| |

Received: 26 March 2019    Revised: 26 October 2019    Accepted: 22 November 2019

DOI: 10.1111/jbi.13793

RE SE ARCH PAPER

Marine latitudinal diversity gradients, niche conservatism

and out of the tropics and Arctic: Climatic sensitivity of small
organisms

Wing-Tung Ruby Chiu1,2  | Moriaki Yasuhara1,2  | Thomas M. Cronin3 | Gene Hunt4 |

Laura Gemery3 | Chih-Lin Wei5

1
School of Biological Sciences, University of
Hong Kong, Hong Kong Abstract
2
Swire Institute of Marine Science, Aim: The latitudinal diversity gradient (LDG) is a consequence of evolutionary and
University of Hong Kong, Hong Kong
ecological mechanisms acting over long history, and thus is best investigated with
3
Florence Bascom Geoscience Center, U.S.
Geological Survey, National Center, Reston,
organisms that have rich fossil records. However, combined neontological-palaeon-
VA, USA tological investigations are mostly limited to large, shelled invertebrates, which keeps
4
Department of Paleobiology, National our mechanistic understanding of LDGs in its infancy. This paper aims to describe the
Museum of Natural History, Smithsonian
Institution, Washington, DC, USA modern meiobenthic ostracod LDG and to explore the possible controlling factors
5
Institute of Oceanography, National Taiwan and the evolutionary mechanisms of this large-scale biodiversity pattern.
University, Taipei, Taiwan
Location: Present-day Western North Atlantic.
Correspondence Taxon: Ostracoda.
Moriaki Yasuhara, School of Biological
Methods: We compiled ostracod census data from shallow-marine environments of
Sciences, University of Hong Kong, Pok Fu
Lam Road, Hong Kong. the western North Atlantic Ocean. Using these data, we documented the marine
Email: [email protected]
LDG with multiple metrics of alpha, beta (nestedness and turnover) and gamma diver-
Funding information sity, and we tested whether macroecological patterns could be governed by different
Seed Funding Programme for Basic
environmental factors, including temperature, salinity, dissolved oxygen, pH and pri-
Research of the University of Hong Kong,
Grant/Award Number: 201311159076 and mary productivity. We also explored the geologic age distribution of ostracod genera
201611159053; Research Grants Council
to investigate the evolutionary mechanisms underpinning the LDG.
of the Hong Kong Special Administrative
Region, China, Grant/Award Number: Results: Our results show that temperature and climatic niche conservatism are im-
HKU 17302518, HKU 17306014, and HKU
portant in setting LDGs of these small, poorly dispersing organisms. We also found
17311316; HKU Earth as a Habitable Planet
Thesis Development Grant evidence for some dispersal-driven spatial dynamics in the ostracod LDG. Compared
to patterns observed in marine bivalves, however, dispersal dynamics were weaker
Handling Editor: Werner Ulrich
and they were bi-directional, rather than following the ‘out-of-the-tropics’ model.
Main conclusions: Our detailed analyses revealed that meiobenthic organisms, which
comprise two-thirds of marine diversity, do not always follow the same rules as larger,
better-studied organisms. Our findings suggest that the understudied majority of bi-
odiversity may be more sensitive to climate than well-studied, large organisms. This
implies that the impacts of ongoing Anthropocene climatic change on marine ecosys-
tems may be much more serious than presently thought.

Equal contribution between W.T.R. Chiu, M. Yasuhara and C.L.Wei

Journal of Biogeography. 2020;47:817–828. wileyonlinelibrary.com/journal/jbi © 2020 John Wiley & Sons Ltd     817 |
 |
818       CHIU et al.

KEYWORDS

benthos, biodiversity, dispersal, macroecology, Ostracoda, temperature

1 |  I NTRO D U C TI O N efforts to quantitatively document their large-scale biogeography,

including their present-day LDGs. It is important to know whether
About two thirds of marine diversity comes from organisms smaller present-day ostracods show standard LDGs that match global me-
than 500  µm (Leray & Knowlton, 2015) and yet our macroeco- ta-analysis predictions (Hillebrand, 2004), or if they show less typical
logical and macroevolutionary studies are largely biased toward patterns (Krug, Jablonski, & Valentine, 2007). In this paper, we begin
larger ‘charismatic’ organisms (Fine, 2015; Jablonski, Huang, Roy, to address this shortcoming.
& Valentine, 2017; Tittensor et al., 2010). The present-day latitudi- It is increasingly clear that biodiversity is scale-dependent, and
nal diversity gradient (LDG), in which biodiversity decreases from decomposing these spatial scales into alpha, beta and gamma com-
the equator to the poles (Roy, Jablonski, Valentine, & Rosenberg, ponents can help us to better understand the mechanisms behind
1998; Willig, Kaufman, & Stevens, 2003), is a fundamental pattern the patterns (McClain & Schlacher, 2015; Rex & Etter, 2010; Worm
of the Earth's ecosystems. These gradients arise through the spa- & Tittensor, 2018). Alpha diversity reflects local coexistence of spe-
tial structure in origination, extinction and dispersal, and in the en- cies, gamma diversity is the regional species pool and beta diversity
vironmental or habitat capacity of species (Jablonski et al., 2017). can tell us about the relationship between alpha and gamma diversi-
The relative importance of these processes is likely to vary among ties (Jost, 2006, 2007, 2010; Whittaker, 1972).
taxonomic groups. Indeed, Jablonski et al. (2017) indicated that sim- Moreover, recent methodological developments (Baselga, 2010,
ilar-shaped LDGs of different taxonomic groups could arise from 2012) enable us to decompose beta diversity into two compo-
different mechanisms. The mechanisms can be better inferred by nents: nestedness and turnover. Nestedness is the extent to which
studying not just present-day organisms, but also their fossils that a species-poor community is a subset of a species-rich community
record macroevolutionary events as these patterns are assembled (Baselga, 2017), and is often used to identify the situation in which
in deep time. However, empirical studies investigating such mech- the species-rich region is the source and the species-poor region is
anisms using both present-day and fossil data remain very limited, the sink (Brault, Stuart, Wagstaff, & Rex, 2013; Stuart et al., 2017;
with the notable exception of bivalve molluscs (e.g. Jablonski et al., Tomašových et al., 2016). One example is the out of the tropics
2013; Jablonski, Roy, & Valentine, 2006). (OTT) model, in which the tropics is the source and the extratropics
It is important to better understand diversity patterns in sys- is the sink. Under this hypothesis (Jablonski et al., 2006), most taxa
tems other than the well-studied macrofaunal taxa, so that we may originate in the tropics and gradually expand their distribution into
get a better sense of outcomes in relation to various ecological and the extratropics while keeping their tropical presence. In contrast
evolutionary characteristics. Ostracods are a well-known meioben- to nestedness, turnover occurs with species replacement along an
thic taxon, and they are among the few meiobenthos with rich fossil environmental gradient and, thus, it is an indication of environmental
records (Horne, Cohen, & Martens, 2002; Mesquita-Joanes, Smith, filtering (Baselga, 2010). This is because species composition along
& Viehberg, 2012; Yasuhara, Tittensor, Hillebrand, & Worm, 2017). an environmental gradient will involve species replacement (turn-
They have the potential to be a good model system among small, over) if different species are specialized for different specific climate
marine invertebrates, comparable to how bivalves have become a ranges.
model taxon of large-scale spatial and temporal diversity analyses. By having a macrofaunal and a meiofaunal model system, we
Bivalves have become a model system because of their (a) well-doc- can compare these diversities and their components, and thus the
umented modern biogeography showing a strong LDG; (b) relatively mechanisms behind the LDGs. We can also use the rich fossil re-
high taxonomic, phenotypic, and functional diversity; and (c) rich cords of bivalves and ostracods to add a temporal dimension to the
fossil record (Jablonski et al., 2017). Ostracods presently fulfill the comparisons. Specifically, we here document the oldest geologi-
latter two criteria. They are estimated to have >20,000 extant spe- cal occurrences of taxa as an indication of where they originate.
cies that belong to a wide range of trophic groups, including deposit Different hypotheses make differing predictions about the spatial
feeders, suspension feeders and scavengers (Horne et al., 2002; distribution of such orginations. For example, the OTT model pre-
Rodriguez-Lazaro & Ruiz-Muñoz, 2012; Yasuhara & Cronin, 2008; dicts that taxa should disproportionately originate at low latitudes.
Yasuhara, Tittensor, et al., 2017). Ostracods also have a fossil re- Here we analyse alpha, beta and gamma diversity of a large
cord that is comparable in richness to that of bivalves, as a result compilation of northwestern Atlantic ostracod census datasets
of their well-calcified carapaces, small size, high abundance and and compare the results with their spatio-temporal distributional
species-level identification achievable from fossilized carapaces data and with the published bivalve results (Tomašových et al.,
(Holmes & Chivas, 2002; Huang et al., 2018; Yasuhara, Iwatani, et 2016). We show that this meiofaunal group has a standard LDG
al., 2017; Yasuhara, Tittensor, et al., 2017). Thus, the factor pre- similar to that of macrofauna and consistent with the global LDG
venting ostracods from becoming a better-established system to meta-analysis predictions (Hillebrand, 2004). However, we found
explore diversity dynamics is criterion a: the paucity of synthetic that beta diversity in these meiofaunal ostracods is more strongly
CHIU et al. |
      819

affected by climatic niche conservatism than in bivalves. Our re- 2.2 | Environmental parameters
sults suggest that the mechanisms behind the LDGs may be differ-
ent between these taxa, and possibly meiofauna and macrofauna Environmental parameters were obtained from various databases
more generally. (Appendix S2 and Table S3). Mean annual values (1955–2012) for
temperature (°C), dissolved oxygen (ml/L), and salinity were obtained
from the World Ocean Atlas (Baranova, 2015). pH data were ob-
2 | M ATE R I A L S A N D M E TH O DS tained from the Global Ocean Data Analysis Project Bottle Data (ver-
sion 2; Olsen et al., 2016). Mean sea surface net primary productivity
2.1 | Present-day shallow-marine ostracod data (NPP, in mgC/m2/day) data (1998–2014) were obtained from Oregon
State University Ocean Productivity Centre (http://www.scien​
We constructed a comprehensive, equator-to-pole compilation of ce.orego​nstate.edu/ocean.produ​ctivi​t y/; Behrenfeld & Falkowski,
ostracod censuses from the western North Atlantic and the Arctic 1997). Sea bottom environmental parameters were interpolated for
Oceans. We standardized ostracod taxonomy and integrated census each standard depth [based on World Ocean Atlas (Baranova, 2015)]
data from previously published studies, which involved restudy of using the Triangulated Irregular Network (TIN) interpolation from
previous collections. We obtained complete ostracod census data the plugin ‘interpolation’ of QGIS (QGIS Development Team, 2016)
from the original faunal slides from the US Geological Survey's with geographic range: latitude = 0–90°N; longitude = 100°W–30°E.
collections, including those of Cronin (1983), Cronin (1990), Hazel Interpolated environmental raster layers were overlaid with ostracod
(1970), and Lyon (1990), and Valentine (1971). Taxonomy was based sites in QGIS (QGIS Development Team, 2016) and Point Sampling
on Cronin (1990) with additional information on high latitude species Tool was used to extract the values of environmental parameters at
from Gemery et al. (2015). each site.
In addition to the census data from the USGS collections,
census data from Kontrovitz (1976), Teeter (1975) and the Arctic
Ostracode Database (Gemery et al., 2015; Yasuhara et al., 2012) 2.3 | Testing the factors governing marine
were added to our dataset after standardizing taxonomy using SEM biodiversity
pictures and descriptions from the literature. Locality information
(longitude, latitude and water depth) was obtained from the orig- Regression models were used to test potential controlling factors
inal cruise reports and literature (Figure 1; Appendix S1). Only driving sample-level diversity [E(S50)], including key environmen-
sites within the geographic range of 0–90°N and 100°W–30°E tal variables of bottom temperature, bottom salinity, bottom pH,
and water depths shallower than 200  m were used. Brackish- bottom dissolved oxygen, NPP and the distance from the sample
water sites (i.e. sites in estuaries or sites including > 10% brackish locality to the shore (km; Table S3). Because low salinity and low
genera such as Cyprideis and Perissocytheridea) were not included dissolved oxygen can depress diversity, sites with bottom salin-
in this dataset. Taxon counts considered one valve or articulated ity < 25 and bottom dissolved oxygen < 2 ml/L were omitted from
carapace as one specimen. Rarefaction, E(S50 ), the expected num- analyses so as to focus on fully marine, fully oxygenated conditions.
ber of species when 50 specimens are drawn at random, was used Distance to shore, dissolved oxygen and NPP were log10 -trans-
to analyse species alpha diversity; thus, only sites with ≥  50 os- formed, but results remain very similar even if all variables, or none
tracod individuals were used. This analysis was computed with of them, were log-transformed.
R (R Development Core Team, 2016) package ‘vegan’ (Oksanen, The Akaike's information criterion corrected for small sample
Blanchet, & Kindt, 2016). size (AICc) (Anderson, Burnham, & Thompson, 2000; Burnham &

F I G U R E 1   Spatial distribution
of ostracod alpha species diversity,
measured as E(S50), in the western North
Atlantic and Arctic Oceans [Colour figure
can be viewed at wileyonlinelibrary.com]
 |
820       CHIU et al.

Anderson, 2002; Hurvich & Tsai, 1989) was used to assess support level. Different sizes of latitudinal moving window (range: 1–10°)
for the models. Generalized least squares (gls) from the R package were also tested.
‘nlme’ (Pinheiro, Bates, & DebRoy, 2014) was used to test for pre-
dictors of diversity while accounting for spatial autocorrelation. We
fit the model assuming either exponential, linear or Gaussian decays 2.6 | Range sizes and per-genus species richness
of spatial correlation with distance (Pinheiro et al., 2014). Of these,
exponential decay was much better supported according to AICc and Latitudinal ranges sizes were computed for each species and genus
so we present those results. We also computed model-averaged re- to the nearest degree. For each latitudinal bin, we computed the
gression coefficients for environmental parameters across all mod- median species and genus latitudinal range size for all taxa found
els based on the model weights using the R (R Development Core in that bin. We also calculated the mean per-genus species rich-
Team, 2016) package ‘MuMIn’ (Barton, 2017). ness in each latitudinal bin as the mean of the number of species
in each genus that occurred anywhere in our database, following
Tomašových et al. (2016). These metrics were also calculated for
2.4 | Multiplicative diversity partitioning the bivalve data from Tomašových et al. (2016), however, trun-
cated to include only the western North Atlantic region for com-
Within each 10-degree moving window of the latitudinal transect, parison with our study.
alpha, beta and gamma diversities were computed with the ‘true’
diversity approach of Whittaker (1972) and Jost (2006) using the R
package ‘simba’ (Jurasinski & Retzer, 2012). In the diversity partition, 2.7 | Geologic ages of ostracod genera
gamma diversity is the total species richness, alpha diversity is the
mean number of species per site and beta diversity is the gamma Geologic age as the oldest stratigraphic occurrence of each os-
divided by alpha diversity within the window. Correlations with lati- tracod genus was obtained mainly from Cronin (1987), Ellis and
tude and temperature were tested with Kendall's tau (τ) correlation Messina (1940–2019), Hazel (1967), Hazel (1983), Moore (1961),
coefficient (Abdi, 2007). Different sizes of latitudinal moving win- Puri (1953), Tabuki and Hanai (1999), Titterton and Whatley
dow (range: 1–10°) were also tested. (1988) and Van Morkhoven (1963). Age-frequency distributions
of geologic ages of ostracod genera were summarized for tropi-
cal (0–23.5°N), subtropical (23.5–35°N), temperate (35–66.5°N)
2.5 | Evaluating beta diversity and Arctic (66.5–90°N) regions, separately, based on species dis-
tributed there regardless of their endemicity (i.e. species were
The metrics of Baselga (2010, 2012) were used to evaluate the included for a region even if their ranges extended into other
relative importance of turnover and nestedness. Sørensen dis- regions). Kolmogorov–Smirnov Tests were applied to test for the
similarity (ßsor) [= the overall ß-diversity per site] was broken down similarities between the age-frequency distributions of geologi-
into turnover (ßsim) and nestedness (ßsne) components. Similarly, cal ages of ostracod genera in the four regions using the function
Jaccard dissimilarity (ßjac) [= the overall ß-diversity of the pooled ks.test in R. The data that support the findings of this study are
assemblage] was broken down into its turnover (ßjtu) and the nest- openly available in Dryad at https​://doi.org/10.5061/dryad.3n5tb​
edness (ßjne) components (Baselga, 2012). Within each 10-degree 2rcp.
moving window along the latitudinal transect, the multiple-site
ß-diversity ßsor and ßjac were calculated by randomly taking five
sites with 100 permutations in order to equalize sampling per 3 | R E S U LT S
latitudinal bin. ß-diversity components, nestedness and turnover,
were calculated with the ‘beta-multi’ function of R package ‘beta- 3.1 | Latitudinal gradient of alpha diversity and its
part’ (Baselga & Orme, 2012). Kendall's tau (τ) correlation coef- controlling factors
ficient (Abdi, 2007) was used to estimate the association between
ß-diversity components, latitude and temperature. BINMATNEST Ostracod diversity of the western North Atlantic showed a noisy
(Rodríguez-Gironés & Santamaría, 2006) and the R package ‘bipar- but significant and standard LDG in sample-level rarefied diversity,
tite’ (Dormann, Gruber, & Fründ, 2008) were used to identify the E(S50) (r2  =  0.02, p  =  .03; n  =  229; Figure 1; Figure S1), with lower
nested rank order of sampling sites with latitude (higher rank num- latitudes showing much higher variation in local richness (see also
bers mean more nestedness in the assemblage), so as to identify Section 3.2 and Figure 2a, which show a significant and standard
the direction and significance of latitude-related trends in nest- LDG in alpha diversity calculated using the multiplicative approach).
edness. Presence–absence data of each species were used for all Multiple regression models predicting alpha diversity from environ-
ß-diversity analyses, and this method has been commonly used in mental parameters find noisy relationships. Model coefficients are
species compositional studies (Stuart et al., 2017). The same multi- generally consistent across the considered models (Table S1), but
ple-site analyses on faunal dissimilarities were performed on genus only bottom water temperature was significant according to the
CHIU et al. |
      821

F I G U R E 2   Ostracod species diversity (a) (b)

patterns with latitude and temperature.
Top panels show (a) latitudinal and (b) τ = 0.68
90
τ = − 0.79
90 P < 0.001
bottom water temperature gradients
P < 0.001
of alpha, beta and gamma diversity

Diversity

Diversity
computed as the ‘true’ diversities for 60 Gamma
τ = − 0.72 60 τ = 0.62
10-degree latitudinal moving window. P < 0.001 P < 0.001 Beta
Kendall's tau (τ) correlation coefficient Alpha
and p-values are shown. Middle panels 30 30
show the relationships between multiple- τ = − 0.5
P < 0.001 τ = 0.52
site ß-diversity (turnover, nestedness P < 0.001
0 0
and overall dissimilarity) along latitudinal
20 40 60 80 0 10 20
gradient with 10-degree moving window
and (c) latitude and (d) bottom water Latitude(°N) Temperature (°C)
temperature. Results are based on Jaccard
(c) 1.00 (d) τ = 0.66
index. Bottom panels show nestedness τ = − 0.67 1.00 P < 0.001
rank order for ostracod assemblages P < 0.001
calculated by BINMATNEST (Rodríguez- 0.75 τ = − 0.65 0.75
Beta Diversity

Beta Diversity
Gironés & Santamaría, 2006) as a function P < 0.001 τ = 0.64
of latitude at (e) species and (f) genus P < 0.001 Turnover
levels. Nestedness is a positive function 0.50
0.50 Nestedness
of latitude. Kendall's tau (τ) correlation
Dissimilarity
coefficient and significance (P) are τ = 0.51
0.25 0.25
shown [Colour figure can be viewed at P < 0.001 τ = − 0.49
wileyonlinelibrary.com] P < 0.001
0.00 0.00
20 40 60 80 0 10 20
Latitude (°N) Temperature (°C)

(e) (f)
200 τ = 0.6 200
P< 0.001
Nestedness Rank

Nestedness Rank

150 150

100 100

50 50

τ = 0.31
0 0 P< 0.001
20 40 60 80 20 40 60 80
Latitude (°N) Latitude (°N)

model-averaged results. This relationship was weak, however, im- that observed in the sample-level, rarefied data (Figure S1), most likely
plying an increase of only 0.15 units of E(S50) for each degree in- because the multiplicative approach averages the number of species
crease in temperature (Table 1). Moreover, although most of the per site within each latitudinal bin, which smooths some of the sam-
best-fitting models include bottom water temperature, the inter- pling variation in local diversity. Gamma diversity declined more rap-
cept-only model also receives modest, but non-negligble support idly than alpha and beta diversity. Alpha, beta and gamma diversity
(ΔAICc = 2.29, model #7 in Table S1). correspondingly increased with temperature (alpha, beta, gamma:
p < .001; Figure 2b). The pattern is consistent across different sizes
of latitudinal bins (range: 1–10°; Figure S2a). Similar trends were also
3.2 | Diversity partitioning found at the genus level (Figures S3a,b, S4a).

Alpha, beta and gamma diversity calculated from the multiplicative

approach all decreased significantly with latitude (alpha, beta, gamma: 3.3 | Beta diversity
p < .001; Figure 2a), which is not surprising because alpha, beta, and
gamma diversities are not independent (Ulrich et al., 2018). The LDG The nestedness component of beta diversity increased with latitude
in alpha diversity under this approach (Figure 2a) is rather neater than (p < .001; Figure 2c) and correspondingly decreased with temperature
 |
822       CHIU et al.

TA B L E 1   Model-averaged parameter estimates and CIs of with the other regions, especially with the subtropical region
shallow-marine ostracod alpha diversity E(S50) in the western North (p = .052; Table S2).
Atlantic

Relative Lower Upper

importance Coefficient CI CI 4 | D I S CU S S I O N
Bottom water 0.76 0.16 0.04 0.37
temperature Our results revealed a noisy but statistically significant latitudinal
(BWT)
gradient in alpha diversity in the northwestern Atlantic (Figures 1, 2a;
Net primary 0.37 −0.65 −5.22 1.71 Figure S1). Our regression results indicated that temperature is the
productivity
most important predictor of alpha diversity (Tables 1; Table S1).
(PP)
Although the temperature–diversity relationship was weak, these
Bottom 0.31 1.41 −7.26 16.22
dissolved findings are consistent with the increasing consensus that tempera-
oxygen (BSO) ture is the most important environmental driver of marine species
Bottom salinity 0.28 −0.04 −1.16 0.88 diversity (Beaugrand, Rombouts, & Kirby, 2013; Tittensor et al.,
(BS) 2010). The strong correlation between temperature and latitude
Distance from 0.62 −0.98 −3.38 0.20 (Figure S8) in our dataset, however, means that it is hard to rule out
land (DST) a role for additional, latitude-related factors beyond temperature
Bottom pH 0.30 −1.43 −19.57 10.05 in structuring this LDG (Snelgrove et al., 2018). The noisiness of
(PH)
our result (Figure S1) is consistent with meta-analyses that showed
Note: Lower CI and Upper CI indicate bounds of 95% confidence latitudinal diversity gradients are weaker in smaller organisms
intervals, accounting for model selection uncertainty using model- (Hillebrand, 2004; Hillebrand & Azovsky, 2001). We found that the
averaging. Relative importance is the sum of Akaike weights of models
latitudinal diversity gradient and temperature–diversity relation-
which include the variable in question. Bold: CIs that exclude zero.
ship is much stronger at gamma (i.e. regional) level (Figure 2a,b; the
same result is found at the genus level, see Figure S3a,b), a finding
(p < .001; Figure 2d). However, beta diversity's turnover component that is also known from previous meta-analyses (Hillebrand, 2004).
and overall dissimilarity decreased with latitude (p < .001; Figure 2c) Beta diversity is higher in lower latitudes (Figure 2a), likely because
and increased with temperature (p  <  .001; Figure 2d). The pattern of higher habitat diversity in the tropics. This is also probably the
was generally consistent across different sizes of latitudinal bins cause of the much larger difference between gamma (i.e. regional
(range: 1–10°; Figure S2b). Similar trends were also found at the species pool) and alpha (local) diversity in the tropics compared to
genus level (Figure S3c,d, S4b). that in higher latitudes (Schlacher et al., 1998; Witman, Etter, &
The rank order of nestedness among samples increased signifi- Smith, 2004).
cantly with latitude (p < .001; Figure 2e), with a similar pattern ob- In our dataset, the turnover component of beta diversity was al-
served at the genus level (p < .001; Figure 2f). ways more dominant than the nestedness component (Figure 2c,d;
Figure S6; the same result was found at the genus level, see Figure
S3c,d). Strong turnover is also evident from the plot of species
3.4 | Range sizes and per-genus species richness ranges (Figure 5). The importance of turnover throughout the en-
tire latitudinal range probably reflects climatic niche conservatism.
Range sizes of species and genera smaller in the tropics and larger Most of the species that occur in the tropics (red in Figure 5) are
in the extratropics, with a sharp change at c. 40°N (Figure 3a). The restricted to the tropics, and species with subtropical occurrences
per-genus species richness showed a unimodal pattern with a peak (red and green in Figure 5) seldom extend their distribution north
at mid-latitudes (20–40°N) and lower values at low (0–20°N) and of 40°N. A considerable number of species never extend their dis-
high (40–80°N) latitudes (Figure 3d). tributions south into the subtropics (blue in Figure 5) and, among
them, there are many ‘subarctic species’ with distributions limited
to >50°N (Figure 5). A smaller number of species are found only in
3.5 | Geologic ages of ostracod genera the Arctic (purple in Figure 5). The preponderance of species with
limited latitudinal ranges (Figure 5) is consistent with the fact that
Geologic age data were available for 70 tropical, 68 subtropical, many ostracod species have narrow temperature ranges in their dis-
57 temperate and 31 Arctic genera. Median ages for the tropi- tribution (Cronin & Ikeya, 1990). Thus, climatic niche conservatism is
cal, subtropical, temperate and Arctic regions were 56, 56, 56 probably an important mechanism for the latitudinal diversity gradi-
and 23 Ma respectively (Figure 4). Kolmogorov–Smirnov pairwise ent of ostracods. Nestedness has a weaker relationship with latitude
tests showed that the age-frequency distributions of tropics, sub- than turnover does, but nestedness still significantly increases as lat-
tropics and temperate regions are similar (p  >  .9; KS-value  <  0.1; itude increases (Figure 2c; Figure S4). The positive nestedness-lati-
Table S2), whereas the Arctic shows relatively larger differences tude relationship in this study indicates that OTT dynamics may play
CHIU et al. |
      823

F I G U R E 3   Range size and species

richness metrics of ostracods and bivalves.
(a) (b)
(a) Ostracod median species (purple) and
genus (yellow) latitudinal range sizes. (b)
Bivalve median species (purple) and genus
(yellow) latitudinal range sizes. (c) Species
richness and (d) mean per-genus species
richness of ostracods (purple) and bivalves
(yellow). All metrics are computed for
1-degree latitudinal bins [Colour figure
can be viewed at wileyonlinelibrary.com]

(°N)

(c) (d)

(°N)

a role, although it is much weaker than the effect of climatic niche There is interest in comparing our results with those of bivalves,
conservatism. which is a well-studied macrofaunal model system with good fos-
Looking further into details, nestedness tends to be highest in sil records comparable to ostracods. The most significant contrast
the mid-latitudes (Figure 2c and e). This trend suggests that not between bivalve and ostracod LDGs is that the LDG in beta diver-
only the OTT but also an ‘Out of the Arctic’ (OTA) process may be sity is more dominated by turnover in ostracods (Figure 2c,d). This
at work. Fossil observations indicate that Arctic species gradually pattern is consistent with ostracod distributions being more con-
extended their distribution southward during the Plio-Pleistocene trolled by climatic niche conservatism, whereas that of bivalves is
glacial-interglacial cycles (Cronin & Ikeya, 1990). By tracking their comparatively more dominated by nestedness and the OTT process
temperature niche, Arctic taxa extended their distribution south- (Tomašových et al., 2016). Indeed, by comparing Tomašových et al.
ward during glacial periods, and their distribution shrank back to the (2016) and our results (Figure 2; Figure S3), we see that the contri-
north during subsequent interglacials. During this back-and-forth bution of nestedness component to the total beta diversity is higher
sloshing of ranges, it is posited that some peripheral populations in bivalves (c. 25% at species level, c. 45% at genus level) than it is
became isolated in the south. Repeating this process through the in ostracods (c. 15% at species and genus levels). This difference is
Plio-Pleistocene glacial-interglacial cycles provided Arctic taxa with also obvious by comparing the fossil data of ostracods with that of
opportunities to adapt and spread to the lower latitudes, just as the bivalves. In bivalves, genera occurring in the present-day Arctic are
opposite occurred for tropical taxa to extend their distribution out geologically older than genera occurring in the present-day tropics
of the tropics (Cronin & Ikeya, 1990). When we divide the latitudi- (Jablonski et al., 2013), reflecting the OTT process in which genera
nal range into two at the peak of nestedness (c. 45°N: Figure 2c), a originate in the tropics and gradually extend their distributions to
positive nestedness-latitude relationship is observed in the lower lat- higher latitudes. In contrast, ostracod genera occurring in the pres-
itudinal range (nestedness ßjne and ßsne for latitude < 45°N: p < .001) ent-day Arctic have very young geological ages (Figure 4), suggesting
and a negative relationship is present in the higher latitudinal range that the majority of them evolved in situ. In fact, the geological ages
(nestedness ßjne and ßsne for latitude ≥ 45°N: p < .001). The overall of ostracod genera occurring in the present-day Arctic are much
nestedness-latitude relationship is positive and thus the OTT process younger than those occurring in the present-day lower-latitude re-
is predominant, but also some OTA dynamic is supported by our data. gions (Figure 4, Table S2), consistent with the OTA dynamic in which
824       | CHIU et al.

(a) F I G U R E 4   Age-frequency distributions

Tropics (N = 70)
20
of geologic ages of ostracod genera
Number of genera
Median = 56 myr currently living in (a) tropics, (b)
15
subtropics, (c) temperate and (d) Arctic
10

areas
5
0

0 100 200 300 400 500

(b)
15

Subtropics (N = 68)
Number of genera

Median = 56 myr
10
5
0

0 100 200 300 400 500

(c)
Temperate (N = 57)
20
Number of genera

Median = 56 myr
15
10
5
0

0 100 200 300 400 500

(d)
15

Arctic (N = 31)
Number of genera

Median = 23myr
10
5
0

0 100 200 300 400 500

Age (myr)

at least some genera originate in the Arctic and spread to the lower Bridge taxa—those that exist in both tropics and extratropics
latitudes. These young geological ages of extant Arctic ostracods are or in any two adjacent latitudinal zones—are important for the OTT
consistent with the recent discovery that the Arctic ostracod fauna, process and play a significant role in the bivalve LDG (Jablonski
although quite distinctive today, was more similar to that of the tem- et al., 2013). In ostracods, there are relatively few species that cross
perate North Atlantic until the mid-Pleistocene Transition (Cronin from the tropics to the subtropics (33 out of 192) and very few of
et al., 2014; DeNinno, Cronin, Rodriguez-Lazaro, & Brenner, 2015). A them make it all the way to the temperate zone (eight out of 192;
similar trend is also known from a molecular study of fishes (Rabosky Figure 4). Moreover, ostracod genera with tropical to Arctic ranges
et al., 2018) which showed that speciation rates were highest at high (Figures S7 and S8, Table S4) often have troughs in diversity and
latitudes, especially in the Arctic Ocean, at least since the Miocene/ occurrence at intermediate latitudes. For example, Hemicytherura
Pliocene boundary (5 Ma). For some taxa such as fishes and ostra- and Propontocypris are rarely found in mid-latitudes (c. 20–40°N for
cods, the endemic Arctic fauna may be quite young and have rapidly Hemicytherura; c. 45–65°N for Propontocypris), Xestoleberis is abun-
developed after the substantial cooling of the Northern hemisphere dant both in the tropical and Arctic latitudes but not in mid-latitudes,
from the initiation of Northern Hemisphere Glaciation (c. 3.2 Ma) or and Cytheropteron is more abundant and also diverse at higher lati-
the Mid-Pleistocene Transition (Zachos, Pagani, & Sloan, 2001). This tudes (Alvarez Zarikian, Stepanova, & Grützner, 2009). These results
timing is consistent with the above-mentioned hypothesis that Plio- suggest that, for latitudinally widespread ostracod genera, there are
Pleistocene glacial-interglacial cycles promoted the OTA process. separate species groups in the tropical and Arctic regions. Thus, we
CHIU et al. |
      825

80

60

First spp.
occurrence
Latitude (°N)

Tropics
Subtropics
40
Temperate
Arctic

20

0 100 200 300 400

Species Rank

F I G U R E 5   Latitudinal ranges for all 426 shallow-marine ostracod species (Appendix S1) in the western North Atlantic and the Arctic
covered in this study. The ostracod species are classified by their lowest-latitude occurrence, i.e. tropical (0–23.5°N), subtropical (23.5–
35°N), temperate (35–66.5°N) and Arctic species (66.5–90°N). The range for each species is computed by connecting the minimum and
maximum latitudes of the species occurrence. Some spp. taxa are kept because of their high abundance or taxonomic uncertainty (c. 2–3
spp., less than 10 individuals), including Argilloecia spp., Bensonocythere spp., Bythoceratina spp., Bythocypris spp., Bythocythere spp., Caudites
spp., Cushmanidea spp., Cytherois spp., Cytheromorpha spp., Cytheropteron spp., Cytherura spp., Eucytherura spp., Hermanites spp., Hulingsina
spp., Krithe spp., Leptocythere spp., Loxoconcha spp., Macrocypris spp., Microcytherura spp., Neonesidea spp., Orionina spp., Paracypris spp.,
Paracytherois spp., Paradoxostoma spp., Perissocytheridea spp., Phlyctocythere spp., Pontocythere spp., Sclerochilus spp., Semicytherura spp. and
Xestoleberis spp. Within each climatic/latitudinal division, the species are ranked by the range width (maximum latitude–minimum latitude)
[Colour figure can be viewed at wileyonlinelibrary.com]

see less evidence for the importance of bridge taxa in ostracods than sensitivity and narrow temperature ranges, and facilitates the assem-
in bivalves. Although the role of polarward bridge taxa in high lati- bly of more localized communities that can be closely related to local
tudes (e.g. from the temperate zone to the Arctic) is not well under- environmental factors including temperature. Direct development
stood, they may not have survived severe Plio-Pleistocene glaciations and limited dispersal characterize not only ostracods, but also other
and may be regionally extinct. This ostracod-bivalve difference in meiofaunal groups such as copepods and nematodes (e.g. Baguley,
bridge taxa is also supported by the per-genus species richness data Montagna, Lee, Hyde, & Rowe, 2006; Brandt et al., 2007; Palmer,
(Figure 3d). Bridge genera tend to have higher speciation rates, as the 1988), and the temperature sensitivity of meiofauna is well-known
per-genus species richness is higher in higher latitudes where bridge not only in ostracods (e.g. Cronin & Ikeya, 1990; Yasuhara & Cronin,
genera dominate in bivalves representing the OTT (Tomašových et 2008), but also in other taxa (Danovaro, Dell'Anno, & Pusceddu,
al., 2016; Figure 3b). In ostracods, the per-genus species richness is 2004; Hunt, Cronin, & Roy, 2005; Yasuhara & Danovaro, 2016). Thus,
highest in the mid latitude and low in the tropics and Arctic. This the contrasts we see between patterns in ostracods and bivalves
unimodal latitudinal pattern (Figure 3d) reflects bridge genera accu- may well hold generality between meiofauna and macrofauna.
mulation in the mid latitude and further supports the hypothesis that
both the OTT and OTA processes have occurred in ostracods.
The differences in mechanisms behind the LDGs between bi- 5 | CO N C LU S I O N S
valves and ostracods may relate to their dispersal abilities and sizes.
Bivalves have a planktonic larva stage whereas ostracods develop 1. The pioneering meta-analysis of Hillebrand (2004) revealed
directly and most marine benthic ostracods cannot swim. Dispersal not only the generality of LDGs but also differences of slope
is important for the OTT model and, thus, the OTT process could be and strength in LDG among taxonomic groups with different
less significant in poorly dispersing taxa such as ostracods (Figure 3a). body sizes and life history traits. He also showed that the
Indeed, species range size is much narrower in ostracods than in bi- gamma (regional) LDGs are much steeper and stronger than
valves, especially in latitudes lower than 45°N (Figure 3a). The lim- the alpha (local) LDGs. Our ostracod results are consistent with
ited dispersal ability of ostracods may also contribute to their climatic this global meta-analysis of the latitudinal diversity gradients.
 |
826       CHIU et al.

This consistency indicates that ostracods are not a contrarian Experimental Marine Biology and Ecology, 330, 327–341. https​://doi.
org/10.1016/j.jembe.2005.12.037
clade (Krug et al., 2007; Yasuhara, Iwatani, et al., 2017) and it
Baranova, O. (2015). World Ocean Atlas, 2013, v2.
supports ostracods as a useful system to explore an integrated Barton, K. (2017). MuMIn: Multi-model inference. R package ver-
neontological-palaeontological analysis of latitudinal diversity sion, 1.40.0. Retrieved from https​://CRAN.R-proje​c t.org/packa​
gradients. ge=MuMIn​
Baselga, A. (2010). Partitioning the turnover and nestedness compo-
2. Our detailed analyses of meiobenthic ostracod data and their
nents of beta diversity. Global Ecology and Biogeography, 19, 134–143.
comparison with macrobenthic bivalve results revealed that small https​://doi.org/10.1111/j.1466-8238.2009.00490.x
organisms, which make up two thirds of marine diversity, do not Baselga, A. (2012). The relationship between species replace-
always follow the same rules as better-studied, large organisms. ment, dissimilarity derived from nestedness, and nestedness.
Small organisms, which are understudied but make up the vast Global Ecology and Biogeography, 21, 1223–1232. https​://doi.
org/10.1111/j.1466-8238.2011.00756.x
majority of biodiversity, may be much more sensitive to climate
Baselga, A. (2017). Partitioning abundance-based multiple-site dissimi-
than the well-studied large organisms. This implies that the im- larity into components: Balanced variation in abundance and abun-
pacts of ongoing Anthropocene climatic change on marine eco- dance gradients. Methods in Ecology and Evolution, 8, 799–808. https​
systems may be much more serious than previously understood. ://doi.org/10.1111/2041-210X.12693​
Baselga, A., & Orme, C. D. L. (2012). betapart: An R package for the study
of beta diversity. Methods in Ecology and Evolution, 3, 808–812.
AC K N OW L E D G E M E N T S Beaugrand, G., Rombouts, I., & Kirby, R. R. (2013). Towards an under-
We thank Laura Wong, Maria Lo and Julia Seidenstein for technical standing of the pattern of biodiversity in the oceans. Global Ecology
support; Zeenatul Basher for support in GIS; Allen Suen for statistical and Biogeography, 22, 440–449. https​://doi.org/10.1111/geb.12009​
Behrenfeld, M. J., & Falkowski, P. G. (1997). Photosynthetic rates de-
consultation; Harry Dowsett, Lynn Wingard, Debra Willard, Michael
rived from satellite-based chlorophyll concentration. Limnology and
Toomey, Christopher Swezey and Robert Stamm for discussion and Oceanography, 42, 1–20. https​://doi.org/10.4319/lo.1997.42.1.0001
improving the manuscript; and Werner Ulrich, Matthew Powell Brandt, A., Gooday, A. J., Brandão, S. N., Brix, S., Brökeland, W.,
and an anonymous reviewer for valuable comments. The work de- Cedhagen, T., … Vanreusel, A. (2007). First insights into the biodiver-
sity and biogeography of the Southern Ocean deep sea. Nature, 447,
scribed in this article was partially supported by grants from the
307–311. https​://doi.org/10.1038/natur​e 05827
Research Grants Council of the Hong Kong Special Administrative Brault, S., Stuart, C. T., Wagstaff, M. C., & Rex, M. A. (2013). Geographic
Region, China (project codes: HKU 17302518, HKU 17306014, HKU evidence for source–sink dynamics in deep-sea neogastropods of
17311316) and the Seed Funding Programme for Basic Research the eastern North Atlantic: An approach using nested analysis. Global
Ecology and Biogeography, 22, 433–439. https​://doi.org/10.1111/
of the University of Hong Kong (project codes: 201311159076,
geb.12005​
201611159053) (to MY), and by the HKU Earth as a Habitable Planet Burnham, K. P., & Anderson, D. R. (2002) Information and likelihood the-
Thesis Development Grant (to RWTC). Any use of trade, firm or ory: A basis for model selection and inference. In: K. P. Burnham,
product names is for descriptive purposes only and does not imply & D. R. Anderson (Eds.), Model selection and multimodel inference: A
endorsement by the U.S. Government. practical information-theoretic approach (pp. 49–97). New York, NY:
Springer.
Cronin, T. M. (1983). Bathyal ostracodes from the Florida-Hatteras slope,
DATA AVA I L A B I L I T Y S TAT E M E N T the Straits of Florida, and the Blake Plateau. Marine Micropaleontology,
The data that support the findings of this study are openly available 8, 89–119. https​://doi.org/10.1016/0377-8398(83)90007-5
in Dryad at https​://doi.org/10.5061/dryad.3n5tb​2rcp. Cronin, T. M. (1987). Evolution, biogeography, and systematics of Puriana:
Evolution and speciation in Ostracoda, III. Journal of Paleontology, 61,
1–71.
ORCID Cronin, T. M. (1990). Evolution of Neogene and Quaternary Marine
Wing-Tung Ruby Chiu  https://orcid.org/0000-0001-6663-9526 Ostracoda, United States Atlantic Coastal Plain: evolution and spe-
Moriaki Yasuhara  https://orcid.org/0000-0003-0990-1764 ciation in Ostracoda, IV. U.S. Geological Survey Professional Paper
1367-C. In Studies related to the Charleston, South Carolina, earthquake
of 1886; Neogene and Quaternary lithostratigraphy and biostratigraphy
REFERENCES (pp. C1–C43). Washington, D.C.: United States Government Printing
Abdi, H. (2007). The Kendall rank correlation coefficient (pp. 508–510). Office.
Sage, Thousand Oaks, CA: Encyclopedia of Measurement and Cronin, T. M., DeNinno, L. H., Polyak, L., Caverly, E. K., Poore, R. Z.,
Statistics. Brenner, A., … Marzen, R. E. (2014). Quaternary ostracode and for-
Alvarez Zarikian, C. A., Stepanova, A. Y., & Grützner, J. (2009). Glacial–in- aminiferal biostratigraphy and paleoceanography in the western
terglacial variability in deep sea ostracod assemblage composition at Arctic Ocean. Marine Micropaleontology, 111, 118–133. https​://doi.
IODP Site U1314 in the subpolar North Atlantic. Marine Geology, 258, org/10.1016/j.marmi​cro.2014.05.001
69–87. https​://doi.org/10.1016/j.margeo.2008.11.009 Cronin, T. M., & Ikeya, N. (1990). Tectonic events and climatic change:
Anderson, D. R., Burnham, K. P., & Thompson, W. L. (2000). Null Opportunities for speciation in Cenozoic marine Ostracoda. In
hypothesis testing: Problems, prevalence, and an alternative. R. M. Ross & W. D. Allmon (Eds.), Causes of Evolution–A Paleontological
The Journal of Wildlife Management, 64, 912–923. https​://doi. Perspective (pp. 210–248). Chicago and London: University of
org/10.2307/3803199 Chicago Press.
Baguley, J. G., Montagna, P. A., Lee, W., Hyde, L. J., & Rowe, G. T. (2006). Danovaro, R., Dell'Anno, A., & Pusceddu, A. (2004). Biodiversity re-
Spatial and bathymetric trends in Harpacticoida (Copepoda) commu- sponse to climate change in a warm deep sea. Ecology Letters, 7, 821–
nity structure in the Northern Gulf of Mexico deep-sea. Journal of 828. https​://doi.org/10.1111/j.1461-0248.2004.00634.x
CHIU et al. |
      827

DeNinno, L. H., Cronin, T. M., Rodriguez-Lazaro, J., & Brenner, A. (2015). Jost, L. (2007). Partitioning diversity into independent alpha and beta
An early to mid-Pleistocene deep Arctic Ocean ostracode fauna components. Ecology, 88, 2427–2439. https​://doi.org/10.1890/06-
with North Atlantic affinities. Palaeogeography, Palaeoclimatology, 1736.1
Palaeoecology, 419, 90–99. https​://doi.org/10.1016/j. Jost, L. (2010). Independence of alpha and beta diversities. Ecology, 91,
palaeo.2014.07.026 1969–1974. https​://doi.org/10.1890/09-0368.1
Dormann, C. F., Gruber, B., & Fründ, J. (2008). Introducing the bipartite Jurasinski, G., & Retzer, V. (2012). simba: A collection of functions for
package: Analysing ecological networks. R News, 8, 8–11. similarity analysis of vegetation data. R package version 0.3–5.
Ellis, B., & Messina, A. (1940–2019). Catalogues of micropalaeontology: Kontrovitz, M. (1976). Ostracoda from the Louisiana continental shelf.
Foraminifers. Ostracoda, Diatoms, New York, Paleontology Press. Tulane Studies in Geology and Paleontology, 12, 49–100.
Fine, P. V. A. (2015). Ecological and evolutionary drivers of geographic Krug, A. Z., Jablonski, D., & Valentine, J. W. (2007). Contrarian clade con-
variation in species diversity. Annual Review of Ecology, Evolution, and firms the ubiquity of spatial origination patterns in the production of
Systematics, 46, 369–392. https​://doi.org/10.1146/annur​ev-ecols​ latitudinal diversity gradients. Proceedings of the National Academy of
ys-112414-054102 Sciences of the United States of America, 104, 18129–18134. https​://
Gemery, L., Cronin, T. M., Briggs, W. M., Brouwers, E. M., Schornikov, E. doi.org/10.1073/pnas.07092​02104​
I., Stepanova, A., … Yasuhara, M. (2015). Arctic Ostracode Database Leray, M., & Knowlton, N. (2015). DNA barcoding and metabarcoding
2015. IGBP PAGES/World Data Center-A for Paleoclimatology Data of standardized samples reveal patterns of marine benthic diversity.
Contribution Series. NOAA/NGDC Paleoclimatology Program, Proceedings of the National Academy of Sciences, 112, 2076–2081.
Boulder CO. https​://doi.org/10.1073/pnas.14249​97112​
Hazel, J. E. (1967). Classification and distribution of the Recent Lyon, S. (1990). Biostratigraphic and Paleoenvironmental Interpretations
Hemicytheridae and Trachyleberididae (Ostracoda) off northeastern from Late Pleistocene Ostracoda, Charleston, South Carolina.
North America. U. S. Geological Survey Professional Paper 564. U.S. Geological Survey Professional Paper 1367-D. In Studies re-
Hazel, J. E. (1970). Atlantic Continental Shelf and Slope of the United States lated to the Charleston, South Carolina, earthquake of 1886; Neogene
- Ostracode zoogeography in the southern Nova Scotian and northern and Quaternary lithostratigraphy and biostratigraphy (D1–D48).
Virginian faunal provinces. U.S. Geological Survey Professional Paper Washington, D.C.: United States Government Printing Office.
529E. McClain, C. R., & Schlacher, T. A. (2015). On some hypotheses of diver-
Hazel, J. E. (1983). Age and correlation of the Yorktown (Pliocene) and sity of animal life at great depths on the sea floor. Marine Ecology, 36,
Croatan (Pliocene and Pleistocene) formations at the Lee Creek 849–872. https​://doi.org/10.1111/maec.12288​
Mine. Smithsonian Contributions to Paleobiology, 53, 81–199. Mesquita-Joanes, F., Smith, A. J., & Viehberg, F. A. (2012). Chapter 2 -
Hillebrand, H. (2004). On the generality of the latitudinal diver- The ecology of ostracoda across levels of biological organisation
sity gradient. The American Naturalist, 163, 192–211. https​://doi. from individual to ecosystem: A review of recent developments and
org/10.1086/381004 future potential. In D. J. Horne, J. A. Holmes, J. Rodriguez-Lazaro,
Hillebrand, H., & Azovsky, A. I. (2001). Body size determines the strength & F. A. Viehberg (Eds.), Ostracoda as proxies for Quaternary climate
of the latitudinal diversity gradient. Ecography, 24, 251–256. https​:// change (pp. 15–35). Amsterdam: Elsevier
doi.org/10.1034/j.1600-0587.2001.240302.x Moore, R. C. (1961). Treatise on invertebrate paleontology, part Q,
Holmes, J. A., & Chivas A. R. (2002). The Ostracoda: Applications in Arthropoda 3 (p. 442). Lawrence: The University of Kansas.
Quaternary Research Washington, D.C.: American Geophysical Union. Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., & O'Hara, R. (2016).
Horne, D. J., Cohen, A., & Martens, K. (2002). Taxonomy, morphology vegan: Community ecology package. R package version, 2.4–4.
and biology of quaternary and living Ostracoda. In J. A. Holmes & Olsen, A., Key, R. M., van Heuven, S., Lauvset, S. K., Velo, A., Lin, X., …
A. R. Chivas (Eds.), The Ostracoda: Applications in Quaternary Research Jutterström, S. (2016). The global ocean data analysis project version
(pp. 5–36). Washington, D.C.: American Geophysical Union. 2 (GLODAPv2) – an internally consistent data product for the world
Huang, H.-H.- M., Yasuhara, M., Iwatani, H., Alvarez Zarikian, C. A., ocean. Earth System Science Data, 8, 297–323.
Bassetti, M. A., & Sagawa, T. (2018). Benthic biotic response to climate Palmer, M. A. (1988). Dispersal of marine meiofauna: A review and con-
changes over the last 700,000 years in a deep marginal sea: Impacts ceptual model explaining passive transport and active emergence
of deoxygenation and the mid-brunhes event. Paleoceanography and with implications for recruitment. Marine Ecology Progress Series, 48,
Paleoclimatology, 33, 766–777. 81–91.
Hunt, G., Cronin, T. M., & Roy, K. (2005). Species–energy re- Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., & Team, R. C. (2014). nlme:
lationship in the deep sea: A test using the Quaternary Linear and nonlinear mixed effects models.R package version,
fossil record. Ecology Letters, 8, 739–747. https​://doi. 3.1–131.
org/10.1111/j.1461-0248.2005.00778.x Puri, H. S. (1953). The ostracode genus trachyleberis and its ally actino-
Hurvich, C. M., & Tsai, C.-L. (1989). Regression and time series model cythereis. The American Midland Naturalist, 49, 171–187. https​://doi.
selection in small samples. Biometrika, 76, 297–307. https​://doi. org/10.2307/2422285
org/10.1093/biome​t/76.2.297 QGIS Development Team. (2016). QGIS geographic information sys-
Jablonski, D., Belanger, C. L., Berke, S. K., Huang, S., Krug, A. Z., Roy, K., tem.Open Source Geospatial Foundation Project. Retrieved from
… Valentine, J. W. (2013). Out of the tropics, but how? Fossils, bridge http://www.qgis.org/
species, and thermal ranges in the dynamics of the marine latitudinal R Development Core Team. (2016) R: A language and environment for sta-
diversity gradient. Proceedings of the National Academy of Sciences, tistical computing. R Foundation for Statistical Computing, Vienna,
110, 10487–10494. https​://doi.org/10.1073/pnas.13089​97110​ Austria. Vienna: R Foundation for Statistical Computing.
Jablonski, D., Huang, S., Roy, K., & Valentine, J. W. (2017). Shaping the Rabosky, D. L., Chang, J., Title, P. O., Cowman, P. F., Sallan, L., Friedman,
latitudinal diversity gradient: New perspectives from a synthesis of M., … Alfaro, M. E. (2018). An inverse latitudinal gradient in spe-
paleobiology and biogeography. The American Naturalist, 189, 1–12. ciation rate for marine fishes. Nature, 559, 392–395. https​://doi.
https​://doi.org/10.1086/689739 org/10.1038/s41586-018-0273-1
Jablonski, D., Roy, K., & Valentine, J. W. (2006). Out of the tropics: Rex, M. A., & Etter, R. J. (2010). Deep-sea biodiversity: Pattern and scale.
Evolutionary dynamics of the latitudinal diversity gradient. Science, Cambridge and London: Harvard University Press.
314, 102–106. https​://doi.org/10.1126/scien​ce.1130880 Rodríguez-Gironés, M. A., & Santamaría, L. (2006). A new algorithm
Jost, L. (2006). Entropy and diversity. Oikos, 113, 363–375. to calculate the nestedness temperature of presence–absence
 |
828       CHIU et al.

matrices. Journal of Biogeography, 33, 924–935. https​://doi. of the United States of America, 101, 15664–15669. https​://doi.
org/10.1111/j.1365-2699.2006.01444.x org/10.1073/pnas.04043​0 0101​
Rodriguez-Lazaro, J., & Ruiz-Muñoz, F. (2012). Chapter 1 - A general in- Worm, B., & Tittensor, D. P. (2018). A theory of global biodiversity.
troduction to ostracods: morphology, distribution, fossil record and Princeton: Princeton University Press.
applications. In D. J. Horne, J. A. Holmes, J. Rodriguez-Lazaro, & Yasuhara, M., & Cronin, T. M. (2008). Climatic influences on deep-sea os-
F. A. Viehberg (Eds.), Ostracoda as proxies for Quaternary climate tracode (Crustacea) diversity for the last three million years. Ecology,
change (pp. 1–14). Amsterdam: Elsevier. 89, S53–S65. https​://doi.org/10.1890/07-1021.1
Roy, K., Jablonski, D., Valentine, J. W., & Rosenberg, G. (1998). Marine lat- Yasuhara, M., & Danovaro, R. (2016). Temperature impacts on deep-
itudinal diversity gradients: Tests of causal hypotheses. Proceedings sea biodiversity. Biological Reviews, 91, 275–287. https​://doi.
of the National Academy of Sciences, 95, 3699–3702. https​://doi. org/10.1111/brv.12169​
org/10.1073/pnas.95.7.3699 Yasuhara, M., Hunt, G., van Dijken, G., Arrigo, K. R., Cronin, T. M., &
Schlacher, T. A., Newell, P., Clavier, J., Schlacher-Hoenlinger, M. A., Wollenburg, J. E. (2012). Patterns and controlling factors of species
Chevillon, C., & Britton, J. (1998). Soft-sediment benthic community diversity in the Arctic Ocean. Journal of Biogeography, 39, 2081–2088.
structure in a coral reef lagoon – the prominence of spatial hetero- https​://doi.org/10.1111/j.1365-2699.2012.02758.x
geneity and 'spot endemism'. Marine Ecology Progress Series, 174, Yasuhara, M., Iwatani, H., Hunt, G., Okahashi, H., Kase, T., Hayashi, H.,
159–174. https​://doi.org/10.3354/meps1​74159​ … Renema, W. (2017). Cenozoic dynamics of shallow-marine biodi-
Snelgrove, P. V. R., Soetaert, K., Solan, M., Thrush, S., Wei, C.-L., versity in the Western Pacific. Journal of Biogeography, 44, 567–578.
Danovaro, R., … Volkenborn, N. (2018). Global carbon cycling on a https​://doi.org/10.1111/jbi.12880​
heterogeneous seafloor. Trends in Ecology & Evolution, 33, 96–105. Yasuhara, M., Tittensor, D. P., Hillebrand, H., & Worm, B. (2017).
https​://doi.org/10.1016/j.tree.2017.11.004 Combining marine macroecology and palaeoecology in understand-
Stuart, C. T., Brault, S., Rowe, G. T., Wei, C.-L., Wagstaff, M., McClain, ing biodiversity: Microfossils as a model. Biological Reviews, 92, 199–
C. R., & Rex, M. A. (2017). Nestedness and species replacement 215. https​://doi.org/10.1111/brv.12223​
along bathymetric gradients in the deep sea reflect productivity: Zachos, J., Pagani, M., Sloan, L.Thomas, E., & Billups, K. (2001). Trends,
A test with polychaete assemblages in the oligotrophic north-west rhythms, and aberrations in global climate 65 Ma to present. Science,
Gulf of Mexico. Journal of Biogeography, 44, 548–555. https​://doi. 292, 686–693. https​://doi.org/10.1126/scien​ce.1059412
org/10.1111/jbi.12810​
Tabuki, R., & Hanai, T. (1999). A new sigillid ostracod from submarine
caves of the Ryukyu Islands, Japan. Palaeontology, 42, 569–593. https​
B I O S K E TC H E S
://doi.org/10.1111/1475-4983.00086​
Teeter, J. W. (1975). Distribution of holocene marine ostracoda from be-
Wing-Tung Ruby Chiu is an environmental scientist investigating
lize. American Association of Petroleum Geologist Studies on Geology, marine and wetland micropalaeontology, macroecology and shell
2, 400–499. geochemistry. She is an alumni from Moriaki Yasuhara's lab in the
Tittensor, D. P., Mora, C., Jetz, W., Lotze, H. K., Ricard, D., Berghe, E. V., & University of Hong Kong and was jointly supervised by Moriaki
Worm, B. (2010). Global patterns and predictors of marine biodiver-
Yasuhara (HKU), Thomas M. Cronin (US Geological Survey) and
sity across taxa. Nature, 466, 1098–1101. https​://doi.org/10.1038/
natur​e 09329 Gene Hunt (Smithsonian Institution) for her PhD thesis. She
Titterton, R., & Whatley, R. (1988). Recent bairdiinae (Crustacea, currently works as a Biology and Science teacher at St. Rose of
Ostracoda) from the Solomon Islands. Journal of Micropalaeontology, Lima's College in Hong Kong.
7, 111–142. https​://doi.org/10.1144/jm.7.2.111
Tomašových, A., Kennedy, J. D., Betzner, T. J., Kuehnle, N. B., Edie, S.,
Kim, S., … Jablonski, D. (2016). Unifying latitudinal gradients in range Author contributions: W.T.R.C. and M.Y. designed the study.
size and richness across marine and terrestrial systems. Proceedings of W.T.R.C., T.C. and L.G. performed taxonomic harmonization
the Royal Society B: Biological Sciences, 283. https​://doi.org/10.1098/ and constructed the ostracod dataset. W.T.R.C., C.L.W., G.H.
rspb.2015.3027
and M.Y. analysed the data. W.T.R.C. wrote the first draft of the
Ulrich, W., Kubota, Y., Kusumoto, B., Baselga, A., Tuomisto, H., & Gotelli,
N. J. (2018). Species richness correlates of raw and standardized manuscript. All authors discussed the results and contributed to
co-occurrence metrics. Global Ecology and Biogeography, 27, 395– revisions.
399. https​://doi.org/10.1111/geb.12711​
Valentine, P. C. (1971). Climatic implication of a late Pleistocene ostrac-
S U P P O R T I N G I N FO R M AT I O N
ode assemblage from southeastern Virginia. U.S. Geological Survey
Professional Paper, 683–D. Additional supporting information may be found online in the
Van Morkhoven, F. P. (1963). Post-Palaeozoic Ostracoda: Their morphology, Supporting Information section.         
taxonomy and economic use. Amsterdam: Elsevier.
Whittaker, R. H. (1972). Evolution and measurement of species diversity.
Taxon, 21, 213–251. https​://doi.org/10.2307/1218190 How to cite this article: Chiu W-TR, Yasuhara M, Cronin TM,
Willig, M. R., Kaufman, D. M., & Stevens, R. D. (2003). Latitudinal gradi-
Hunt G, Gemery L, Wei C-L. Marine latitudinal diversity
ents of biodiversity: Pattern, process, scale, and synthesis. Annual
Review of Ecology, Evolution, and Systematics, 34, 273–309. https​:// gradients, niche conservatism, and out of the tropics and
doi.org/10.1146/annur​ev.ecols​ys.34.012103.144032 Arctic: Climatic sensitivity of small organisms. J Biogeogr.
Witman, J. D., Etter, R. J., & Smith, F. (2004). The relationship between 2020;47:817–828. https​://doi.org/10.1111/jbi.13793​
regional and local species diversity in marine benthic communities:
A global perspective. Proceedings of the National Academy of Sciences