Greater Atlantic Region Policy Series


Download Greater Atlantic Region Policy Series


Preview text

Greater Atlantic Region Policy Series
[19-03]
Mortality Rate Estimates for Sea Turtles in Mid-Atlantic and Northeast Fishing Gear,
2012-2017
Carrie Upite1, Kimberly Murray2, Brian Stacy3, Lesley Stokes4, Sara Weeks2
1NMFS Greater Atlantic Regional Fisheries Office; 2NMFS Northeast Fisheries Science Center; 3NMFS Office of Protected Resources, University of Florida; 4NMFS Southeast Fisheries Science Center
Greater Atlantic Region Policy Series | NOAA Fisheries | Greater Atlantic Regional Fisheries Office 55 Great Republic Drive | Gloucester, MA 01930

ABSTRACT
The NMFS Northeast Sea Turtle Injury Workgroup reviewed all sea turtle interactions from 2012 to 2017 recorded by the Northeast Fisheries Observer Program (n=157) and At-Sea Monitoring (n=15) and interactions reported to the Sea Turtle Disentanglement Network (n=217). The workgroup first determined if the interaction occurred while the turtle was alive (e.g., not a carcass on the seafloor) and whether any injuries or other apparent effects were attributable to the interaction. If so, interactions were assigned to one of three injury categories with associated post-interaction mortality rates, or a determination of 100% mortality was applied according to the criteria in NMFS (2017). Sea turtle records are presented by major gear type for the fishery observer records (trawl, gillnet, dredge, pot gear) and by vertical fishing line, fish trap, or aquaculture gear for the entanglement records. The results are delineated by rolling 5-year periods (20122016; 2013-2017) to retain consistency with how previous regional results were presented and for applicability in Section 7 consultations. For the most recent 5-year time period (2013-2017), the resulting estimated mortality rate for observable interactions in trawl gear is 48%, gillnet gear is 73%, dredge gear is 40%, vertical line gear is 55% (or 61% if we include turtles that were not disentangled that we assumed died), and fish trap gear is 57%. The limited information on the aquaculture record precluded a mortality rate estimate for that gear. While NMFS calculated previous mortality rate estimates for trawl, gillnet, and dredge gear, this is the first estimate of post-interaction mortality rates in vertical fishing lines and fish traps (using entanglement data).
KEYWORDS
Sea turtles, Fisheries, Post-interaction mortality, Observer, Entanglement, Bycatch
The Greater Atlantic Region Policy Series is a secondary publication series based in the NOAA Fisheries Greater Atlantic Regional Fisheries Office in Gloucester, MA. Publications in this series
include works in the areas of marine policy and marine policy analysis. Please visit www.greateratlantic.fisheries.noaa.gov/policyseries/ for more information.
This document may be cited as:
Upite, C., Murray, K., Stacy, B., Stokes, L., Weeks, S. 2019. Mortality Rate Estimates for Sea Turtles in Mid-Atlantic and Northeast Fishing Gear, 2012-2017. Greater Atlantic Region Policy Series 19-03. NOAA Fisheries Greater Atlantic Regional Fisheries Office www.greateratlantic.fisheries.noaa.gov/policyseries/. 15p.

TABLE OF CONTENTS
BACKGROUND ............................................................................................................................ 3 METHODS ..................................................................................................................................... 3 RESULTS ....................................................................................................................................... 5 Fishery Observer Data (NEFOP and ASM).................................................................................. 5
Trawl Gear ............................................................................................................................... 6 Gillnet Gear ............................................................................................................................. 7 Dredge Gear............................................................................................................................. 8 Pot Gear ................................................................................................................................... 9 Summary of Fishery Observer Data 2012-2017...................................................................... 9 Sea Turtle Disentanglement Network Data................................................................................. 10 Vertical Fishing Line ............................................................................................................. 10 Fish Trap Gear ....................................................................................................................... 12 Aquaculture ........................................................................................................................... 12 DISCUSSION ............................................................................................................................... 13 ACKNOWLEDGEMENTS.......................................................................................................... 14 REFERENCES CITED................................................................................................................. 15
1

LIST OF TABLES
Table 1. The estimated rate of post-interaction mortality for the respective risk categories, as defined in NMFS (2017).......................................................................................................... 4
Table 2. The number of observer records reviewed as well as the records excluded from overall mortality rate estimates from 2012 to 2017............................................................................. 5
Table 3. The mortality rate assessment for sea turtles observed in trawl gear from 2012 to 2016. 6 Table 4. The mortality rate assessment for sea turtles observed in trawl gear from 2013 to 2017. 7 Table 5. The mortality rate assessment for sea turtles observed in gillnet gear from 2012 to 2016.
................................................................................................................................................. 7 Table 6. The mortality rate assessment for sea turtles observed in gillnet gear from 2013 to 2017.
................................................................................................................................................. 8 Table 7. The mortality rate assessment for sea turtles observed in dredge gear from 2012 to 2017.
................................................................................................................................................. 9 Table 8. Overall estimated mortality rate by gear type from 2012 to 2016 and 2013 to 2017....... 9 Table 9. The number of entanglement records reviewed from 2012 to 2017............................... 10 Table 10. The mortality rate assessment for sea turtles found entangled in vertical fishing line
gear from 2012 to 2016. ........................................................................................................ 11 Table 11. The mortality rate assessment for sea turtles found entangled in vertical fishing line
gear from 2013 to 2017. ........................................................................................................ 11 Table 12. The mortality rate assessment for sea turtles found entangled in fish trap gear from
2012 to 2016. ......................................................................................................................... 12 Table 13. The mortality rate assessment for sea turtles found entangled in fish trap gear from
2013 to 2017. ......................................................................................................................... 12
2

BACKGROUND
Mortality from incidental capture (bycatch) in fisheries can occur during the interaction or after turtles are released alive. The latter is referred to as post-interaction mortality (PIM) and results from delayed effects of physiological disturbances or trauma caused by capture. A determination of PIM is needed to characterize the full impact of federal fisheries on sea turtles, which is necessary under Section 7 of the Endangered Species Act (ESA).
NMFS defined criteria for estimating PIM in various fisheries (NMFS 2017). NMFS developed these criteria using available scientific studies in conjunction with veterinary and other expert opinion, primarily based on behavior and the presence and severity of injuries (Stacy et al. 2016). Information used to apply these criteria is collected by observers onboard commercial fishing vessels, through either the Northeast Fisheries Observer Program (NEFOP) or At-Sea Monitoring (ASM), or by personnel specifically trained and permitted to disentangle and release sea turtles. The latter personnel are part of the Northeast U.S. Sea Turtle Disentanglement Network (STDN), which was established in 2002 to reduce injuries and mortalities caused by entanglements primarily in the vertical lines of pot/trap fishing gear and to collect data on the interactions. The range of the STDN extends from Maine through Virginia. The STDN consists of a number of responders (see acknowledgement section), but the Center for Coastal Studies responds to the vast majority of entanglements in the Northeast.
Each NMFS region annually reviews records of incidental captures in trawl, net, and pot/trap fisheries to determine PIM (NMFS 2017). Previous estimates of PIM have been calculated for Northeast and Mid-Atlantic gillnet, trawl, and dredge fishing gear using Northeastspecific criteria and NEFOP/ASM data (Upite 2011, Upite et al. 2013, Upite et al. 2018). This document includes updated mortality rate estimates for those gear types using the national PIM criteria (NMFS 2017). Upite et al. (2018) reviewed one observed pot gear interaction, but the sample size was too small to produce a valid mortality rate estimate. The NMFS (2017) criteria included injuries often associated with the vertical lines of pot/trap gear, specifically those affecting the neck and appendages. Using the STDN data and the NMFS (2017) criteria, PIM for vertical fishing lines and fish traps can now be estimated.
METHODS
We applied NMFS’ national PIM criteria (NMFS 2017) to records of incidental captures in Northeast and Mid-Atlantic U.S. trawl, gillnet, dredge, pot, vertical fishing line, and fish trap gear from 2012 through 2017. Information reviewed included the turtle photos, video (when available), and observer comments on the vessel and trip information logs, incidental take logs, and sea turtle biological sample logs (for fishery observer cases) or STDN responder comments on the Sea Turtle Entanglement Reporting Form (for entanglement cases). Each workgroup member determined whether the turtle was likely captured while alive (ante-mortem) and whether any injuries or other apparent effects were attributable to the interaction. Interactions not attributed to the observed haul/tow/set were noted as such and excluded from further analysis. For ante-mortem interactions, turtles were placed into one of four categories with associated PIM rates (Table 1). Workgroup members reviewed each case to make independent injury determinations, which were then compared and discussed to develop consensus determinations. In order to be consistent with previous mortality estimates in Northeast and Mid-Atlantic fishing gear (Upite et al. 2013, Upite et al. 2018) and for applicability to Section 7 consultations,
3

mortality estimates were calculated over rolling 5-year periods by gear type. Two separate 5-year periods (2012-2016 and 2013-2017) are included in this document. The criteria in NMFS (2017) apply to all sea turtle species and life phases, and we combined all species in the mortality rate estimates for each gear type.
Table 1. The estimated rate of post-interaction mortality for the respective risk categories, as defined in NMFS (2017).

Low Risk

Intermediate risk

High risk

Incompatible with survival
(deceased)

Category

1Aa

1Bb

2

3

Estimated rate of

post-interaction

10%

20%

50%

80%

mortality

100%

a1A: fisheries at minimal risk of causing decompression sickness (DCS). Applies to fisheries operating at a depth
less than 40 m (22 fathoms). b1B: fisheries at risk of causing DCS. Applies to fisheries operating at a depth of 40 m (22 fathoms) or greater.

For those interactions where there was insufficient information on which to base the assessment, an injury category was not assigned. These cases were considered “unknown” and excluded from further analysis.
There were cases where a credible report verified an entanglement, but no response was mounted, the turtle was not located, or details were unclear on whether the turtle was disentangled. In these cases, the workgroup assumed the turtle was not released from the gear and eventually succumbed to the interaction (i.e., it was considered deceased). Because we made an assumption on mortality without a review of the injuries, but the entanglement report was credible and it is a reasonable assumption that the turtle died without any reported disentanglement, we calculated estimated mortality rates with and without these “assumed dead” cases.
In addition, there were other instances where the entanglement data were reviewed, but the case was not included in the overall mortality rate. This occurred when available information was unclear or sparse, but some assessment of the animal’s condition was possible. In such cases, we assigned a PIM category determination to make the most use of all available observations; however, we had less confidence in the assignments and, therefore, excluded these results from the overall mortality rate. These cases are noted as Less Confident (LC) 1, 2, or 3 in the results. The following conditions had to be met for a case to receive a less confident determination (as opposed to an unknown determination): there was a photo or other reliable information (e.g., interview) to confirm the entanglement (regardless of whether the configuration or injury were visible); there was a description of the areas of the body involved; there was a notation of presence/absence of visible injury; and there was some description of behavior that informed assessment. If these criteria for “Less Confident” were not met, the case determination was unknown.

4

For each Category 1 case, the depth was determined by the value recorded in the observer database (typically for the haul or trip, not where the interaction occurred). Following previously published procedures (NMFS 2017), cases with depths less than 40 m were given a 10% mortality rate, while the mortality rate for cases at depths of 40 m or greater was 20% to reflect the relative risk of decompression sickness. All of the observed entanglements occurred at depths less than 40 m, so the mortality rate for each of those Category 1 cases was 10%.

RESULTS

Fishery Observer Data (NEFOP and ASM)
This evaluation includes records documented by fishery observers in Northeast and MidAtlantic U.S. fisheries from 2012 to 2017. Table 2 depicts the number of observed sea turtle interactions we reviewed by year and gear type for each of the two 5-year time periods featured in this report. Observed sea turtle species included the Northwest Atlantic Ocean loggerhead Distinct Population Segment (Caretta caretta), the North Atlantic green Distinct Population Segment (Chelonia mydas), leatherback (Dermochelys coriacea), and Kemp’s ridley (Lepidochelys kempii) turtles, as well as unidentified sea turtle species. The mortality rates include an assessment of PIM (e.g., our determination of Categories 1, 2, 3) and those turtles found dead as a result of the interaction.
Table 2. The number of observer records reviewed as well as the records excluded from overall mortality rate estimates from 2012 to 2017. T=trawl; G=gillnet; D=dredge; P=pot.

NEFOP records ASM records TOTAL records reviewed Insufficient information
Not attributable to observed fishery
Total records with determinations 2012-2016 Total records with determinations 2013-2017

2012 34
5 39 T GD 27 12 0 3
T G 2 1
0

2013 24

1 25 T G 20 4 2

D T 1 23

T 2 4

T D 3 1

2014 27
9 36 G DP 12 0 1 1 G 1 0
138

2015

2016

31

20

31

20

T GDT GD

14 15 2 12 5 3

1

0

G

1

0

2

D 2

120

2017 21
21 T GD 13 6 2
2 G D 1 1
1 D 1

5

Trawl Gear

2012-2016

From 2012-2016, there were 96 observed interactions in trawls. After the records with insufficient information (n = 4) and not attributable to the current gear interaction (n = 3) were removed, PIM determinations were made for 89 interactions involving trawl gear. The resulting estimated mortality rate for observable interactions in trawl gear from 2012-2016 is 47% (Table 3).

Table 3. The mortality rate assessment for sea turtles observed in trawl gear from 2012 to 2016.

Loggerhead

Category 1

10% mortality
16

20% mortality
17

Category 2 50%
mortality
18

Category 3 80%
mortality
17

Leatherback

0

0

1

0

Kemp’s ridley

4

0

0

2

Green

1

0

1

1

Unidentified

0

1

0

0

TOTAL

21

18

20

20

Percentage of 24%

20%

22%

22%

turtles in each

category^

Dead turtles

2.1

3.6

10

16

(total *

mortality %)

^ The combined percentages do not equal 100% because of rounding.

100% mortality
8 1 0 1 0 10 11%
10

TOTAL

Estimated Mortality
Rate

76 2 6 4 1 89

41.7

47%

2013-2017

The NEFOP and ASM databases included 82 observed interactions in trawl gear from 2013-2017. Post-interaction mortality determinations were made for 77 interactions involving trawl gear, after those cases with insufficient information (n = 2) and not attributable to the current gear interaction (n = 3) were removed. The resulting estimated mortality rate for observable interactions in trawl gear from 2013-2017 is 48% (Table 4).

6

Table 4. The mortality rate assessment for sea turtles observed in trawl gear from 2013 to 2017.

Loggerhead
Leatherback
Kemp’s ridley
Green
Unidentified
TOTAL Percentage of turtles in each category Dead turtles (total * mortality %)

Category 1

10% mortality
13 0 6 1 0 20 26%

20% mortality
12 0 0 0 1 13 17%

2

2.6

Category 2 50%
mortality
12 1 1 1 0 15 19%
7.5

Category 3 80%
mortality
15 1 3 1 0 20 26%
16

100% mortality
7 1 0 1 0 9 12%
9

TOTAL
59 3 10 4 1 77
37.1

Estimated Mortality
Rate
48%

Gillnet Gear
2012-2016
For gillnet gear, there were 48 records reviewed from 2012 to 2016. After the records with insufficient information (n = 3) were removed, PIM determinations were made for 45 interactions involving gillnet gear. The resulting mortality rate for observable interactions in gillnet gear is 78% (Table 5).

Table 5. The mortality rate assessment for sea turtles observed in gillnet gear from 2012 to 2016.

Loggerhead Leatherback Kemp’s ridley Green Unidentified TOTAL Percentage of turtles in each category^ Dead turtles (total * mortality %)

Category 1

10% mortality
3 0 3 1 1 8 18%

20% mortality
0 0 0 0 0 0 0

Category 2 50%
mortality
4 0 0 0 0 4 9%

0.8

0

2

Category 3 80%
mortality
2 0 0 1 0 3 7%
2.4

^ The combined percentages do not equal 100% because of rounding.

100% mortality
17 2 5 0 6 30 67%
30

TOTAL
26 2 8 2 7 45
35.2

Estimated Mortality
Rate
78%

7

2013-2017
From 2013-2017, there were 42 observed interactions in gillnets and after records with insufficient information (n = 3) were excluded, PIM determinations were made for 39 interactions. The resulting mortality rate for observable interactions in gillnet gear from 20132017 is 73% (Table 6).

Table 6. The mortality rate assessment for sea turtles observed in gillnet gear from 2013 to 2017.

Loggerhead Leatherback Kemp’s ridley
Green Unidentified TOTAL Percentage of turtles in each category Dead turtles (total * mortality %)

Category 1

10% mortality
4 1 4 0 1 10 26%

20% mortality
0 0 0 0 0 0 0

1

0

Category 2 50%
mortality
1 0 0 1 0 2 5%
1

Category 3 80%
mortality
1 0 0 1 0 2 5%
1.6

100% mortality
15 1 5 0 4 25 64%
25

TOTAL
21 2 9 2 5 39
28.6

Estimated Mortality
Rate
73%

Dredge Gear
The dredge fishing gear cases only involved scallop dredges. The resulting mortality rate for observable interactions in dredge gear for both 2012-2016 and 2013-2017 is 40% (Table 7). However, given the small sample size of observer records for this gear type (n=3; Table 2), there is uncertainty with this mortality rate estimate.

8

Preparing to load PDF file. please wait...

0 of 0
100%
Greater Atlantic Region Policy Series