A25-09241044_Cooperative_Agreement_with_the_University_of_Arizona_for_Genetic_Analyses_for_Apache_Trout_Samples_Collected_from_t.pdf

Agreement #: A25-09241044

COOPERATIVE AGREEMENT

BETWEEN THE ARIZONA GAME AND FISH DEPARTMENT

AND

THE ARIZONA BOARD OF REGENTS ON BEHALF OF THE UNIVERSITY OF ARIZONA

This Cooperative Agreement (“Agreement”) is entered into between the Arizona Game and Fish

Commission (“Commission”) and its administrative agency the Arizona Game and Fish Department

(“Department”) and the Arizona Board of Regents on behalf of the University of Arizona

(“Cooperator” or “University of Arizona”) (collectively “Parties” and singularly “Party”) for the purpose of undertaking a genetic study to better understand the genetic status of Apache Trout in the

West Fork Little Colorado River.

WHEREAS, the Commission is authorized to enter into this Agreement pursuant A.R.S. § 17-

231(B)(7);

WHEREAS, the Commission has statewide responsibility for management actions related to fish and wildlife population within the State of Arizona;

WHEREAS, the Arizona Game and Fish Department (“Department”) acts under the authority of the Commission and implements the Department’s Mission Statement in a variety of ways including, but not limited to, implementing sound fish and wildlife management programs and implementing studies to determine the genetic integrity of fish and wildlife populations;

WHEREAS, the University of Arizona is authorized to enter into this agreement pursuant to

A.R.S.§ 15-1625(B);

WHEREAS, the Parties recognize and encourage a continued commitment to improving the understanding of the genetics of Arizona’s fish and wildlife populations;

Purpose and Mutual Benefits

In 2018, the Department identified potential Apache × Rainbow Trout hybrids in the West Fork Little

Colorado River (WFLCR). In response, Department staff collected 16 fin clips from Apache Trout in the lower sections of the WFLCR. The fin clips were sent to the U.S. Fish and Wildlife Service’s

(USFWS) Southwestern Native Aquatic Resources and Recovery Center (SNARRC) for genetic analysis to determine if the Apache Trout were genetically pure, or hybridized with Rainbow Trout.

The results of the analyses completed at SNARRC suggested the individuals were Apache Trout ×

Rainbow Trout hybrids. Following these initial test results, Department staff collected an additional

103 fin clips from a 3.0 km stretch of the WFLCR and sent them to SNARRC for analysis. The

SNARRC analyzed both the new 2019 samples and the 2018 samples, however this time the results indicated that all the individuals tested were genetically pure Apache Trout. In 2021, the USFWS, White Mountain Apache Tribe, and Department initiated a project with SNARRC to collect genetic and morphological data from Apache Trout populations to get a better understanding of genetic purity of populations rangewide. As part of this study, SNARRC reanalyzed the 2018 and 2019 samples from the WFLCR and this time concluded that all individuals had mixed Apache Trout and Rainbow Trout ancestry, indicating that they were Apache Trout × Rainbow Trout hybrids. Given the conflicting results from the three studies completed by SNARRC, the Department is interested in having an independent lab replicate the genetic analyses completed in the 2021 study to verify the results received from SNARRC. The objectives of this study are to:

1) Determine the genetic purity of Apache Trout in the West Fork Little Colorado River.

2) Verify the accuracy of the genetic testing results from SNARRC.

Department staff collected 188 fin clips from Apache Trout in the WFCLR in September and October of 2024. These samples will be provided to the University of Arizona to complete the genetic analyses.

The University of Arizona will compile the results of their analyses into a technical report that will help guide future management actions for Apache Trout in the WFLCR.

NOW, THEREFORE, it is the understanding of the Parties that:

1. The Department agrees:

A. To Provide Apache Trout genetic samples for analysis.

B. Reimburse the University of Arizona in an amount not to exceed $13,000 for activities outlined in this agreement. Funding under this agreement is not available for reimbursement of the Cooperator’s purchase of equipment. Equipment is defined as having a purchase cost of $5,000 or more per unit.

C. Make payment within 30 days of receipt of invoice(s) received by June 5, 2026 from the cooperator.

D. Serve as collaborators on all aspects of this project. Provide input on project planning, methods, and implementation, as well as interpretation of results and assistance with project report and manuscript preparation and/or review.

E. Retain the right to use, publish or share data resulting from the project.

2. The University of Arizona agrees:

A. Oversee and implement the scope of work as outlined in Exhibit A

B. Use the funding provided under this agreement exclusively to implement project activities outlined under this agreement.

C. Designate administrative and technical staff for this project.

D. Data and results generated from this project will be used solely for the purpose of testing for hybridization between Apache Trout and Rainbow Trout, and no other project without written permission from the Department.

E. Submit a draft technical report by May 1, 2025 documenting the results of the genetic analysis for Department review and a finalized technical report to the Department by June

1, 2026.

F. Submit all invoices by June 5, 2026. Payments will be made on a fixed-price basis and the

University of Arizona shall invoice the Department upon the Department’s completion and approval of the final technical report.

G. Data and results generated from this project will not be distributed to other organizations, individuals, or the public, or put on the internet without written permission from the

Department.

H. Data will not be sold or used for commercial gains.

3. The Parties mutually agree:

A. The Parties agree to use good faith efforts to collaborate on any joint publication(s) in accordance with scientific custom and publisher’s requirements. The publishing Party agrees to note the contributions of the non-publishing Party through acknowledgment or co-authorship, as appropriate. Either Party shall have the right, upon mutual written agreement, to release information or to publish any findings, analyses, writings, or material resulting from this Agreement or a Program Addendum, including its inventions or other discoveries, except for Confidential Information or sensitive information on the locations of Threatened and Endangered species. The publishing Party will furnish the non-publishing Party with a copy of any proposed publication in advance of the proposed publication date and grant the non-publishing Party thirty (30) days for review and comment. Within this period, the non-publishing Party may request the publishing Party in writing to remove any Confidential Information contained therein and/or delay such publication for a maximum of an additional eight (8) months in order to protect the potential patentability of any invention described therein. Such delay will not, however, be imposed on the filing of any student thesis or dissertation. The non-publishing Party’s failure to object to a publication or presentation within the thirty (30) day period above will be deemed acceptance by the non-publishing Party.

SWG-21 Implementation Grant (FY26)

Subrecipient unique entity identifier (UEI): ED44Y3W6P7B9

Federal award identification number (grant number or federal award number):

F25AF02137

Federal award start date: 07/01/2025

Federal award end date: 6/30/2026

Amount of federal funds obligated by the agreement: $13,000

Amount of federal funds obligated to each third party to the agreement: $13,000

Total amount of the federal award/grant): $1,527,205

Federal award/grant project description: AZ SWG-21 Implementation

Name of federal awarding agency: U.S. Fish and Wildlife Service

Catalog of Federal Domestic Assistance (aka CFDA) number and name: 15.634 State

Wildlife Grants

Identification of whether the award was for research and development: No

Negotiated and approved indirect cost rate for the federal award to AGFD: 51.99%

Negotiated and approved indirect cost rate for the third party: 55.0%

4. Effective Date and Duration. This Agreement is effective as of the last signature date and expires five (5) years from that date. Unless terminated earlier by the Parties, this Agreement shall be automatically renewed at expiration for an additional five (5) year term.

July 1, 2025 - June 30, 2026

5. Termination Generally. Either Party may terminate this Agreement upon thirty (30) days’ written notice to the other Party. Upon termination, all work performed pursuant to this

Agreement shall cease. If terminated, the Cooperator shall not incur any new obligations that would be paid from funds provided under this Agreement.

6. Notices. All written notices concerning this Agreement shall be delivered in person or sent by certified mail, return receipt requested, to the Parties as follows:

A. For the Department:

Vanessa Bilodeau

Agreements and Contracts Administrator

5000 W Carefree Highway

Phoenix, Arizona 85086

623-236-7530 vbilodeau@azgfd.gov

Zachary Beard

Statewide Native Aquatics Program Manager

5000 W Carefree Highway

Phoenix, Arizona 85086

623-236-7653 zbeard@azgfd.gov

B. Dr. Jessica A. Rick

1064 E Lowell St.

Tucson, Arizona 85721

520-621-0059 jrick@arizona.edu

Marcel Villalobos, Director, Postaward

Sponsored Projects Services

845 N. Park Ave., Rm 538

Tucson, AZ 85721-0158

520-626-6000 sponsor@email.arizona.edu

7. Non-discrimination. In carrying out the terms of this agreement, the Parties agree to comply, to the extent they apply, with Executive Orders No. 2023-01 and No. 99-09 prohibiting discrimination in employment, the provisions of which are incorporated herein by reference, and all other applicable Federal and State laws, rules and regulations, including the Americans with Disabilities Act.

8. Audit. Pursuant to A.R.S. § 35-214, all books, accounts, reports, files, electronic data, and other records relating to this Agreement shall be subject at all reasonable times to inspection and audit by the State of Arizona for five (5) years after completion of this Agreement.

9. Arbitration. To the extent required pursuant to A.R.S. § 12-1518, the Parties agree to use arbitration to resolve any dispute arising under this Agreement, with each Party to bear its own attorneys’ fees and costs.

mailto:vbilodeau@azgfd.gov mailto:jrick@arizona.edu

10. Termination for Conflict of Interest. This Agreement is subject to termination pursuant to

A.R.S. § 38-511.

11. Termination for Non-Availability of Funds. Every obligation of the Parties under this

Agreement is conditioned upon the availability of funds appropriated or allocated for the payment of such obligation. If funds for the continuance of this Agreement are not allocated or are not available, this Agreement shall terminate automatically on the date of expiration of funding. In the event of such termination, the Parties shall incur no further obligation or liability under this Agreement other than for payment of services rendered prior to the expiration of funding.

12. Other Agreements. This Agreement in no way restricts either Party from participating in similar activities with other public or private agencies, organizations, or individuals.

13. Compliance with Applicable Law. All work performed pursuant to this Agreement shall be in compliance with all applicable state and federal laws and regulations.

14. Integration. This Agreement constitutes the entire agreement between the Parties pertaining to the subject matter herein and accurately sets forth the rights, duties, and obligations of each

Party. All prior or contemporaneous agreements and understandings, oral or written, are hereby superseded and merged herein. The provisions of this Agreement may be abrogated, modified, rescinded, or amended in whole or in part only by mutual written consent executed by the Parties.

15. Severability. In the event that any provision of this Agreement or portion thereof is held invalid, illegal, or unenforceable, such provision or portion thereof shall be severed from this

Agreement and shall have no effect on the remaining provisions of this Agreement, which shall remain in full force and effect.

16. Illegal Immigration. Compliance Requirements for A.R.S. § 41-4401, Government

Procurement: E-Verify Requirement

1. The contractor warrants compliance with all Federal immigration laws and regulations relating to employees and warrants its compliance with Section A.R.S. § 23-214, Subsection A. (That subsection reads: “After December 31, 2007, every employer, after hiring an employee, shall verify the employment eligibility of the employee through the E-

Verify program.)

2. A breach of a warranty regarding compliance with immigration laws and regulations shall be deemed a material breach of the contract and the contractor may be subject to penalties up to and including termination of the contract.

3. Failure to comply with a State audit process to randomly verify the employment records of contractors and subcontractors shall be deemed a material breach of the contract and the contractor may be subject to penalties up to and including termination of the contract.

4. The State Agency retains the legal right to inspect the papers of any employee who works on the contract to ensure that the contractor or subcontractor is complying with the warranty under paragraph 1.

5. The Parties agree to comply with Executive Order 2005-30, “Ensuring Compliance with

Federal Immigration Laws by State Employers and Contractors,” the provisions of which are hereby incorporated by reference.

IN WITNESS WHEREOF, the Parties hereto have executed this Agreement as of the date below:

The Arizona Board of Regents on behalf of the University of Arizona

Christopher Barnhill

Contracts Manager

[Date]

ARIZONA GAME AND FISH DEPARTMENT

Clayton Crowder

Assistant Director, Wildlife Management

__12/9/25_____________________________

[Date]

12/11/2025

EXHIBIT A. Scope of Work

Scope of Work for the Arizona Game and Fish Department

Title:

Analysis of genetic samples from Apache Trout from the West Fork Little Colorado River

University of Arizona Contact and Principal Investigator:

Jessica A. Rick, Assistant Professor, School of Natural Resources and the Environment; University of

Arizona; 520-621-0059; jrick@arizona.edu

Background

In 2018, Arizona Game and Fish Department staff identified potential Apache × Rainbow Trout hybrids in the West Fork Little Colorado River (WFLCR). In response, Department staff collected sixteen fin clips from Apache Trout in the lower sections of the WFLCR. The fin clips were sent to the

Southwestern Native Aquatic Resources and Recovery Center (SNARRC) for genetic analysis to determine if the Apache Trout were genetically pure, or hybridized with Rainbow Trout. Geneticists at

SNARRC genotyped samples at mitochondrial locus (ND4) and 11 nuclear loci (microsatellites). The resulting data was compared to an existing database of Apache, Rainbow, and Gila Trout. The results found conflicting nuclear and mitochondrial signals, suggesting that all samples came from Apache

Trout × Rainbow Trout hybrids (Mussman 2018). Following these test results, Department staff collected an additional 103 fin clips from a 3.0 km stretch of the WFLCR in 2019 and sent them to

SNARRC for further genetic analysis. Both the 2018 and 2019 samples were re-genotyped at 11 microsatellite loci and one mitochondrial locus (ND4). Following this reanalysis, individual multilocus genotype analyses consistently revealed limited evidence of Apache Trout × Rainbow Trout introgression or backcrossing. Bayesian clustering analyses (e.g., STRUCTURE and NEWHYBRIDS) and a discriminant analysis of principal components suggested that samples from the WFLCR appeared more similar to two wild populations of Apache Trout from the East Fork White River and

Coleman Creek, Arizona. Based on these results the authors concluded that the Apache Trout in the

WFLCR were genetically pure, reversing the findings of the 2018 analysis (Weathers and Mussmann

2020). In 2021, the U.S. Fish and Wildlife Service, White Mountain Apache Tribe, and Arizona Game and Fish Department initiated a project with SNARRC to collect genetic and morphological data from

Apache Trout populations to get a better understanding of the genetic purity of populations rangewide and to investigate whether or not there are any distinguishable morphological characteristics that can be used to identify hybrid Apache Trout × Rainbow Trout in the field. As part of this study the

WFLCR samples from 2018 and 2019 were reanalyzed, this time using double-digest restriction-site associated DNA (ddRAD) data, which were used to identify single nucleotide polymorphisms (SNPs) distributed throughout the genome that could be used to distinguish between Rainbow Trout and

Apache Trout ancestry (see appendix 1 for full methodology). This analysis found that mixed Apache

Trout and Rainbow Trout ancestry was apparent from the samples collected in 2018 and 2019 from the

WFLCR (see appendix 2 for results), reversing the conclusions from the 2020 report. Given the conflicting results from the three studies completed by SNARRC the Department is interested in having an independent lab replicate the genetic analyses completed in the 2021 study to verify the results received from SNARRC.

Methods

On September 10, 2024 and October 15-16, 2024 Department staff collected 188 fin clips from Apache

Trout in the WFLCR. Apache Trout were sampled using single pass backpack electrofishing from 32 sites stretching from the upper Barrier on the WFLCR to the Fort Apache Indian Reservation

Boundary. The 32 sites identified in the WFLCR were all 100 m in length and were selected using a systematic sampling approach (Dauwalter et al. 2024). Sites were sampled starting at the downstream end and continuing upstream until 10 Apache Trout were captured or the entire 100 m site was sampled. If 10 Apache Trout were captured, sampling ceased at that site. The Apache Trout captured from each site had fin clips taken for genetic sampling. Fin clips were stored in vials, with 70% ethanol and brought back to Department headquarters for storage. Apache Trout total lengths varied from 42 -

254 mm. Using the methods described in (Appendix 3), University of Arizona staff will test the

Apache Trout genetic samples from the WFLCR for mixed Apache Trout and Rainbow Trout ancestry.

The results of this analysis will provide clarity on the previous genetic work completed by SNARRC.

Objectives

1) Determine the genetic purity of Apache Trout in the West Fork Little Colorado River

2) Verify the accuracy of the genetic testing results from SNARRC.

Deliverables

1) Final technical report documenting the results of the genetic analysis by June 1, 2026.

Literature Cited

Dauwalter, D., T. Gatewood, Z. Jackson, K. Manuell, Z. Beard, R. Follmuth. 2024. Apache Trout

Monitoring Plan: Revised. Trout Unlimited, Arlington, Virginia.

Mussman, S. 2018. Genetic Analysis of Apache Trout (Oncorhynchus apache) from the West Fork

Little Colorado River, AZ. U.S. Fish and Wildlife Service Southwestern Native Aquatic

Resources and Recovery Center - Report for the Arizona Game and Fish Department.

Weathers, T. C., and S. Mussman. 2020. Determining Apache Trout (Oncorhynchus apache) x

Rainbow Trout (Oncorhynchus mykiss) hybrid introgression within the West Fork of the Little

Colorado River, AZ. U.S. Fish and Wildlife Service Southwestern Native Aquatic Resources and Recovery Center - Report for the Arizona Game and Fish Department.

Appendix 1 - The methods used by the U.S. Fish and Wildlife Service analyze genetic and morphometric samples collected rangewide for Apache Trout in 2021.

Hybrid Production

Reciprocal hybrid crosses of Apache Trout broodstock and Eagle Lake strain Rainbow Trout were conducted. Apache Trout milt was collected from ten male fish at Williams Creek NFH and diluted at a 1:5 ratio (milt:extender) in StorFish Extender (IMV Technologies). The diluted milt was shipped overnight on ice to Bozeman Fish Technology Center (FTC) where milt from three individuals was used to fertilize eggs from each of three female Rainbow Trout sourced from Ennis NFH. The same process was used to ship Rainbow Trout milt from Ennis NFH to Williams Creek NFH. Fertilization was conducted off-site at each hatchery to mitigate the risk of hybrid offspring being erroneously incorporated into annual hatchery production. Fertilized eggs were then shipped overnight to the

Southwestern Native Aquatic Resources and Recovery Center where they were reared in Z-habitats matched to the hatchery water temperature of each egg group (Ennis NFH = 12.2 °C; Williams Creek

NFH = 11.1 °C). Fish were reared until reaching approximately 2-3 cm total length. All hybrid offspring were then euthanized using MS-222 and preserved in 95% ethanol.

Laboratory Methods

Genomic DNA was extracted from fin tissue samples preserved in 95% ethanol using the Nucleospin®

Tissue Kit (Macherey-Nagel, Allentown, PA). Double-digest restriction-site associated DNA (ddRAD) libraries were then prepared following Peterson et al. (2012). In summary, individual samples were ligated with 8 basepair (bp) barcodes and then pooled into libraries of 48 individuals each prior to size selection with a Pippin Prep (model PIP0001, Sage Science). DNA fragments of 246 – 296 bp

(excluding Illumina adapter sequence length) were retained by the Pippin Prep. This size range was optimized through in silico digest of a Rainbow Trout genome (NCBI assembly GCF_013265735.2)

(Chafin et al. 2018; Gao et al. 2021). Each library received a unique 6 bp index sequence that was added during 12 cycles of PCR amplification with Phusion polymerase (New England Biolabs).

Library sizes were verified on an Agilent 2100 Bioanalyzer. Library quantification was conducted with a NEBNext Library Quant Kit for Illumina (New England Biolabs). Paired-end 150 bp sequencing was conducted on an Illumina NovaSeq 6000 at either the University of Illinois at Urbana-Champaign Roy

J. Carver Biotechnology Center or the University of Oregon Genomics and Cell Characterization Core

Facility.

ddRAD Data Assembly

Genetic data were demultiplexed using the process_radtags module of Stacks v2.65 (Rochette et al.

2019). Reads with uncalled bases or Phred quality scores < 10 were discarded and a single nucleotide mismatch to known barcode sequences was permitted. Samples were aligned to the Rainbow Trout genome using the ‘--very-sensitive’ setting in Bowtie v2.45 (Langmead and Salzberg 2012).

Genotyping was conducted using default settings in the gstacks module of Stacks. The populations module was then used to filter loci for O. apache, O. clarkii, O. gilae, and O. mykiss individuals. Loci were retained if they were present in at least 70% of individuals from at least 12 sample groups. This was done to minimize potential bias that could be introduced by the unequal representation of species ancestry in our dataset. A minor allele frequency filter of 0.01 was also applied, and a single SNP per ddRAD locus was retained for downstream analysis. Individuals with higher-than-expected proportions of missing data (i.e., individuals with <300,000 Illumina reads passing quality filters) were identified and removed from the dataset (Cerca et al. 2021). The populations module was used to refilter the dataset as necessary to fulfill assumptions of various analytical programs as described below.

Hybrid Analysis

Species ancestry proportions for all O. apache, O. clarkii, O. gilae, and O. mykiss individuals were first assessed in the program Admixture v1.30 (Alexander et al. 2009; Alexander and Lange 2011).

This program was operated through the AdmixPipe v3.2 pipeline (Mussmann et al. 2023) to model species ancestry assuming presence of four species (K = 4) with 20 replicates performed. Data were thinned to maintain a minimum distance of 50,000 bp between SNPs. The output was assessed in

CLUMPAK (Kopelman et al. 2015) to check for multimodality of results among Admixture replicates.

The Admixture results were used to identify non-admixed Apache Trout and Rainbow Trout individuals. These included 48 Apache Trout (N = 12 each from East Fork White River, Upper Big

Bonito Creek, Ord Creek, and Soldier Creek) and 48 Rainbow Trout (N = 8 from each of the six domestic strains collected from Ennis NFH). These individuals were used as reference data for all subsequent programs that required diagnostic loci to distinguish Apache Trout from Rainbow Trout

(i.e., NewHybrids and Introgress).

Multiple approaches were used to identify hybrids and evaluate patterns of hybridization. Individuals were classified into discrete ancestry categories using NewHybrids v2.0 (Anderson and Thompson

2002). We followed the methods of Pritchard et al. (2016) to evaluate fifteen different ancestry categories that can be discerned from three generations of hybridization (Table 2). Input data were prepared for NewHybrids by rerunning the populations module of Stacks to identify SNPs common to both Apache Trout and Rainbow Trout (i.e., populations was rerun using previously described settings, but Cutthroat Trout, Gila Trout, and Chinook Salmon were excluded from the dataset). The resulting data were reduced to 200 loci with the greatest among-species differentiation and lowest linkage disequilibrium as determined in genepopedit (Stanley et al. 2017). Two hundred SNPs were selected to retain enough informative loci for hybrid identification while simultaneously reducing the dataset in size to facilitate computational efficiency in NewHybrids. Four independent replicates of NewHybrids were run using 100,000 Markov chain Monte Carlo (MCMC) burn-in generations followed by

1,000,000 MCMC generations of data collection. Reference genotypes for non-admixed individuals were designated using the ‘z’ option.

The power of the 200-SNP panel to accurately identify hybrids was assessed through simulation. The

200-SNP panel was used to simulate three generations of hybrid genotypes using the hybridize function of Adegenet v2.1.10 (Jombart 2008; Jombart and Ahmed 2011). One hundred genotypes were simulated for each ancestry category. Simulated data were analyzed in NewHybrids using the same methods as applied to the empirical dataset. A confusion matrix was prepared to identify classification errors associated with each ancestry category.

A hybrid index was calculated using the est.h function in the Introgress v1.2.3 R package (Gompert and Buerkle 2010). This method does not assign individuals to discrete categories as accomplished in

NewHybrids but allows for approximation of hybrid categories when plotted against interspecific heterozygosity values (i.e., the proportion of loci at which an individual is heterozygous for the two parental species alleles). This method is more computationally efficient than NewHybrids, so all fixed differences between Apache Trout and Rainbow Trout were retained. Interspecific heterozygosity was calculated using the calc.intersp.het function in Introgress and results were visualized using the triangle.plot function.

We evaluated historical introgression of Apache Trout populations by Cutthroat Trout and Rainbow

Trout using the four-taxon D-statistic test (Durand et al. 2011). Genetic loci were mapped onto a four-taxon phylogenetic tree comprised of the species of interest (Apache Trout), a sister taxon (Gila Trout), the species involved in a potential introgression event (Cutthroat Trout or Rainbow Trout), and an outgroup species (Chinook Salmon). The outgroup species was used to determine the ancestral (A) and derived (B) alleles at a locus. The number of genetic loci showing evolutionary events in conflict with a parsimonious explanation of evolution (i.e., ABBA and BABA genealogies) were presumed to result from hybridization and introgression. These were counted, and a χ2 goodness of fit test with one degree of freedom was performed to determine if the number of discordant genealogies differed significantly from the number expected by random chance. Statistically significant tests indicated that introgression had occurred. Data were refiltered in the populations module of Stacks to apply a weak minor allele frequency filter (0.01) and include all loci present in at least 70% of individuals from a minimum of four sample groups listed in Table 1. This was done to include all loci that are potentially informative in a four-taxon analysis (Eaton et al. 2017). All D-statistic calculations were performed in

COMP-D (Mussmann et al. 2020). SNP frequency formulas were employed to allow inclusion of heterozygous loci in D-statistic calculations (Durand et al. 2011; Eaton et al. 2015). A sequential

Bonferroni adjustment (Holm 1979) was conducted to account for type I error that could result from testing multiple hypotheses (α = 0.05).

Finally, a principal components analysis (PCA) was conducted in hierfstat v0.5 (Goudet 2005) using a matrix of individual genotypes to compare all Apache Trout and Rainbow Trout individuals. For this analysis, the populations module of Stacks was used to identify loci found in at least 90% of all

Apache Trout and Rainbow Trout individuals. This was done to identify variable loci while minimizing the misleading effects that missing genotype data can have on a PCA (Yi and Latch 2022).

These loci were randomly subsampled to retain 1,000 SNPs spaced at least 50,000 bp apart with a minor allele frequency ≥ 0.2 (Danecek et al. 2011). This was done to both reduce computation time and mitigate the impacts of pseudoreplication on genetic analyses that can occur in large genome-scale datasets (Waples et al. 2022).

Appendix 2. The results from the U.S. Fish and Wildlife Service analysis of genetic and morphometric samples collected rangewide for Apache Trout in 2021.

We aligned double-digest restriction-site associated DNA (ddRAD) data to the Rainbow Trout reference genome and used the Stacks pipeline to identify single nucleotide polymorphisms (SNPs) distributed throughout the genome that would help evaluate structuring among Apache Trout populations and distinguish Rainbow Trout ancestry from Apache Trout ancestry. After data filtering steps were completed, we retained 27,024 SNPs distributed throughout the genomes of the two species.

These SNPs were run through the program Admixture to assign proportions of Apache Trout and

Rainbow Trout ancestry to each fish sampled for this project. The bar plot below shows the results of that analysis. Blue = Rainbow Trout; Orange = Apache Trout.

The AZ21Hybrid and MT21Hybrid sample groups represent first-generation (F1) hybrids of Apache and Rainbow Trout created by crossing Arlee strain Rainbow Trout (Ennis National Fish Hatchery) with Apache Trout broodstock (Williams Creek National Fish Hatchery). Ancestry assignments for these samples match expectations for F1 hybrids, with each sample showing approximately 50% ancestry assigned to each species. These known-ancestry sample groups serve as positive controls which demonstrate that our methods are accurately identifying mixed-ancestry individuals.

Mixed ancestry fish were apparent in both Paddy Creek and West Fork of the Little Colorado River.

Some of these individuals have approximately 50% Rainbow Trout ancestry; however additional analyses are needed to classify these individuals to hybrid classes (i.e., are these fish first generation hybrids, second generation, etc.?).

Admixture identified small amounts of Rainbow Trout ancestry in some other Apache Trout populations. Additional analyses are ongoing to determine whether these small amounts of Rainbow

Trout ancestry could have resulted from past hybridization with Rainbow Trout. Alternatively, these patterns could either be statistical artifacts or result from natural evolutionary processes such as incomplete lineage sorting. However, Coon Creek appears the most likely of these populations to have undergone hybridization with Rainbow Trout within the past several generations.

Sample Groups:

Appendix 3. A detailed description of the genetic methodology to be used by the University of

Arizona to test for mixed Apache Trout and Rainbow Trout ancestry in the West Fork Little

Colorado River.

Laboratory Methods

Genomic DNA will be extracted from fin tissues in the Rick Genomics Lab at the University of

Arizona using New England Biolabs Monarch gDNA Extraction Kits, following the standard protocol with modifications to improve DNA yield as necessary. Double-digest restriction-site associated DNA

(ddRAD) libraries will then be prepared using MseI and EcoRI restriction enzymes following

Parchman et al. (2012). In summary, individual samples are ligated with unique 8-10 nucleotide barcodes (to allow individual identification) and then pooled into a single sequencing library prior to size selection with a Pippin Prep (Sage Science) for 350-450bp DNA fragments. Library sizes and size selection success will be verified on an Agilent 2100 Bioanalyzer at the University of Arizona Genetics

Core. Paired end 150 bp sequencing will then be conducted on an Element AVITI sequencing system at the University of Arizona Genetics Core.

Sequencing and Data Analysis

Genetic data will be filtered to remove sequencing reads with uncalled bases or with quality scores <

20, and demultiplexed to assign reads to individual fish using cutadapt. Reads will then be aligned to the rainbow trout reference genome (Pearse et al. 2019) using bwa (Li 2013) and variable sites will be identified using bcftools (Danecek et al. 2021). Using vcftools (Danecek et al. 2011), individuals will high amounts of missing data will be removed from the dataset and variable sites will be filtered to retain only those that are biallelic and have genotypes in a majority of the samples.

Hybrid Identification

To identify any evidence for hybridization within the 2024 samples, we will use a hierarchical

Bayesian clustering algorithm, implemented in the program entropy (Shastry et al. 2021). In this analysis, for each individual fish, we will estimate q (proportion of ancestry) and Q (interspecific ancestry) for a k = 2 model assuming two genetic clusters (i.e., Apache trout and rainbow trout). We will run three replicate chains and assess convergence of the model through examining trace plots. We will then use the bivariate relationship of q and Q to classify individuals into categories of hybrid crosses (following Mandeville et al. 2019, Rosenthal et al. 2022), to determine whether there is evidence for first-generation hybrids, second generation hybrids, backcrosses to each parental species, etc. The genetic affinities among individuals in our dataset will further be visualized using principal component analyses (PCA) of genetic dissimilarity matrices and discriminant analysis of principal components (DAPC).

Literature Cited

Danecek, P., Auton, A., et al. (2011). The variant call format and VCFtools. Bioinformatics, 27(15), 2156–2158.

Danecek, P., Bonfield, J. K., et al. (2021). Twelve years of SAMtools and BCFtools. GigaScience, 10(2), 1–4.

Li, H. (2013). Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM, 00(00), 1–3.

Mandeville, E. G., Walters, A. W., et al. (2019). Variable hybridization outcomes in trout are predicted by historical fish stocking and environmental context. Molecular Ecology, 28(16), 3738–3755.

Parchman, T. L., Gompert, Z., et al. (2012). Genome-wide association genetics of an adaptive trait in lodgepole pine. Molecular Ecology, 21(12), 2991–3005.

Pearse, D. E., Barson, N. J., et al. (2019). Sex-dependent dominance maintains migration supergene in rainbow trout. Nature Ecology and Evolution, 3(12), 1731–1742.

Rosenthal, W. C., Fennel, J. M., et al. (2022) Hybridization decreases native cutthroat trout reproductive fitness. Molecular Ecology, 31(16): 4224-4241.

Shastry, V., Adams, P. E., et al. (2021). Model-based genotype and ancestry estimation for potential hybrids with mixed-ploidy. Molecular Ecology Resources, 21(5), 1434–1451.