Introgressive hybridization may erode phenotypic divergence along environmental gradients, collapsing locally adapted populations into a hybrid swarm. Alternatively, introgression may promote phenotypic divergence by providing variation on which natural selection can act. In freshwater fishes, water flow often selects for divergent morphological traits in lake versus stream habitats. We tested the effects of introgression on lake-stream morphological divergence in the minnow Owens Tui Chub (Siphateles bicolor snyderi), which has been rendered endangered by introgression from the introduced Lahontan Tui Chub (Siphateles bicolor obesa). Using geometric morphometric analysis of 457 individual Tui Chub from thirteen populations, we found that both the native and introgressing parent taxa exhibited divergent body and caudal fin shapes in lake vs. stream habitats, but their trajectories of divergence were distinct. In contrast, introgressed populations exhibited intermediate body and caudal fin shapes that were not differentiated by habitat type, indicating that introgression has eroded phenotypic divergence along the lentic-lotic gradient throughout the historic range of the Owens Tui Chub. Individuals within hybrid populations were less morphologically variable than those within parent populations, suggesting hybrid adaptation to selective agents other than water flow or loss of variance by drift.

Fish Sampling

We collected fish in July and August in 2019 and 2020, except for one site (East Walker River) which we sampled in February 2021, using beach seines and backpack electrofishing. We sampled one lake and one stream population from each of the parent subspecies (Owens and Lahontan), and four lake and five stream populations of putatively introgressed populations (see Appendix S1 for site descriptions). We sampled a total of 457 individual Tui Chub (mean = 35 fish/site). For the Lahontan and hybrid populations, we humanely euthanized the fish upon collection. For the endangered Owens populations, we took photos and measurements of live anesthetized fish, then revived and released them. We measured all fish to the nearest millimeter (standard length), weighed to the nearest centigram (wet weight), and photographed from three perspectives (dorsal, ventral, and lateral) for morphometric analysis. 

Morphometric Analysis

For each fish, we placed sixteen ventral, eighteen lateral, and seven dorsal homologous landmarks (Armbruster 2012) using the R package ‘StereoMorph’ (Olsen & Westneat 2015) (Figure 2). A single researcher placed all landmarks to avoid potentially confounding effects. We performed generalized procrustes analysis (GPA) to align and adjust the raw landmarks, providing centroid size and shape coordinates (Gower 1975, Rohlf & Slice 1990) using the R package ‘geomorph’ (Adams et al. 2020). Sexual dimorphism has not been reported in Tui Chub so we did not perform separate morphometric analyses by sex. To calculate caudal fin aspect ratio, we digitally measured caudal fin height and surface area using the software ‘ImageJ’ (Schneider et al. 2012) (Figure 2). We calculated caudal fin aspect ratio (AR) as the squared height of the fin divided by the surface area: AR = h2/SA (Langerhans & Reznick 2010).

The uploaded dataset contains GPA-aligned landmark coordinates.  

Column descriptions:

id: a unique identifier for each fish (factor)

location: the collection site of the fish (factor)

loc.type: whether the site is considered a lake or stream (factor)

putative.subspecies: whether the population is Owens Tui Chub, Lahontan Tui Chub, or Hybrid (introgressed) (factor)

sl.mm: fish standard length reported in millimeters (numeric)

weight.g: fish wet weight reported in grams (numeric)

caudal.AR: caudal fin aspect ratio (unitless) (numeric)

D_X1 through D_X7 : x-coordinates for GPA-aligned dorsal landmarks 1 through 7, respectively. 

D_Y1 through D_Y7 : y-coordinates for GPA-aligned dorsal landmarks 1 through 7, respectively. 

V_X1 through V_X16 : x-coordinates for GPA-aligned ventral landmarks 1 through 16, respectively. 

V_Y1 through V_Y16 : y-coordinates for GPA-aligned ventral landmarks 1 through 16, respectively. 

L_X1 through L_X18: x-coordinates for GPA-aligned lateral landmarks 1 through 18, respectively. 

L_Y1 through L_Y18: y-coordinates for GPA-aligned lateral landmarks 1 through 18, respectively. 

Missing Values:

Fish standard length (sl.mm) is missing for 10 fish from White Mountain Research Station (WMRS): id #500-509