Viability Criteria for Steelhead within the Puget Sound Distinct Population Segment

Introduction

Under the U.S. Endangered Species Act (ESA), the National Oceanographic and Atmospheric Administration's National Marine Fisheries Service (NMFS) is required to identify measurable and objective delisting criteria as part of recovery planning. These delisting criteria must describe the conditions under which a listed species or distinct population segment (DPS) is no longer in danger of extinction (endangered) or likely to become so in the foreseeable future (threatened). We define a viable DPS as one that is unlikely to be at risk of extinction in the foreseeable future; for this purpose, we adopted the viable salmonid population (VSP) criterion of a 100-year timeline (McElhany et al. 2000, Viable Salmon Populations and the Recovery of Evolutionarily Significant Units, NOAA Tech. Memo. NMFS-NWFSC-42) to evaluate risk of extinction. Ultimately, the identification of delisting criteria requires the consideration of technical analyses relating to viability, which are contained in this report, and policy decisions such as acceptable levels of risk, which are not. It presents the biological viability criteria recommended by the Puget Sound Steelhead Technical Recovery Team (PSS TRT). The framework and the analyses it supports do not set targets for delisting, nor do they explicitly identify specific populations or groups of populations for recovery priority. Rather, the framework and associated analyses are meant to provide a technical foundation for those charged with recovery of listed steelhead (Oncorhynchus mykiss) in Puget Sound, from which they can develop effective recovery plans at the watershed scale (and higher) that are based on biologically meaningful criteria.

This technical memorandum develops viability criteria for the Puget Sound Steelhead DPS as identified in the NMFS status review (Hard et al. 2007, Status Review of Puget Sound Steelhead, NOAA Tech. Memo. NMFS-NWFSC-81). The DPS was listed as a threatened species under the ESA in May 2007. Under the ESA, a threatened species is �"any species which is likely to become an endangered species within the foreseeable future throughout all or a significant portion of its range” and an endangered species is �"any species which is in danger of extinction throughout all or a significant portion of its range.” The viability of a threatened species is therefore at some risk, and in an attempt to quantify that risk, the current document was developed by the PSS TRT, which was composed of scientists from federal, state, tribal, and local government agencies with expertise in steelhead biology and management.

The primary purpose of this document is to recommend objective, measurable, biological criteria for assessing the recovery and progress toward recovery of the Puget Sound Steelhead DPS. In doing so, we applied these criteria to an assessment of the current biological status of the DPS. We do not provide recommendations for criteria for delisting of the DPS under the ESA, nor do we evaluate whether the DPS should or should not be listed. In addition to evaluating the biological requirements presented here, listing and delisting decisions require evaluations regarding particular listing factors and conservation measures, which are beyond the scope of this report. Thus the criteria presented here provide a necessary but not sufficient set of criteria to determine the ESA listing status of the DPS.

Approach to Developing the Criteria

We developed viability criteria based on the considerations laid out in the VSP document (McElhany et al. 2000, Viable Salmon Populations and the Recovery of Evolutionarily Significant Units, NOAA Tech. Memo. NMFS-NWFSC-42, online at http://www.nwfsc.noaa.gov/assets/25/6190_06162004_143739_tm42.pdf), which identified four key population parameters that influence the persistence of populations.

• Sizeâ€"the abundance of all life stages of the species (often only measured as adults);

• Growth rate (productivity)â€"production over the entire life cycle, often measured as recruits or returns per spawner, or as long-term population growth rate (ë);

• Spatial structureâ€"distribution of individuals among spawning and rearing habitat areas, and connectivity among those areas; and

• Diversityâ€"variation in traits (phenotypic and genotypic) among individuals within and among populations in the DPS.

The goal of recovery is not merely to meet a set of defined criteria, but rather to restore or repair ecological processes that lead to long-term sustainability of the resource. Our approach views the DPS as a complex structure with important processes operating at scales ranging from individual breeding aggregations to the entire DPS. Along this continuum, we identified three hierarchical units for defining attributes related to biological processes that characterize DPS status, from smallest to largest: demographically independent populations (DIPs), major population groups (MPGs), and the entire DPS. We developed VSP-based criteria that pertain to each of the 32 constituent DIPs within the three MPGs (Northern Cascades, Central and South Puget Sound, and Hood Canal and Strait of Juan de Fuca) that compose the Puget Sound DPS. These criteria were developed using the Puget Sound Steelhead DPS population structure the TRT identified in its previous publication (Myers et al. 2015, Identifying Historical Populations of Steelhead within the Puget Sound Distinct Population Segment, NOAA Tech. Memo. NMFS-NWFSC-128).

To achieve full recovery, steelhead populations in the Puget Sound DPS need to be robust enough to withstand natural environmental variation and even some catastrophic events, and they should be resilient enough to support harvest and habitat loss due to human population growth. Recovery therefore requires sufficient abundance and productivity of the DPS and diversity among and within its constituent populations distributed across the DPS's range. Achieving ESA goals requires biological sustainability into the foreseeable future (as well as the absence of threats, which are outside the scope of this report). DPS sustainability implies that the number and distribution of sustainable populations are sufficient and that DPS-level diversity is conserved. Avoiding endangerment requires biological persistence, that is, assurance that all significant parts of the DPS have a high likelihood of persisting over a VSP time horizon, which is 100 years. Achieving viability at each level implicitly requires meeting all criteria for the lower levels.

Our analysis of steelhead viability rests on two primary components: one component focused on evaluating the quantitative abundance information available for individual populations, and the other component aimed at assessing viability at multiple hierarchical scales (DIP¨MPG¨DPS) by incorporating qualitative and quantitative information pertaining to all four VSP criteria. For the first component.a quantitative analysis of individual population data.we applied two distinct population viability analyses (PVAs) to the DIPs that have sufficient information on abundance and productivity to support them. The first of these PVAs, based loosely on a conventional spawner-recruit analysis, estimates extinction risk from combinations of abundance and productivity across the parameter space. The second of these PVAs, based on an autoregressive state-space analysis of abundance time series, uses estimates of demographic stochasticity to predict future abundance and extinction risk.

For the second component, we developed a knowledge-based decision support system to incorporate these PVAs into a framework to characterize viability that can 1) accommodate quantitative and qualitative information, including habitat-based (rather than strictly biological) metrics, and 2) estimate viability at multiple hierarchical scales (i.e., DIPs, MPGs, and the entire DPS). This framework is based on a Bayesian network (BN, aka Bayesian belief network or Bayes net) of probabilities that links the key aspects of viability.abundance, productivity, diversity, and spatial structure.across the DPS and all its constituent populations. The network framework we designed links criteria at a variety of scales and aggregates them from population-level criteria, through MPG-level criteria, to criteria for the entire DPS.

Identification and Application of Viability Criteria

For each DIP, the components of population-level diversity are: 1) VSP abundance.adult and juvenile abundance relative to estimated capacity, and the probability that abundance would reach a specified quasi-extinction threshold (an abundance considered to be tantamount to functional extinction) within 100 years; 2) VSP productivity.the number of smolts per spawner and the number of adults per smolt (both of which influence long-term population growth rate), and the frequency of repeat spawners; 3) VSP diversity.the degree of hatchery influence (using a simple propagule-pressure analysis), variation in adult run timing, the fraction of smolts from resident fish, and current age structure relative to the historic distribution; and 4) VSP spatial structure.the fraction of intrinsic potential rearing and spawning habitat occupied, where intrinsic potential is the area of habitat suitable for steelhead rearing or spawning, at least under historical conditions.

A BN constructed for each DIP estimated DIP viability from each of these VSP parameters using Bayesian probabilities. The BNs for each DIP were then combined to produce a composite BN for estimating the viability of each MPG, with weights provided to consider DIP representation of each major life history type (summer run versus winter run). The composite BNs for each MPG were then combined to estimate the viability of the DPS as a whole, with the contribution of each MPG weighted equally. In combination with the requirement that all MPGs within a viable DPS must be viable, this approach ensured that for an DPS to be viable, a sufficient fraction of its constituent DIPs spread across all its MPGs, representing each major life history type, must be viable.

In defining viability criteria and metrics, we used this hierarchical decision support system to provide an assessment of the current status of the DPS by applying the metrics to data available through 2011. This analysis is provided for illustration to demonstrate how the criteria might be used in the future to assess progress toward recovery. The assessment framework evaluates viability using objective measures of spawner abundance, productivity, hatchery influence, spawner and juvenile distribution, and several measures of population diversity (e.g., resident fish contribution, spawn timing, and relative effective population size). Some of these measures were estimated from physical habitat data, in particular metrics extracted from Geographic Information System (GIS) data layers that measure intrinsic potential steelhead production. (The algorithm used by the TRT to estimate intrinsic potential spawning or rearing area and potential steelhead production is described in Appendix C of this report). Fully quantifiable criteria are not available for several metrics, so we used TRT members' judgment to evaluate those. In evaluating DIP-level viability, we gave the combined diversity and spatial structure metrics equal weight of the more quantifiable abundance and productivity metrics. We then used recent observations of population performance (primarily trends in abundance) to evaluate how certain we can be that the Puget Sound Steelhead DPS is sustainable under current conditions.

The PSS TRT considered two issues specific to steelhead viability for which little information is available in Puget Sound populations: the demographic contributions of a resident life history and the degree of iteroparity (repeat spawning). In determining the viability of steelhead DIPs in the Puget Sound DPS, the TRT considered the potential influence of co-occurring resident O. mykiss on anadromous steelhead demographics. The TRT concluded that in basins where anadromous O. mykiss abundance is below the quasi-extinction threshold, the risk of extinction is not necessarily 100% if resident O. mykiss are present below natural, long-standing migration barriers. Inclusion of resident fish in the viability criteria is dependent, however, on several conditions. First, the abundance of resident fish must be large enough to be self-sustaining (this will be largely dependent on the numbers of resident females). Second, there must be some evidence of interbreeding between the anadromous and resident forms (this will most likely be established by genetic analysis of the relationship between the resident fish and the most proximate anadromous population). To accommodate these considerations, the TRT included in the BN for DIP viability a diversity node describing the proportional contribution of smolts from resident fish.

Iteroparity is another primary characteristic of O. mykiss, with potentially substantial demographic consequences. Simulation analyses of the demographic consequences of iteroparity were conducted to develop a suite of age-structured matrix population projection models with annual time steps that incorporate density dependence, varying levels of repeat spawning, and varying levels of density-independent stochasticity and fishing mortality. The stochastic models were used to evaluate the influence of varying levels of repeat spawning on population abundance and age structure under varying levels of fishing.

The modeling indicated that the average proportion of repeat spawners in an adult steelhead population is relevant to population abundance and stream capacity in two ways. First, relative to a population with no repeat spawners, for a given average capacity of juveniles, a population with repeat spawners will have a larger average adult spawning population. Second, a given average number of adult spawners can be sustained by fewer juveniles when repeat spawners are present than when they are not. It is therefore likely that viable populations (DIPs) of steelhead can be sustained in smaller stream basins than is the case for semelparous Pacific salmon (Oncorhynchus spp.). The model results also indicated that repeat spawning provides increased levels of resilience compared to populations without repeat spawning. From a population rebuilding and recovery perspective, it appears that for small population sizes such as the ones considered in this modeling exercise, the value of specific levels of repeat spawning to population resilience (and repeat spawning) are most likely to be realized under a no-fishing scenario, regardless of the level of environmental variation (at least for the range and kind of variation employed in the models).

While these analyses are exploratory, they and the Integral Projection Model analysis applied to a wild Alaskan steelhead population in Appendix G of this technical memorandum tend to reinforce the TRT's conviction that iteroparity is an important consideration in a comprehensive evaluation of viability for anadromous coastal steelhead. The team therefore included in the BN for DIP viability a productivity node describing the influence of iteroparity on aspects of viability.

Viability Criteria for Puget Sound Steelhead

The listed unit under the ESA for O. mykiss, including anadromous steelhead, is the DPS, and this is the unit that must be considered for delisting. Thus delisting criteria must ultimately address the extinction risk of the DPS. In approaching the development of criteria for assessing viability (and subsequent delisting) of DPSs, we have relied on the language in the ESA, information described in the listing decision, concepts outlined in the VSP document, and published research describing salmon populations and their past or potential responses to environmental changes. The ESA in Section 4.2.1 lists five potential factors for decline that must be considered in species listing decisions:

1. The present or threatened destruction, modification, or curtailment of its habitat or range;
2. Overutilization for commercial, recreational, scientific, or educational purposes;
3. Disease or predation;
4. The inadequacy of existing regulatory mechanisms; and
5. Other natural or man-made factors affecting its continued existence.

NMFS identified all five factors as contributing to the endangerment of Pacific salmonids, including steelhead. In considering how viability criteria might inform population delisting requirements, the TRT also attempted to consider these factors for decline.

For the Puget Sound Steelhead DPS, the TRT developed in this report criteria for recovery of steelhead that share some similarities:

1. The viability (as reflected in abundance, productivity, diversity, and spatial structure) of a majority of steelhead populations in each of the MPGs across the DPS is detectably higher than currently, using conventional PVA.
2. At least 40% of steelhead populations in each of the three MPGs within the DPS achieve viability, depending on the historical biological characteristics and acceptable risk levels for populations within each region. (The threshold of 40% corresponds to an average probability of DIP viability of 67% if the DIPs that are not viable have probabilities of viability <55%.)
3. At least 40% of populations from each major life history type (summer run, winter run) historically present within each of the MPGs are viable using these criteria.
4. Natural production of steelhead from tributaries to Puget Sound not identified as primary spawning or rearing habitat for any of the 32 identified populations is sustained to provide ecological diversity and productivity sufficient to support DPS-wide recovery.

From these considerations, the PSS TRT developed the following constructs for developing viability criteria at the three hierarchical levels of DIP, MPG, and DPS.

DIP Viability

1. Using a simple three-bin classification (not viable, intermediate, viable), a DIP is scored 1, 2, or 3, depending on whether the probability of viability as computed by its viability BN is <40%, 40â€"85%, or >85%, respectively. For a DIP to be considered viable, its probability of viability must be at least 85%; the TRT considered this estimate to be sufficiently high to insure persistence over 100 years. To estimate this value, the mean abundance and productivity criteria for the candidate DIP are each double weighted; this produces DIP viability scores ranging from 6 to 18 points. DIPs with scores less than 11 are considered not viable, those with scores between 11 and 14 are considered to have intermediate viability, and those with scores greater than 14 (>85%, rounded up) are considered viable.

MPG Viability

2. An MPG is considered viable if 40% (rounded up) of its DIPs (including extinct as well as extant historical populations) are viable and mean DIP viability exceeds a threshold for viability.
   1. DIPs exhibiting distinct life history strategies (i.e., summer run versus winter run) will be considered separate components of the MPG. Therefore, a minimum of 40% of summer-run and 40% of winter-run populations within an MPG must be viable to achieve MPG viability.
   2. DIPs containing both winter-run and summer-run subpopulations predominantly exhibit the winter-run life history strategy in Puget Sound and will be considered winter-run for the purpose of estimating 2a. This approach should be reevaluated as data on summer-run populations becomes available.
   3. A viable MPG must, in addition to the criterion outlined in 2a and 2b, have a geometric mean (averaged over all its DIPs) score of at least 2.2 to be considered viable, using the 1â€"3 scale for individual DIPs described under the DIP Viability heading above.

DPS Viability

3. A DPS is considered viable only if all its component MPGs are viable.

The TRT applied the criteria to the 32 DIPs in the Puget Sound Steelhead DPS to provide a status assessment of the current viability of these units. This assessment indicates low viability for many DIPs, all MPGs, and the DPS as a whole. Nearly all DIPs have insufficient current abundance and productivity scores to be considered viable. Most DIPs also have low scores for diversity and spatial structure, largely because of extensive hatchery influence, low breeding population sizes, and freshwater habitat fragmentation or loss.

The Puget Sound Steelhead DPS is not considered to be viable by the TRT. The TRT concluded that the DPS is currently at very low viability; most of the 32 DIPs, all three MPGs, and the DPS as a whole are at low viability. Nearly all DIPs in both the Central and South Puget Sound MPG and the Hood Canal and Strait of Juan de Fuca MPG are not viable. Most of the DIPs score low for all four VSP criteria, and nearly all DIPs have insufficient current abundance and productivity scores to be considered viable. Most DIPs also have low scores for diversity and intermediate scores for spatial structure. The pattern of low viability is widespread throughout Puget Sound, across all three MPGs, and includes both summer-run and winter-run populations. The populations with highest viabilityâ€"and those with highest abundance and diversityâ€"are in northeastern Puget Sound (Northern Cascades MPG).