You are on page 1of 6

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

Refinements of the 2008 FCRPS Biological Opinion’s Adaptive Management and Contingency Planning Processes: Unexpected Severe Decline and Early Warning Triggers

Background

As part of the administration’s review of the FCRPS BiOp, scientists suggested that a refinement of the BiOp’s adaptive management and contingency planning processes could provide additional surety that the BiOp is implemented in a precautionary fashion through 2018. Specifically, the scientists suggested that additional “early warning” triggers be developed that would be sensitive to 1) unexpected declines in adult abundance and 2) natural disasters or environmental degradation (either biological or environmental) in combination with preliminary abundance indicators. They further advised that these triggers be based on simple metrics that are readily available.

This document describes two triggers – as refinements of the adaptive management and contingency planning processes – that are responsive to the scientists’ advice, are transparent to ongoing regional processes, and are not likely to result in a series of “false-positive” events.

The first trigger (Unexpected Severe Declines) relies upon 4-year rolling averages 1 of the estimated numbers of naturally produced adults at key locations (typically dams where fish can readily be counted) in the Snake and Columbia Rivers.

The second trigger (Early Warning) is a surrogate for the Unexpected Severe Decline trigger which considers both recent abundance information (in relation to the Unexpected Severe Decline triggers) as well as biological or environmental information that strongly suggests that substantially reduced productivity would be expected to continue for several additional years. The remainder of this document describes the formulation of these triggers in greater detail.

It is important to remember that triggering the proposed Unexpected Severe Decline or Early Warning Triggers within the term of the 2008 FCRPS BiOp is not an expected or even likely outcome. Indeed, under the 2008 BiOp, the abundance of these ESUs - on average - are

DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite

expected to increase over time. However, inclusion of these triggers as part of the 2008 BiOp's adaptive management and contingency implementation processes provides additional assurances that the 2008 BiOp is implemented in a precautionary fashion from the perspective of the ESA- listed salmon and steelhead species.

Deleted: if the assumptions in Deleted: are proven true, Deleted: would be

1 Note: Four succeeding years of declining abundance was considered as a trigger, but was rejected in favor of a four year average abundance trigger based on a review of the available data. A trigger using four succeeding years of decline was not robust and resulted in many “false negative” results. It was most often tripped following “peak” years when the abundance in the fourth year was not particularly low. Conversely, even a small, occasional uptick in numbers failed to trigger an action in past years when abundances were very low.

NOAA Document page 000049

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

Unexpected Severe Decline Trigger for Chinook Salmon and Steelhead

The purpose of the Unexpected Severe Decline Trigger is to detect unexpected and severe declines in the abundance of ESUs 2 so that rapid response actions can be implemented in a timely fashion to minimize or mitigate for an unforeseen downturn.

The metric of exceptionally low abundance measured over a four-year period was selected as a trigger for rapid responses for several reasons. First, this metric is relatively easy to measure in a rapid manner and is easily interpretable. The four-year period corresponds approximately to a generation for most ESUs, and the International Union for Conservation of Nature (IUCN) recommends examining population declines over a time period representing one generation. Furthermore, the Interior Columbia TRT used a four-year period over which to calculate quasi- extinction probabilities.

Step 1: The proposed approach uses US v. Oregon Technical Advisory Committee (TAC) run reconstructions (estimates of naturally produced adult Chinook Salmon 3 and steelhead) based on adult dam counts. Counts at Lower Granite Dam are used for Snake River ESUs, counts at Priest Rapids or Rock Island dams are used for Upper Columbia River ESUs and counts at Prosser Dam are used for the Yakima River population of Mid-Columbia River steelhead. The available data varies by ESU within the 1975 to 2008 time frame.

Step 2: The proposed approach uses four-year rolling averages of the TAC data (e.g., the average of the 1997 through 2000 returns make up the 2000 four-year average, 1998 through 2001 returns make up the 2001 four-year average, etc.). The four-year rolling averages are then sorted from high to low and plotted to create an exceedence curve - which depicts the percent of years in the data set in which the four-year rolling average was greater than a particular abundance level. See Figure 1 for Chinook salmon ESUs and Figure 2 for steelhead ESUs.

Step 3: An examination of the resultant exceedence curves for Chinook salmon (Figure 1) indicate that of the observed four-year average abundances, about 15-25% are relatively high; about 5-15% are relatively low, and the remainder are close to average, showing relatively little variation. The pattern is less clear for steelhead ESUs, whose distribution is more continuous.

The lowest four-year average abundances correspond to the historically low abundance levels observed during the early to mid- 1990s which precipitated the ESA-listing of these ESUs. Declines to these levels would indicate that ESUs have declined to levels that would be unlikely to occur, given the analysis in the BiOp. Thus, we propose that the 90 th percentile (red line on Figures 1 and 2) be used as a “hard” trigger for implementing Rapid Response Actions (see

2 ESUlevel adult abundance information is the most readily available information at present. Future refinements of the Unexpected Severe Decline trigger could potentially be extended to the Major Population Group (MPG) or the population level. 3 Chinook “Jacks” are excluded from this data as they are predominantly males which return to spawn after spending only a single year in the ocean.

NOAA Document page 000050

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite
DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite

NOAA Document page 000051

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

Contingency Planning and RM&E Document); and the 80 th percentile (orange line on Figures 1 and 2) be used as a “soft” trigger that would engage closer examination and potential implementation of Rapid Response Actions.

The 90% percentile exceedence level was selected as a threshold level because: 1) based on the curves (Figures 1), this is a level below which mean four-year abundances for Chinook salmon dropped rapidly. This level represents a marked departure from median abundance levels (especially for Chinook salmon), but is also somewhat above the lowest observed four-year period (for both Chinook salmon and steelhead) [ DOES THIS MEAN THE 1990 LEVEL’S? IF SO, MENTION THAT HERE, IE., THIS IS LARGER THAN THE LEVELS WE SAW IN THE

1990’S].

In addition, the Action Agencies conducted a prospective analysis of the likelihood that the Snake River spring/summer Chinook salmon ESU would fall below certain abundance

thresholds representing four year running averages.

The estimation was accomplished by fitting

a Beverton-Holt production function to the ESU level data, then projecting forward 24 years. Four thousand simulated trajectories were used in the probability calculation. The base case expected fraction was estimated from brood years 1978-1994. The "future" trajectories were initialized with the geometric mean of the time series of spawners from 1994-2003. The results of this analysis were very similar to the exceedence curves developed by NOAA.

DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite

Deleted:

Taken together, use of the 90 th percentile as a threshold protects against false negatives while kicking in well before historic low levels are reached.

DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite

As an additional precaution, the 80 th percentile will serve as a “soft” trigger requiring closer examination of the available data and the readying of Rapid Response actions for more rapid implementation if the ESU(s) in question continue to decline.

Table 1 summarizes the four-year average abundance levels corresponding (closest value or average of two nearest values) to the 90 th and 80 th percentiles in Figures 1 and 2.

Table 1. Summary of ESUSpecific Unexpected Severe Decline Triggers (Average 4year Abundance of Naturally Produced Adults)

Species

90 th Percentile Trigger

80 th Percentile Trigger

Deleted: The Action Agencies conducted a prospective analysis of the likelihood that the Snake River spring/summer Chinook salmon ESU would fall below certain abundance thresholds representing four year running The estimation was

averages. accomplished by fitting a Beverton-Holt production function to the ESU level data, then projecting forward 24 years. Four thousand simulated trajectories were used in the probability calculation. The base case expected fraction was estimated from brood years 1978-1994. The "future" trajectories were initialized with the geometric mean of the time series of spawners from 1994-2003. The results of this analysis indicated that an abundance threshold of 4,500 average adults would be expected to occur in only about 10% of the years.¶

SR fall Chinook

341

410

SR spring/summer Chinook

4,855

7,574

UCR spring Chinook

460

1,125

SR steelhead (A-Run)

6,795

7,814

SR steelhead (B-Run)

1,345

1,852

NOAA Document page 000052

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

UCR steelhead

903

1,097

MCR steelhead (Yakima R.)

766

970

Early Warning Trigger for Chinook Salmon and Steelhead

The purpose of the Early Warning Trigger is an additional failsafe that could be triggered before the Unexpected Severe Decline numbers are exceeded. Under this trigger, the agencies will evaluate the adult abundance numbers described above as well as environmental indicators that would suggest a continuing decline. The purpose of this trigger guards against unexpected and severe declines in the abundance of ESUs 4 that can be anticipated as a result of environmental indicators so that rapid response actions can be implemented in a timely fashion to minimize or mitigate for an unforeseen downturn. More specifically, this trigger would anticipate a continuing decline by evaluating two years of adult return information, preliminary biological information (e.g., jack counts predicting expected adult returns) and environmental indicators (e.g., indicators of ocean conditions predicting very low abundance of adult returns for recent outmigrants) or natural disasters (wide-spread forest fires, increased distribution and virulence of pathogens, new invasive species, etc.) that would be likely to substantially reduce the productivity (and abundance) of ESUs in the future.

DRAFT Pre ‐ Decisional Document / Attorney ‐ Client Privileged work product / Do Not Cite
Deleted: Deleted: detect likely
Deleted:
Deleted: detect likely

Deleted:

In other words, detect if it is

likely that an ESU, based on

NOTE: Initial assessments suggest that juvenile monitoring (numbers, sizes, condition, etc.) of interior Columbia River basin ESUs would likely provide information that could complement the adult monitoring information and further enhance the Early Warning Trigger in the future. Additional work will be required in order to inventory the current monitoring program, determine what additional monitoring might be needed, and assess how best to collect and use this information to inform the Early Warning Trigger at the ESU, MPG, or population scale.

Step 1: Determine if the most recent two-year average of adult returns is near the threshold levels used for the Unexpected Severe Decline Trigger (above).

Step 2: Determine if there are any biological or environmental indicators that would suggest that ESUs are likely to experience low abundance in the next two or more years. This information could include, as an example, extremely low jack counts (a preliminary biological indicator that next years returns will be much lower than average) and ocean indicators (both biological and environmental) that indicate that recent outmigrants are likely experiencing extremely poor ocean conditions that would be expected to result in substantially reduced numbers of naturally produced adults in the next two years.

4 ESUlevel adult abundance information is the most readily available information at present. Future refinements of the Early Warning trigger could potentially be extended to the Major Population Group (MPG) or the population level.

NOAA Document page 000053

DRAFT PreDecisional Document / AttorneyClient Privileged work product / Do Not Cite or Distribute August 4, 2009

Step 3: Assess whether there have been any "natural disasters" such as wide-scale forest fires, volcanic eruptions, rapid increases in the distribution or virulence of fish pathogens, or mud- slides that would be likely to substantially reduce the productivity of freshwater habitat or severely limit the ability of adults or juveniles to migrate to or from this habitat.

After evaluating each of the factors in steps 1-3, a determination would need to be made as to whether or not there is a reasonable likelihood that future adult returns would fall to levels triggering the Unexpected Severe Decline Trigger (see above) or the existing BA/BiOp trigger. If the determination is yes, then the Early Warning Trigger would affect the implementation of Rapid Response Actions.

NOAA Document page 000054