1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Received: 11 July 2022 Accepted: 10 August 2022

DOI: 10.1111/conl.12921

LETTER

Integrating local ecological knowledge, ecological

monitoring, and computer simulation to evaluate
conservation outcomes

Michelle María Early-Capistrán1 Elena Solana-Arellano1

F. Alberto Abreu-Grobois2 Gerardo Garibay-Melo3 Jeffrey A. Seminoff4
Andrea Sáenz-Arroyo5,6 Nemer E. Narchi7
1 Departamento de Ecología Marina, Centro de Investigación Científica y de Educación Superior de Ensenada, Ensenada, Baja California, México
2 Unidad Académica Mazatlán, Instituto de Ciencias del Mar y Limnología, Universidad Nacional Autónoma de México, Mazatlán, Sinaloa, México
3 Next Generation Sonoran Desert Researchers (N-Gen), Ensenada, BC, México
4 NOAA, Southwest Fisheries Science Center, La Jolla, California, U.S.A.
5 Departamento de Conservación de la Biodiversidad. El Colegio de la Frontera Sur (ECOSUR), San Cristóbal de Las Casas, Chiapas, México
6 Centro de Ciencias de la Complejidad (C3) de la Universidad Nacional Autónoma de México (UNAM), CDMX, México
7 CoLaboratorio de Oceanografía Social/Centro de Estudios de Geografía Humana, El Colegio de Michoacán—Sede La Piedad, La Piedad, Michoacán,
Mexico

Correspondence
Elena Solana-Arellano, Departamento de Abstract
Ecología Marina, Centro de Investigación Successful conservation of long-lived species requires reliable understanding of
Científica y deEducación Superior de
Ensenada, Carretera Ensenada-Tijuana long-term trends and historical baselines. We present a framework for evaluating
No. 3918, Zona Playitas, 22860, Ensenada, abundance trends and conservation outcomes for long-lived marine species by
Baja California, Mexico.
integrating local ecological knowledge (LEK), ecological monitoring, and com-
Email: [email protected]
puter simulation, tested on a case-study of long-lived and heavily exploited green
turtles (Chelonia mydas) in the Eastern Pacific. Models fit to LEK and monitor-
ing data indicate that turtle abundance is increasing, but only after ∼40 years of
safeguarding the species’ nesting and foraging habitats in Mexico. However, cur-
rent abundance is at ∼60% of baseline levels and historic population structure
has not been reestablished, indicating the need for sustained, long-term con-
servation actions. We demonstrate the potential of linking LEK and ecological
science to provide critical information for conservation, by establishing reference
baselines and gauging population status with a long-term historical perspective,
while promoting equitable and sustainable futures.

KEYWORDS
conservation, local ecological knowledge, long-term data, marine historical ecology, nonlinear
modeling, sea turtles, transdisciplinary research

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the
original work is properly cited.
© 2022 The Authors. Conservation Letters published by Wiley Periodicals LLC.

Conservation Letters. 2022;e12921. wileyonlinelibrary.com/journal/conl 1 of 10

https://doi.org/10.1111/conl.12921
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 10 EARLY-CAPISTRÁN et al.

1 INTRODUCTION As with many long-lived species (e.g., McDougall et al.,

2017), green turtle conservation status evaluation is com-
Successful conservation of highly migratory, long-lived plicated by the lack of knowledge on fundamental param-
marine taxa requires efforts and policies implemented eters including maturation age, life stage duration, and
across large spatiotemporal scales, along with locally population connectivity (Casale & Heppell, 2016; Seminoff
grounded data and practices (Mazaris et al., 2017; Vierros & Shanker, 2008). Green turtles are highly migratory and
et al., 2020). Globally, many large marine taxa have shown occupy a broad range of habitats across different life stages,
patterns of slow changes over millennia, rapid depletion separated by hundreds or thousands of kilometers (see
after the industrial revolution, and accelerated decline in Seminoff et al., 2015). Recruitment age to foraging areas
the 20th and 21st centuries (Longo et al., 2015; Lotze & varies widely, with estimates between 1 and 7 years. Simi-
Worm, 2009). In these scenarios, understanding conser- larly, estimates of age at maturity range from 15 to 50 years
vation status requires comparisons between current and globally, with ∼30 years as a broadly accepted estimate
preexploitation abundances. Achieving this is particularly (Avens & Snover, 2013).
challenging for species impacted by small-scale fisheries or Coastal foraging habitats are focal points for conserva-
subsistence hunting, for which technical or baseline abun- tion strategies as they (i) host turtles of both sexes and a
dance data are scarce or unavailable (e.g., Selgrath et al., broad range of age classes, providing fundamental insights
2018). into population structure, dynamics, and trends and (ii)
Long-term data generated from Local Ecological Knowl- are where green turtles spend most of their lives dur-
edge (LEK) and historical sources is critical for evalu- ing developmental and inter-breeding periods (Bjorndal
ating conservation outcomes for long-lived and histor- et al., 2005; Seminoff & Shanker, 2008; Seminoff et al.,
ically exploited marine megafauna such as sea turtles, 2003). However, globally, monitoring is heavily skewed
slow-growing fishes, and marine mammals (e.g., Lotze towards nesting beaches, providing limited demographic
& Worm, 2009; McClenachan et al., 2012; Sáenz-Arroyo data as only adult females and nesting output are quanti-
et al., 2005). Integrative approaches can contribute to fied (Patrício et al., 2021; Seminoff & Shanker, 2008). Given
sound conservation and management by establishing that sea turtle populations consist primarily of juveniles,
local baselines and recovery targets, evaluating population defining conservation and management policies based on
status, integrating the cultural dimensions of human- the adult fraction of the population is unrealistic (Casale
environment interactions, and supporting equitable and & Heppell, 2016). Thus, research on foraging habitats is
inclusive practices (e.g., Lee et al., 2018; Poe et al., 2014). crucial for informing conservation policy and practice.
However, there is a need for innovative methodologies We focus on the East Pacific green turtle (hence-
to integrate diverse knowledge systems and for guide- forth, green turtle), a distinct population segment (DPS)
lines to address practical challenges (e.g., Moon et al., (Seminoff et al., 2015). Critical foraging habitats along
2019). We present a framework for evaluating abun- the Baja California peninsula in northwestern Mexico
dance trends and conservation outcomes for long-lived provide a valuable case study given (i) their biological
marine species by integrating LEK, ecological monitor- importance for the East Pacific DPS (Seminoff et al.,
ing, and computer simulation, demonstrated through the 2015) and (ii) green turtles’ cultural importance for the
case study of green turtles (Chelonia mydas) in the Eastern region’s inhabitants as food staples and medicine since
Pacific. the earliest phases of human occupation in the termi-
Sea turtle species are highly susceptible to over- nal Pleistocene (∼12,000 BP) (e.g., Early-Capistrán et al.,
exploitation given their complex life-history and prolonged 2018). Research in historical ecology has demonstrated
maturation time. Worldwide, the long history of commer- that small human populations and geographic isolation
cial exploitation of green turtle meat and eggs has depleted kept captures sustainable until the mid-twentieth cen-
or extirpated several stocks (e.g., Chaloupka et al., 2008), tury (see Early-Capistrán et al., 2018, 2020a and references
with some populations collapsing as far back as the 18th therein). From 1960 to 1980, intensive commercial fishing
century (e.g., McClenachan et al., 2012). Local, regional, supplied green turtle meat to fast-growing cities along the
and international conservation efforts, including nesting U.S.–Mexico border, driving foraging populations to near
beach and habitat protection, bans or restrictions on direct extinction. This process coupled with intense egg collec-
use, and by-catch regulations, have reversed declines in tion at nesting beaches (Delgado-Trejo & Alvarado Díaz,
several populations (e.g., Hawai’i, Atlantic, Indo-Pacific) 2012; Early-Capistrán et al., 2018) (Figure S1). Long-term
(Broderick et al., 2006; Mazaris et al., 2017). However, this conservation and research began in the late 1970s as green
cosmopolitan species exhibits substantial inter-population turtles diminished, and in the early 1980s the fishery col-
variation, such that some stocks may increase as others lapsed (Early-Capistrán et al., 2020a; Seminoff et al., 2008)
decline or become locally extinct (Mazaris et al., 2017). (Figure 1).
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EARLY-CAPISTRÁN et al. 3 of 10

F I G U R E 1 General chronology of green turtle fishing and conservation in Bahía de los Ángeles. Fishing efficiency increased along with
demand between 1960 and 1980, leading to population collapse. Long-term conservation and research began in the late 1970s in response to
diminishing populations. Fishing permits for green turtles were suspended by the Federal Government in 1983 as the fishery collapsed.
Regulations and conservation measures across green turtle habitats increased from the late 1970s onward, reaching full legal protection with
the total ban in 1990 (Early-Capistrán et al., 2020a; Márquez, 1996; Seminoff et al., 2008)

Green turtles are classified as endangered under Mex- to long-lived taxa to understand long-term change and
ican legislation and by the IUCN, and all sea turtle inform management.
exploitation in Mexico is banned since 1990 (DOF, 1990;
IUCN, 2021). Decades of nesting beach and habitat pro-
tection across the species’ range in Mexico supported 2 METHODS
population increases since the early 2000s (Seminoff et al.,
2015). Unlike regions where large-scale human impacts 2.1 Study site
occurred centuries ago (cf. McClenachan et al., 2012), the
Baja California peninsula’s unique geographic and histor- Bahía de los Ángeles (BLA), Baja California, Mexico
ical conditions provide an important case study with an (28.951917◦ , −113.562433◦ ) is a warm-temperate foraging
established historical baseline within living memory (see area in the Gulf of California, which was one of the
Early-Capistrán et al., 2018, 2020a). most important green turtle fishing grounds in the East-
We demonstrate a novel framework for linking ecolog- ern Pacific (Early-Capistrán et al., 2018) (Figure S1). This
ical science, LEK, and computer simulation, promoting index site hosts significant in-water foraging aggrega-
conservation processes by harnessing local and scientific tions and systematic monitoring, which began in 1995 and
capacities (Game et al., 2015). We integrated scientific has continued as a collaboration between government,
monitoring data with historical catch rate reconstruc- academic, and nongovernmental institutions (Figure 1;
tions generated collaboratively with local fishers, resulting Early-Capistrán et al., 2020a). Establishing spatial and
in (to our knowledge) the longest available standardized numerical links between foraging areas and nesting rook-
time-series (1952−2018) worldwide for a critical sea turtle eries is challenging, due to (i) foraging areas’ frequent
foraging habitat. Our approach can be adapted and applied incorporation of individuals from multiple genetic stocks;
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 of 10 EARLY-CAPISTRÁN et al.

(ii) limited animal-borne telemetry data for the East Pacific of missing data (i.e., blank observations or unavailable val-
DPS; and (iii) limited data on foraging sites (Dutton et al., ues) (Commercial Fishing = 48%, Conservation = 42%).
2019; Seminoff et al., 2015). The index nesting beach for This systematic method is robust for scenarios with up to
this population is Colola, Michoacán, Mexico (18.297392◦ , 75% missingness (Takahashi, 2017) (Supporting Informa-
−103.410956◦ ), ∼1500 km southeast of BLA (Seminoff et al., tion 6; Figure S4). We assumed that values were missing
2003) (Figure S1). Colola is the only green turtle nesting site at random (i.e., missingness was not systematically related
in the Northeast Pacific with long-term data (>30 years), to catch rates) (Thurstan et al., 2014).
and accounts for 56%–71% of green turtles in the Gulf of Each missing value was replaced with a plausible value
California (Delgado-Trejo, 2016; Koch, 2013). generated using Markov Chain Monte-Carlo simulations
To establish baseline abundance and analyze long-term to sample subsets of observed values (van Buuren &
change before scientific monitoring (1952−1982), we used Groothuis-Oudshoorn, 2011) (Supporting Information 6).
a published data set of LEK-derived catch-per-unit-effort We generated m complete data sets, equivalent to the per-
(CPUE) estimates (Early-Capistrán et al., 2020b). For the- centage of missing values for each process, running 1500
oretical framing, full methodological details, data sets, and iterations (Bodner, 2008) (Figures S5 and S6). Observed
code used for LEK-derived CPUE estimation, see Early- values were retained.
Capistrán and collaborators (2020a, 2020b) and Supporting We fitted each of the m complete data sets separately to
Information 1 and 2. the model:
To evaluate changes in abundance after the implemen-
tation of conservation measures (1995−2018), we used 𝑌 𝑎⋅𝑒(𝛽𝑥) , (1)
CPUE values from in-water monitoring. Data were pro-
vided by author Seminoff (2003; NOAA, Unpublished where Y is the response variable (CPUE), x is the indepen-
raw data), Comisión Nacional de Áreas Naturales Pro- dent variable (year), and α and β are fitted constants. We
tegidas (CONANP) (unpublished raw data), and Grupo pooled results of the m fitted models using Rubin’s Rules
Tortuguero de Bahía de los Ángeles (unpublished raw (Dong & Peng, 2013) to obtain parameter estimates with
data). For details on monitoring procedures and data, see standard errors that (i) describe the uncertainty of imputed
Supporting Information 3. missing data and (ii) are unbiased and have valid statisti-
During in-water monitoring, turtles were captured with cal properties (van Buuren & Groothuis-Oudshoorn, 2011)
set-nets of the same design as in commercial green turtle (Figure S4) (Supporting Information 6.1–6.3; Table S3).
fishing (Seminoff et al., 2003). To enable direct comparison Validity was ratified through residual analysis (ei ∼ N(0,
across time, CPUE values from monitoring and LEK were σ2 )) (Nguyen et al., 2017) (Supporting Information 6.4).
standardized to one 12-h in-water set for a single 100 m We developed an ad hoc method to (i) visualize a pooled
net (Seminoff et al., 2003). Mean annual CPUE values trend line (broadly equivalent to regression line) and (ii)
were analyzed to evaluate central tendencies (see Early- derive 95% confidence intervals using Rubin’s Rules (Dong
Capistrán, 2020a, 2020b] and Supporting Information 2–4; & Peng, 2013; Nguyen et al., 2017) (Supporting Information
Figures S2 and S3 and Tables S1 and S2 for details). 6.5).

2.2 Analyses 2.2.2 Changes in population structure

2.2.1 Nonlinear models and Multiple We used curved carapace length (CCL) size distributions
Imputation by Chained Equations (MICE) from monitoring data to analyze changes in population
structure after the implementation of conservation mea-
We evaluated CPUE trends for two distinct processes: sures. CCL values were converted to life stages based
Commercial Fishing (1952−1983) and Conservation on mean nester size at Colola (82.0 cm CCL) (Seminoff
(1978−2018). We appended LEK values from 1978−1982 et al., 2015); i.e., adults > 82.0. We used Mann–Whitney
to the monitoring data set to interpolate values in the U tests (α = 0.05) to compare size and life stage composi-
temporal gap between commercial fishing (1983) and tion over time (Period 1: 1995−2005; Period 2: 2009−2018;
in-water monitoring (1995). This is justified, as conser- Figure S7).
vation efforts began in the late 1970s and increased over Given the high variability in size composition across
time (Figure 1) (Márquez, 1996; Seminoff et al., 2008) time, we used LEK to describe general trends before scien-
(Supporting Information 5). tific monitoring. LEK-based size composition is described
We ran Multiple Imputation by Chained Equations using weight (kg), which fishers recollected accurately as
(MICE) with the mice package in R 4.0.4 (van Buuren & the basis of payment for catches (see Early-Capistrán et al.,
Groothuis-Oudshoorn, 2011) to handle the high percentage 2020a).
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EARLY-CAPISTRÁN et al. 5 of 10

F I G U R E 2 Long-term trends in mean annual catch-per-unit-effort (CPUE) Annotations show key events in sea turtle conservation and
management: (1) start of permanent sea turtle research efforts at Bahía de los Ángeles (1979); (2) start of nesting beach protection at Colola,
Michoacán (1979); (3) suspension of green turtle fishing permits by the Federal Government (1983); (4) permanent ban on all sea turtle
capture in Mexico (1990); (5) start of in-water monitoring at Bahía de los Ángeles (1990). Shaded area shows data derived from Local
Ecological Knowledge (LEK). Values below x-axis represent missing data

3 RESULTS Despite the upward trend, maximum CPUE during

scientific monitoring (11.2 turtles/12 h; 2018) represents
Our results suggest increasing green turtle abundance after ∼60% of the maximum CPUE in the commercial fish-
∼40 years of conservation measures and ∼30 years of ery (18.5 turtles/12 h; 1959) (Figures 2 and 3). Further-
full legal protection (Figures 2 and 3). Both processes, more, median CPUE is significantly lower in scientific
Commercial Fishing and Conservation, are described by a monitoring (median = 0.66) than commercial fishing
nonlinear model (Equation 1). Commercial Fishing shows (median = 3.47) (Mann–Whitney U = 232, p < 0.05,
high α and negative β (α = 24.271, β = −0.820, p < 0.01; 95% C..[0.81−7.75]) (Supporting Information 8; Table S5).
R2 = 0.845), describing exponential decline. In contrast, Thus, green turtle abundance in BLA is comparable to the
Conservation shows low α and positive β (α = 0.002, mid-1960s but remains below baseline levels.
p < 0.01; β = 0.136, p = 0.746; R2 = 0.711), describing
prolonged latency followed by accelerated increase. Both
models have robust residuals and good fits (Figure 3; 3.1 Life stage and size distribution
Table 1).
The BLA population declined at a rate of 8.4% annually Patterns in life stage and size distribution suggest that
during commercial fishing, contrasting with 4.8% annual population growth is driven by juvenile recruitment, and
increase in response to conservation measures (Support- that historical population structure—with a slight juve-
ing Information 7; Table S4). Thus, declines occurred nile bias and substantial numbers of large adults—has
75% faster than increases. Notably, there is a prolonged not been reached (Figures 4 and 5). Abundance increases
latency between the initial implementation of conserva- after 2011 coincide with shifts toward a higher proportion
tion measures (1979), initial signs of increase (∼2000), and of juvenile turtles. Period 2 (2009−2018; 72.22% juveniles)
significant recovery after 2011 (Figure 2). The 33-year lapse shows a strong juvenile bias compared with Period 1
between nesting beach protection and increases at forag- (1995−2005; 55.71% juveniles). Median CCL was smaller in
ing habitats roughly corresponds with one generation (∼30 Period 2 (median = 75.5) than Period 1 (median = 80.8)
years) and one period of age-at-recruitment (∼3−5 years) (Mann–Whitney U = 71406, p < 0.001, 95% CI [1.99−5.20])
(Avens & Snover, 2013). (Supporting Information 9; Tables S6 and S7).
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 of 10 EARLY-CAPISTRÁN et al.

F I G U R E 3 Trends in catch-per-unit-effort (CPUE) for 1952−2018 using Multiple Imputation by Chained Equations (MICE). Data points
show mean value for each year (mean imputed values are for illustrative purposes only). Pooled parameter estimates and R2 values are shown
for Commercial Fishing phase (1952−1983; blue box) and Conservation (1978−2018; green box). Solid trend line shows pooled predicted values
across all m imputed models, and dashed lines show 95% confidence intervals for the pooled trend line (Supporting Information 6). Standard
errors were pooled according to Rubin’s Rules to account for within-model and between-model variance (Dong & Peng, 2013) (Supporting
Information 6.5). Marginal rug plots show density distributions of imputed values for Commercial Fishing (blue rug plot, n = 630) and
Conservation (green rug plot, n = 816) (see also Figures S5 and S6). Pooled 95% confidence intervals for parameter estimates and R2 values are
reported in Table 1.

TA B L E 1 Results of Multiple Imputation by Chained Equations (MICE) analysis with nonlinear model
Pooled std.
Parameter Pooled estimate Pooled 95% CI error p Value Pooled df Pooled R2
Commercial Fishing (1952–1982; LEK data) Model: y ∼ α (β•x)
; df (complete data) = 29; m = 48; e ∼ N(0, σ ) for mean residuals
2

α 24.271 [19.669 to 28.873] 2.189 <0.01 17.800 0.845 [0.692 to

0.926]
β −0.0820 [−0.101 to 0.00928 <0.01 24.060
−0.0628]
Conservation (1978–2018; LEK and Monitoring data*) Model: y ∼ α (β•x) ; df (complete data) = 38; m = 42; e ∼ N(0, σ2 ) for mean residuals
α 0.002 [−0.0122 to 0.00703 0.746 25.325 0.711 [0.364 to
0.0168] 0.890]
β 0.136 [ 0.0760 to 0.196] 0.0286 <0.01 17.376
Note: Bold type indicates significant results at α = 0.05. Parameter estimates, confidence intervals, standard error, and R2 values were pooled using Rubin’s Rules
to account for uncertainty of the missing data and variance within and between the m imputed models (Dong & Peng, 2013) (Supporting Information 6). 95%
confidence intervals for R2 values are shown in brackets. Pooled degrees of freedom are included to account for the effects of missing data (Supporting Information
6.6) (van Buuren & Groothuis-Oudshoorn, 2011). See also Figure 3.
*LEK values from 1978 to 1982 were appended to the monitoring data set to allow for interpolation of values in the temporal gap between LEK and monitoring
data sets.
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EARLY-CAPISTRÁN et al. 7 of 10

F I G U R E 4 Trends in size distribution from Local Ecological Knowledge (LEK) and monitoring data. Adapted from Early-Capistrán
et al. (2020a), table 4

F I G U R E 5 Total captures per year in scientific monitoring for adult (Curved Carapace Length, CCL > 82.0 cm) and juvenile (CCL ≤
82.0 cm) green turtles. Size at maturity is based on mean size of nesting females at Colola (Figueroa et al. cited in Seminoff et al., 2015).
Shaded area shows a time span with 30+ years of nesting beach protection in Colola (initiated in 1979) and 20+ years since the total ban on
sea turtle capture in Mexico (initiated in 1990). Values below the x-axis represent missing data

4 DISCUSSION trends at BLA are encouraging but remain below base-

line levels. Our results show that intensive, technologically
We demonstrate that integrating LEK, ecological moni- efficient commercial fishing—even by a small fleet—can
toring, and computer simulation can gauge conservation have severe impacts in short timeframes. In contrast, pop-
outcomes for long-lived and heavily exploited species ulation recovery requires decades of sustained protection
like sea turtles. Spatially and temporally widespread con- across habitats. This coincides with agent-based models,
servation measures (e.g., nesting and foraging habitat which suggest that time required for green turtle recov-
protection, by-catch regulation, protection from unsus- ery is approximately twice that of human impact (Piacenza
tainable use, etc.) are generating positive results for East et al., 2017). This pattern of fast decline and slow popu-
Pacific green turtles (Mazaris et al., 2017). Population lation growth is consistent with other long-lived marine
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 of 10 EARLY-CAPISTRÁN et al.

taxa, including large sharks, sirenids, and cetaceans the importance of foraging habitats as focal points of
(cf. Chaloupka et al., 2008; McClenachan et al., 2012). conservation research and practice.
Our results correspond with consensus among senior Patterns of historical green turtle abundance, decline
fishers that green turtles are abundant but below num- from overfishing, and growth following conservation
bers observed as young harpooners (Early-Capistrán et al., efforts are documented in several populations (Broderick
2020a). Furthermore, historical trends in size class distri- et al., 2006; Mazaris et al., 2017). LEK-based approaches
bution suggest that large adults were extirpated during the will become increasingly relevant as sea turtle popula-
commercial fishery (see Early-Capistrán et al., 2020a). The tions grow (e.g., Christianen et al., 2021), particularly
smaller mode size and absence of large adults (>150 kg) as conservation conflicts can arise when management
in ecological monitoring (Figure 4) suggest that historical frameworks overlook or discount sea turtles’ cultural and
population structure has not been reached. Considering material importance (cf. Barrios-Garrido et al., 2018). Our
green turtles’ slow growth rates, decades of continued results were only possible thanks to long-term, collabo-
protection at foraging areas may be required to increase rative efforts with the BLA community, whose LEK was
juvenile survivorship and prolonged adult growth in order indispensable for establishing baseline levels and evalu-
to recover historical population structure (Seminoff et al., ating conservation outcomes. Importantly, LEK must be
2003). Our results are consistent with theoretical models recognized for its inherent value and become integral to
that suggest ecological recovery may take at least four gen- conservation policy and practice, with self-determination
erations in populations where multiple life stages were as the guiding principle of scientific collaboration (Mawyer
impacted (Piacenza et al., 2017). Likewise, reduction in & Jacka, 2018).
mean body sizes has been observed in recovering green To succeed, future conservation and policy measures
turtle populations, as shifting recruitment dynamics drive for long-lived and highly migratory species must inte-
changes in population structure (sensu Hays et al., 2022). grate international and basin-wide approaches built upon
Synchronous over-exploitation across life stages and locally grounded efforts (e.g., Vierros et al., 2020). Social
habitats from 1960−1980 drove steep population declines dimensions to conservation are expanding as human
by simultaneously decreasing adult survivorship, hatch- impacts on marine environments increase. Thus, multi-
ling production, and juvenile recruitment (sensu Semi- ple forms of expertise—including local collaboration—are
noff & Shanker, 2008). Indeed, population trajectories at essential for creating diverse and inclusive approaches to
BLA largely coincide with trends at the Colola index conservation (e.g., Carman & González Carman, 2020).
nesting site, which was largely uninhabited until the Our methods provide a tangible and adaptable system that
1950s. Like BLA’s foraging population, the Colola rook- fosters innovative and equitable approaches to marine con-
ery has grown substantially since ∼2010: nesting rates servation science and practice. LEK, held by people living
increased 508% from 1982 (3383 nests/year) to 2015 (15,196 with and from the sea, is indispensable for comprehending
nests/year). Nevertheless, the degree of relative increase long-term change and building sustainable futures.
remains unclear due to the absence of preexploitation
baseline data (Delgado-Trejo, 2016). The availability of syn- AC K N OW L E D G M E N T S
chronous time-series for nesting and foraging habitats We thank the community of Bahía de los Ángeles and Pos-
will enable new insights into population dynamics (e.g., grado en Ciencias del Mar y Limnología, UNAM, where
life stage duration, reproductive phenology, etc.) and for this research was conducted. We thank CONANP and
development of age–structure models for this population. Grupo Tortuguero de Bahía de los Ángeles for monitor-
Our results suggest a robust causal relationship between ing data. We thank R. Ávalos, J. Candela, L. Crowder, and
abundance trends and key events in conservation and V. Koch. Data and code are available at www.github.com/
management policies (Figures 1 and 2). BLA’s foraging earlycapistran/miceNls.
population shows an encouraging trend related to decades
of conservation measures enacted over broad spatiotem- FUNDING
poral scales. Decades of fishing bans and nesting beach Research was funded by CONACYT doctoral fellow-
protection have generated a positive feedback loop of ship (#289695) to M.M.E.C. and approved by CICESE’s
increased survivorship and recruitment across life stages Bioethics Committee (#2S.3.1).
and habitats. Demographic models suggest that impacts
on adults and large sub-adults can generate dramatic pop- ORCID
ulation declines and, likewise, restoration of these life Michelle María Early-Capistrán https://orcid.org/0000-
stages can have the greatest contribution to recovery (e.g., 0003-2626-9262
Crowder et al., 1994; Piacenza et al., 2017), highlighting Nemer E. Narchi https://orcid.org/0000-0002-3508-3913
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EARLY-CAPISTRÁN et al. 9 of 10

REFERENCES reveal origin of the green turtles in San Diego Bay. Marine Biology,
Avens, L., & Snover, M. (2013). Age and age estimation in sea turtles. 166, 3.
In J. Wyneken, K. J. Lohmann, & J. A. Musick (Eds.), The biology Early-Capistrán, M.-M., Sáenz-Arroyo, A., Cardoso-Mohedano, J.-G.,
of sea turtles (pp. 97–133). Boca Raton, USA: CRC Press,. Garibay-Melo, G., Peckham, S. H., & Koch, V. (2018). Reconstruct-
Barrios-Garrido, H. A., Palmar, J., Wildermann, N., Rojas Izales, D., ing 290 years of a data-poor fishery through ethnographic and
Diedrich, A., & Hamann, M. (2018). Marine turtle presence in archival research: The East Pacific green turtle (Chelonia mydas)
the traditional pharmacopoeia, cosmovision, and beliefs of Wayuú in Baja California, Mexico. Fish Fish, 19, 57–77.
indigenous people. Chelonian Conservation and Biology, 17, 177. Early-Capistrán, M.-M., Solana-Arellano, E., Abreu-Grobois, F. A.,
Bjorndal, K. A., Bolten, A. B., & Chaloupka, M. Y. (2005). Evaluating Narchi, N. E., Garibay-Melo, G., Seminoff, J. A., Koch, V., & Saenz-
trends in abundance of immature green turtles, Chelonia mydas, Arroyo, A. (2020a). Quantifying local ecological knowledge to
in the greater Caribbean. Ecological Applications, 15, 304–314. model historical abundance of long-lived, heavily-exploited fauna.
Bodner, T. E. (2008). What improves with increased missing data PeerJ, 8, e9494.
imputations? Structural Equation Modeling: A Multidisciplinary Early-Capistrán, M.-M., Solana-Arellano, E., Abreu-Grobois, F. A.,
Journal, 15, 651–675. Narchi, N. E., Garibay-Melo, G., Seminoff, J. A., Koch, V., & Saenz-
Broderick, A. C., Frauenstein, R., Glen, F., Hays, G. C., Jackson, A. Arroyo, A. (2020b). Quantitative datasets and r code: Quantifying
L., Pelembe, T., Ruxton, G. D., & Godley, B. J. (2006). Are green local ecological knowledge to model historical abundance of long-
turtles globally endangered? Global Ecology and Biogeography, 15, lived, heavily-exploited fauna. PeerJ, 8, e9494.
21–26. Game, E. T., Schwartz, M. W., & Knight, A. T. (2015). Pol-
van Buuren, S., & Groothuis-Oudshoorn, K. (2011). MICE : Multivari- icy relevant conservation science. Conservation Letters, 8, 309–
ate imputation by chained equations in R. Journal of Statistical 311.
Software, 45. Hays, G. C., Taxonera, A., Renom, B., Fairweather, K., Lopes, A.,
Carman, M., & González Carman, V. (2020). Going beyond diverse Cozens, J., & Laloë, J. -O. (2022). Changes in mean body size in
worldviews for conservation: Response to Kohler et al. Conserva- an expanding population of a threatened species. Proceedings of
tion Biology, 34, 286–288. the Royal Society, 289, 20220696.
Casale, P., & Heppell, S. (2016). How much sea turtle bycatch is IUCN. (2021). The IUCN Red List of Threatened Species [WWW Doc-
too much? A stationary age distribution model for simulating ument]. IUCN Red List of Threatened Species. Version 2021–2.
population abundance and potential biological removal in the URL https://www.iucnredlist.org/en
Mediterranean. Endangered Species Research, 29, 239–254. Koch, V. (2013). 12 años de monitoreo de la tortuga negra (Chelonia
Chaloupka, M., Bjorndal, K. A., Balazs, G. H., Bolten, A. B., Ehrhart, mydas) en zonas de alimentación y crianza en el noroeste de méx-
L. M., Limpus, C. J., Suganuma, H. S., Troëng, S., & Yamaguchi, ico (Reporte técnico). Mexico: Grupo Tortuguero de las Californias
M. (2008). Encouraging outlook for recovery of a once severely A.C., La Paz, B.C.S.
exploited marine megaherbivore. Global Ecology and Biogeogra- Lee, L. C., Thorley, J., Watson, J., Reid, M., & Salomon, A. K. (2018).
phy, 17, 297–304. Diverse knowledge systems reveal social-ecological dynamics
Christianen, M. J. A., van Katwijk, M. M., van Tussenbroek, B. I., that inform species conservation status. Conservation Letters, 12,
Pagès, J., Ballorain, K., Kelkar, N., Arthur, R., & Alcoverro, T. e12613.
(2021). A dynamic view of seagrass meadows in the wake of suc- Longo, S., Clausen, R., & Clark, B. (2015). The tragedy of the com-
cessful green turtle conservation. Nature Ecology and Evolution, 5, modity: Oceans, fisheries, and aquaculture. Rutgers University
553–555. Press.
Crowder, L. B., Crouse, D. T., Heppell, S. S., & Martin, T. H. (1994). Lotze, H. K., & Worm, B. (2009). Historical baselines for large marine
Predicting the impact of turtle excluder devices on loggerhead sea animals. Trends in Ecology & Evolution, 24, 254–262.
turtle populations. Ecological Applications, 4, 437–445. Márquez, R. (1996). Las tortugas marinas y nuestro tiempo. La ciencia
Delgado-Trejo, C. (2016). Recovery of the black sea turtle in Michoacán, para todos. Mexico City: Fondo de Cultura Económica.
Mexico: Final report to the U.S. Fish and wildlife service (Techni- Mawyer, A., & Jacka, J. K. (2018). Sovereignty, conservation and
cal report). Morelia, Mexico: U.S. Fish and Wildlife Service and island ecological futures. Environmental Conservation, 45, 238–
Universidad Michoacana San Nicolás Hidalgo. 251.
Delgado-Trejo, C., & Alvarado Díaz, J. (2012). Current conservation Mazaris, A. D., Schofield, G., Gkazinou, C., Almpanidou, V., & Hays,
status of the black sea turtle in Michoacán, Mexico. In J. A. G. C. (2017). Global sea turtle conservation successes. Science
Seminoff, & B. P. Wallace (Eds.), Sea turtles of the Eastern Pacific: Advances, 3, e1600730.
Advances in research and conservation (pp. 263–278). Tucson: McClenachan, L. E., Ferretti, F., & Baum, J. K. (2012). From archives
University of Arizona Press,. to conservation: Why historical data are needed to set baselines for
Diario Oficial de la Federació (DOF). (1990). Acuerdo por el que se marine animals and ecosystems: From archives to conservation.
establece veda para las especies y subespecies de tortuga marina en Conservation Letters, 5, 349–359.
aguas de jurisdicción Federal del Golfo de México y Mar Caribe, así McDougall, C. A., Welsh, A. B., Gosselin, T., Anderson, W. G., &
como en las del Océano Pacífico, incluyendo el Golfo de California. Nelson, P. A. (2017). Rethinking the influence of hydroelectric
DOF 31/05/1990. development on gene flow in a long-lived fish, the Lake Sturgeon
Dong, Y., & Peng, C.-Y. J. (2013). Principled missing data methods for Acipenser fulvescens. PLoS ONE, 12, e0174269.
researchers. SpringerPlus, 2, 222. Moon, K., Blackman, D. A., Adams, V. M., Colvin, R. M., Davila,
Dutton, P. H., LeRoux, R. A., LaCasella, E. L., Seminoff, J. A., Eguchi, F., Evans, M. C., Januchowski-Hartley, S. R., Bennett, N. J.,
T., & Dutton, D. L. (2019). Genetic analysis and satellite tracking Dickinson, H., Sandbrook, C., Sherren, K., st. John, F. A. V., van
 1755263x, 0, Downloaded from https://conbio.onlinelibrary.wiley.com/doi/10.1111/conl.12921 by Nemer Narchi - UABC - Universidad Autonoma de Baja California , Wiley Online Library on [13/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 of 10 EARLY-CAPISTRÁN et al.

Kerkhoff, L., & Wyborn, C. (2019). Expanding the role of social Danemann & E. Ezcurra (Eds.), Bahía de los Ángeles: Recursos
science in conservation through an engagement with philosophy, naturales y comunidad: Línea base 2007 (pp. 457–494). México
methodology, and methods. Methods in Ecology and Evolution, 10, D.F.; San Diego, California: Secretaría de Medio Ambiente y
294–302. Recursos Naturales ; San Diego Natural History Museum, Tlal-
Nguyen, C. D., Carlin, J. B., & Lee, K. J. (2017). Model checking pan,.
in multiple imputation: An overview and case study. Emerging Seminoff, J. A., & Shanker, K. (2008). Marine turtles and IUCN red
Themes in Epidemiology, 14. listing: A review of the process, the pitfalls, and novel assessment
Patrício, A., Hawkes, L., Monsinjon, J., Godley, B., & Fuentes, M. approaches. Journal of Experimental Marine Biology and Ecology,
(2021). Climate change and marine turtles: Recent advances and 356, 52–68.
future directions. Endangered Species Research, 44, 363–395. Takahashi, M. (2017). Statistical inference in missing data by MCMC
Piacenza, S. E., Richards, P. M., & Heppell, S. S. (2017). An agent- and non-MCMC multiple imputation algorithms: Assessing the
based model to evaluate recovery times and monitoring strategies effects of between-imputation iterations. Data Science Journal, 16,
to increase accuracy of sea turtle population assessments. Ecolog- 37.
ical Modelling, 358, 25–39. Thurstan, R. H., Campbell, A. B., & Pandolfi, J. M. (2014). Nine-
Poe, M. R., Norman, K. C., & Levin, P. S. (2014). Cultural dimen- teenth century narratives reveal historic catch rates for Australian
sions of socioecological systems: Key connections and guiding snapper (Pagrus auratus). Fish and Fisheries, 2–16.
principles for conservation in coastal environments. Conservation Vierros, M. K., Harrison, A. - L., Sloat, M. R., Crespo, G. O., Moore, J.
Letters, 7, 166–175. W., Dunn, D. C., Ota, Y., Cisneros-Montemayor, A. M., Shillinger,
Sáenz-Arroyo, A., Roberts, C. M., Torre, J., & Cariño-Olvera, M. G. L., Watson, T. K., & Govan, H. (2020). Considering Indigenous
(2005). Using fishers’ anecdotes, naturalists’ observations and grey peoples and local communities in governance of the global ocean
literature to reassess marine species at risk: The case of the Gulf commons. Marine Policy, 119, 104039.
grouper in the Gulf of California, Mexico. Fish and Fisheries, 6,
121–133.
Selgrath, J. C., Gergel, S. E., & Vincent, A. C. J. (2018). Shifting gears: S U P P O RT I N G I N F O R M AT I O N
Diversification, intensification, and effort increases in small-scale Additional supporting information can be found online
fisheries (1950-2010). PLoS ONE, 13, e0190232. in the Supporting Information section at the end of this
Seminoff, J. A., Allen, C. D., Balazs, G., Dutton, P. H., Eguchi, T., Haas, article.
H. L., Hargrove, S., Jensen, M., Klemm, D. L., Lauritsen, A. M.,
MacPherson, S. L., Opay, P., Possardt, E. E., Pultz, S., Seney, E.,
Van Houtan, K. S., & Waples, R. S. (2015). Status review of the green
turtle (Chelonia mydas) under the endangered species act (NOAA How to cite this article: Early-Capistrán, M. M.,
technical memorandum no. NOAA-TM-NMFS-SWFSC-539). San Solana-Arellano, E., Abreu-Grobois, F. A.,
Diego, USA: NOAA. Garibay-Melo, G., Seminoff, J. A., Sáenz-Arroyo, A.,
Seminoff, J. A., Jones, T. T., Resendiz, A., Nichols, W. J., & Chaloupka,
& Narchi, N. E. (2022). Integrating local ecological
M. Y. (2003). Monitoring green turtles (Chelonia mydas) at a
coastal foraging area in Baja California, Mexico: Multiple indices
knowledge, ecological monitoring, and computer
to describe population status. Journal of the Marine Biological simulation to evaluate conservation outcomes.
Association of the UK, 83, 1355–1362. Conservation Letters, , e12921.
Seminoff, J. A., Reséndiz-Hidalgo, A., Jiménez de Reséndiz, B., https://doi.org/10.1111/conl.12921
Nichols, W. J., & Todd-Jones, T. (2008). Tortugas marinas. In G.