2007), centrarchids are captured in low

numbers in minnow traps compared to their visual abundance. We found few centrarchids and no

smallmouth bass in our sampling, in part because of gear bias, but also because the sites were

previously shown to have Umpqua chub, and thus few centrarchids. On a broad scale, Umpqua

chub are found at 3rd and 4th order stream sites while smallmouth bass are absent from 3rd order streams and are detected at over 90% of 6th order streams in the Umpqua River (Markle et al.,

unpublished data).

The nominal Elk Creek Umpqua chub population had the most extensive distribution and

appeared to have the highest densities in previous work (Simon 2008). This population may gain

protection from Cunningham Dam (within 2.3 river km of 43.6587º N, 123.3902º W) on Elk

Creek that coincides with the upstream distribution of smallmouth bass and downstream

distribution of Umpqua chub. Our site-specific CPUE data did not suggest higher densities in Elk

Creek. However, we caution that the site-specific nature of this study should preclude larger

spatial extrapolation.

Our mark-recapture data were limited. Scherrer (2007) has been successful using minnow

traps with sampling densities of one trap every 100-250 m2 of surface area. We used comparable

sampling densities so our low recapture rates might indicate a large population, more extensive

movement in streams than in ponds that Scherrer sampled, handling mortality, or numerous other

possibilities. However, given that the design should have been able to detect displacement from 0

to 700 m, it is instructive to note that all four recaptures moved less than 100 m and three were

collected within 10 m of their original trap location. Clearly, a larger targeted effort would be

required to better describe the spatial extent of local populations and their abundance.

Otolith ageing

Otolith sections had well-formed alternating opaque and translucent bands that were

relatively easily to count, but the position of the first annulus was somewhat difficult. An opaque

core was surrounded by an opaque area with growth checks and we assigned the first annulus to

the outer edge of the opaque area surrounding the core. Umpqua chub from 23-65 mm FL

ranged in age from 1-7 yr. Growth in length was relatively rapid until age 2 after which growth

slowed and there was greater overlap in length-at-age. Our samples from the six populations

differed in size frequencies so we could not detect if there were differences in growth.

Consequently, differences in mean growth rates between populations are attributable to the age

structure of Oregon chub in the samples from those drainages. Smith River was the only

drainage that contained fish younger than age-2, while the sample from 3-C Rock consisted of a

larger proportion of fish older than 3 years. The closely related Oregon chub have a similar growth trajectory (asymptotic length of 62 mm versus 60 mm) and mature at 40 mm or age 2

(Scheerer and McDonald 2003). Depending on mortality rates, average Umpqua chub generation time could be expected to be about 4 yr.

Microsatellite analyses

In agreement with spatial patterns from surveys, our analyses of genetic variance and the neighbor joining tree suggest that the six nominal Umpqua chub populations are, in fact,

genetically differentiated from one another. In pairwise comparisons, only Calapooya Creek had a significant isolation by distance relationship. Pairwise estimates of Fst were highest for Smith

River (mean=0.100, range 0.070 – 0.121) and relatively uniform (CV=22.7%) for distances

ranging from 105-280 km. Excluding the very low Calapooya Creek and Ollala Creek pairwise

comparison, estimates of Fst outside Smith River were still less than half (mean=0.045, range

0.028 – 0.069), but also relatively uniform (CV=26.2 - 32.2%) for distances ranging from 57-241

km. Even if the latter is restricted to distances >105 km, the mean is still only 0.048.

These data may suggest isolation of Smith River fish prior to isolation of upstream

populations. There are several potential causes for Smith River isolation: salinity > 5 ppt due to

sea-level rise about 2500+ yr ago; higher salinity due to episodic tsunami events, the most recent

of which was about 330 yr ago; and predation from striped bass introduced about 70 yr ago. Our

microsatellite data are consistent with early isolation of Smith River, but we cannot currently apply a molecular clock to the microsatellite data to determine if one or more of these events might have been responsible for their isolation.

Future work

Future work should include system-wide monitoring, perhaps every five years unless there is evidence of additional population loss. At a minimum, the existence and spatial extent of each of the six populations should be documented during monitoring. Ideally, unbiased estimates of abundance and age structure should be collected, but our experience to date suggests that both

may be difficult to obtain. Refinement of the snorkel monitoring used by Simon (2008) and

further work with minnow trap sampling should both include evaluation of sampling efficiency

for different size classes and in different habitats. A rigorous evaluation of threats to each population should be completed. The Smith River population deserves special scrutiny as it is the

23 most genetically differentiated of the six populations and the most isolated. Their isolation may be due to salinity of the estuarine water near the mouth of Smith River or to striped bass predation rather than to smallmouth bass predation. As a result, threats to their existence appear to differ from the other five populations.

Acknowledgments

Authorization to collect samples in the Umpqua was permitted under 4d take permit

#14762. Funding for this project was provided by two United State Fish and Wildlife Service

Grants: 13420-9-J907 in 2008 and E-2-54 in 2010, along with in-kind match from Oregon

Department of Fish and Wildlife and Oregon State University. We are grateful for field assistance from Caulder Lonie, Kalene Onikama, Jay Potter, Jonathan Pender, Jeff Young, Brian

Jenkins, and Laura Jackson.

Literature Cited

Ardren WR, Amata L, Whelan J, DeHaan PW (2007) Characterization of 16 highly variable

tetranucleotide microsatellite loci for Oregon cub (Oregonichthys crameri) and cross

amplification in Umpqua chub (O. kalawatseti). Molecular Ecology Notes, 7, 808-810.

Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F (2004) GENETIX 4.05, logiciel sous

Windows TM pour la genetique des populations. Laboratoire Genome, Populations,

Interactions, Montpellier, France.

Beamish, RJ, and DA Fournier. 1981. A method for comparing the precision of a set of age

determinations. Can. J. Fish. Aquat. Sci.38:982-983.

Briggs, G. G. 1994. Coastal crossing of the elastic strain zero-isobase, Cascadia margin, south-

central Oregon coast. Masters thesis, Portland State University, 251 p.

Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation

procedures. Evolution 21, 550-570.

Campana, SE. 2001. Accuracy, precision and quality control in age determination, including a

review of the use and abuse of age validation methods. J. Fish. Biol. 59:197-242.

Daily, K. 1990. Smallmouth Bass Predation on Indigenous Fish Species in the Umpqua, Rogue

and John Day River Basins. Oregon Department of Fish and Wildlife.

Felsenstein J (2005) PHYLIP (phylogenetic inference package) version 3.6. Distributed by the

author. Department of Genome Sciences, University of Washington, Seattle, Washington.

Hey, J., and R. Nielsen. 2004. Multilocus methods for estimating population sizes, migration

rates and divergence time, with applications to the divergence of Drosophila

pseudoobscura and D. persimilis. Genetics 167:747-760.

Hey, J., and R. Nielsen. 2007. Integration within the Felsenstein equation for improved Markov

chain Monte Carlo methods in population genetics. Proceedings of the National Academy

of Sciences of the United States of America 104:2785–2790.

Ivanova NV, deWaard J, Hebert PDN (2006) An inexpensive, automation-friendly protocol for

recovering high-quality DNA. Molecular Ecology Notes, 6, 998-1002.

Markle, D. F., T. N. Pearsons and D. T. Bills, 1991. Natural history of Oregonichthys (Pices:

Cyprinidae) with a description of a new species from the Umpqua River of Oregon.

Copeia, 1991 (2): 277-293.

MacRae, P.S., and D.A. Jackson. 2001. The influence of smallmouth bass predation and habitat

complexity on structure of littoral zone fish assemblages. Canadian Journal of Fish and

Aquatic Science 58: 342-351.

Moore, K., K. Jones, J. Dambacher, C. Stein et al. 2010. Methods for Stream Habitat Surveys

(version 20.1). Aquatic Inventories Project, Corvallis, OR 97333.

Nelson, A. R. 1992. Holocene tidal-marsh stratigraphy in south-central Oregon- evidence for

localized sudden submergence in the Cascadia subduction zone. Quaternary Coasts of

the United States: Marine and Lacustrine Systems, Society for Sedimentary Geology

Special Publication No. 48: 287-301.

Page RDM (1996) TREEVIEW: An application to display phylogenetic trees on personal

computers. Computer Applications in the Biosciences,12 ,357-358.

Pearsons, T. N. (1989) Ecology and decline of a rare western minnow: the Oregon chub

(Oregonichthys crameri). Unpubl. M.S. thesis, Oregon State University, Corvallis,

Oregon.

Raymond M, Rousset F (1995) An exact test for population differentiation. Evolution, 49, 1280-

1283.

Rice WR (1989) Analyzing tables of statistical tests. Evolution, 43, 223-225.

Scheerer, P. D. and P. J. McDonald, 2003. Age, growth and timing of spawning of an

endangered minnow, the Oregon chub (Oregoinichthys crameri), in the Willamette

Basin, Oregon. Northwestern Naturalist, 84: 68-79.

Scheerer, P. D. 2007. Improved Status of the Endangered Oregon Chub in the Willamette

River, Oregon. In: Status, distribution and conservation of native freshwater fishes of

Western North America: a symposium proceedings. Transactions of the American

Fisheries Society Symposium 53:91-102.

Scheerer, P. D., and P. J. McDonald. 2003. Age and growth and timing of spawning of an

endangered minnow, the Oregon chub (Oregonichthys crameri), in the Willamette Basin,

Oregon. Northwestern Naturalist 84:68-79.

Secor, DH, JM Dean, and EH Laban. 1992. Otolith removal and preparation for microstructural

examination. In: DK Stevenson and SE Campana, (eds), Otolith microstructural

examination and analysis. Can. Spec. Pub. Fish. Aquat. Sci. 117, pp 19-57.

Simon, D. C. 1998. Distributional surveys of Umpqua chub Oregonichthys kalawatseti, 1987 and

1998. Final Report submitted to U. S. Fish and Wildlife Service, Oregon State Office

order no. 1448-13420-97-M216. December 16, 1998. 13pp.

Simon, D. C. 2008. Distributional surveys of Umpqua chub Oregonichthys kalawatseti and

smallmouth bass Micropterus dolomeiu the Umpqua Basin, Oregon. Final report

submitted to Pacific Northwest Cooperative Ecosystems Studies Unit, Cooperative

Agreement No. HAA003D00, for Bureau of Land Management, Roseburg, OR, August

14, 2008, 21 pp.

Simon, D.C. and D. F. Markle, 1999. Evidence of a relationship between smallmouth bass

(Micropterus dolomeiu) and decline of Umpqua chub (Oregonichthys kalawatseti)in the

Umpqua basin, Oregon. Northwestern Naturalist, 80:110-113.

Terwilliger M, T. Reece, & D. F. Markle. 2010. Historic and recent age structure and growth of

endangered Lost River and shortnose suckers in Upper Klamath Lake, Oregon,

Environmental Biology of Fishes.

Weir BS, Cockerham CC (1984). Estimating F-Statistics for the analysis of population structure.

Evolution, 38, 1358–1370.