Download Linking long-term, large-scale climatic and environmental variability to patterns of

Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Contents lists available at ScienceDirect
Journal of Experimental Marine Biology and Ecology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j e m b e
Linking long-term, large-scale climatic and environmental variability to patterns of
marine invertebrate recruitment: Toward explaining “unexplained” variation
Bruce A. Menge a,⁎, Tarik C. Gouhier a, Tess Freidenburg a,b, Jane Lubchenco a
a
b
Department of Zoology, Oregon State University, Corvallis, OR 97331–2914, United States
MPA Monitoring Enterprise, California Ocean Science Trust, 1330 Broadway, Suite 1135, Oakland, CA 94612, United States
a r t i c l e
i n f o
Keywords:
Barnacles
Climate change
Climatic perturbation
Coastal ecosystem
El Niño
ENSO
La Niña
Long-term time series
Mussels
NPGO
Oregon
PDO
Recruitment
Rocky intertidal
Spatial scale
a b s t r a c t
The rate of input of new individuals into communities is a key component of community dynamics. Detection
of ecosystem responsiveness to climatic perturbations such as El Niño-Southern Oscillation (ENSO), Pacific
Decadal Oscillation (PDO) and North Pacific Gyre Oscillation (NPGO) can provide insight into how ecosystems
will respond to climate change. Although these climate patterns are ocean-based, understanding of their
influences on key ecosystem processes in temperate coastal marine ecosystems is limited. Recent analyses
documented orders-of-magnitude increases in mussel recruitment in the 2000s relative to the previous
decade. To evaluate the generality of this pattern across other intertidal species with planktotrophic
dispersive larvae, and determine the extent to which variation in recruitment was explained by climatic and
environmental (upwelling) variation, we analyzed patterns of barnacle and mussel recruitment across space
(at up to 10 sites across 250 km) and through time (up to 20-years). Compared to mussel recruitment,
barnacle recruitment varied far less interdecadally, and showed minimal change in seasonality. Using three
analytical approaches, multiple regression, quantile regression and wavelet analysis, we found that
recruitment of the barnacles Balanus glandula and Chthamalus dalli and mussels Mytilus spp. varied in
generally similar ways with climate variation as expressed in ENSO, NPGO, and PDO indices, and with
upwelling, as expressed in the Bakun upwelling index. In regression approaches, NPGO and upwelling had the
strongest associations with recruitment of all species, with MEI and PDO also having an influence on C. dalli.
The strength of the association with the environment varied with recruitment magnitude, however, being
generally stronger at intermediate and in some cases high recruitment densities and weak at low recruitment
densities. Time series (wavelet) analysis showed that with all environmental indices, the most consistent
strong relationships occurred at annual periodicities, but multi-year periodicities were also important with
PDO and upwelling indices. Patterns were complex, however, with periods of strong and weak associations at
different time periodicities varying differently among the different environmental measures, and in some
cases varying from strongly positive to strongly negative associations at the same periodicities through time.
These relatively consistent relationships between large-scale climate and upwelling, and recruitment
explained a substantial fraction of the variance (37–40%). The ~ 60% of the variance in recruitment not
explained in our analyses is likely due to myriad other factors such as shorter-scale variation in upwelling,
internal tidal bores, sea breezes, wave forces, larval behavior, and post-settlement processes. Thus, our
analysis adds insights into potential influences on recruitment that occur at ocean basin- to regional scales,
and complements the body of work that focuses on the more local-scale and shorter-term mechanistic
influences that also help determine the pace at which larvae recruit to the adult habitat.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
The potential ecological consequences of changing climate on
marine and coastal ecosystems are likely to be extensive and profound
(Southward et al., 1995; McGowan et al., 1998; Peterson and Schwing,
⁎ Corresponding author.
E-mail address: [email protected] (B.A. Menge).
0022-0981/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jembe.2011.02.003
2003; Feely et al., 2004; Harley et al., 2006; Helmuth et al., 2006;
Helmuth et al., 2011-this issue; Johnson et al., 2011-this issue;
Philippart et al., 2011-this issue; Schiel, 2011-this issue), yet
uncertainties remain about the mechanistic linkages between
physical and biological change. For example, El Niño-Southern
Oscillation (ENSO) cycles can offer valuable insights into the potential
influences of large scale changes in climate (Glynn, 1988; Chavez
et al., 1999), but effects vary with the intensity of the ENSO, perhaps in
a non-linear manner (Holmgren et al., 2001). One issue is the extent
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
to which ENSO characteristics are affected (in duration, intensity, and
frequency) by both anthropogenic climate change and longer-term
climatic cycles (Kleeman and Power, 2000; Mann et al., 2000), varying
over decades rather than the 3–7 year ENSO pattern. The detection of
the Pacific Decadal Oscillation (PDO) in the 1990s suggested that
decadal cycles in abundance of fish stocks of commercial interest are
at least partly due to changes associated with large-scale oceanic
variability (Francis and Hare, 1994; Hare and Francis, 1994; Mantua
et al., 1997). This finding added another layer of complexity to our
efforts to understand the link between climatic pattern and the
structure and functioning of ecological systems. Most recently, a third
climate cycle of intermediate length (the 10–15 year North Pacific
Gyre Oscillation), has also been associated with changes in zooplankton and phytoplankton abundance in coastal environments (Di
Lorenzo et al., 2008; Menge et al., 2009).
Evidence suggests that temperate pelagic ecosystems may be
highly sensitive to the PDO cycle (McGowan et al., 1998; Peterson
and Schwing, 2003). Abundance of zooplankton off southern
California, of northern-origin copepods off the Oregon coast, and
survival of Coho salmon (Peterson and Schwing, 2003) were all low
during a “warm” phase of the PDO (~ 1977–1998). In 1997–1998
several climate shifts occurred more or less simultaneously: one of
the century's two most intense El Niños occurred, followed in 1999
by an intense La Niña. Coincidentally, in 1998, oceanographic
conditions seemed to shift to a “cold” phase of the PDO, although in
2002–2007 the PDO temperature anomalies had shifted back to
warmer temperatures, and then shifted back to cooler temperatures
in 2008–2009 (http://jisao.washington.edu/pdo/PDO.latest). Such
fluctuations lend uncertainty as to whether or not the switch to cool
conditions has yet occurred. Overlaid on all this was a steadily rising
global temperature due to anthropogenic influences (Houghton
et al., 2001). The ecological changes associated with these shifts led
Peterson and Schwing (2003) to suggest that northeast Pacific
ecosystems had entered a “new climate regime.” Other recent
changes in the northeast Pacific (the sudden onset of persistent
summer hypoxia in 2002 (Grantham et al., 2004), the “strange
summer” of 2005 (Brodeur et al., 2006; Kudela et al., 2006; Schwing
et al., 2006; Sydeman et al., 2006; Barth et al., 2007); the unusually
intense upwelling summer of 2006 (Chan et al., 2008)) are consistent
with the possibility of a regime shift in this region.
The supply of new individuals is obviously crucial for population
persistence, and both models and evidence from a variety of systems
suggest that under certain conditions, recruitment can be an
important determinant of community structure (Underwood and
Denley, 1984; Gaines and Roughgarden, 1985; Menge and Sutherland,
1987; Steele, 1997; Tilman, 1997; Connolly and Roughgarden, 1999;
Menge, 2000; Gouhier et al., 2010). One assumption made in some
models is that propagule supply varies in a consistent way along some
environmental gradient, such as upwelling intensity (e.g., Connolly
and Roughgarden, 1999). That is, among-species differences (e.g., in
magnitude or seasonality) are unimportant or “average out,” and
thus patterns of recruitment of different species tend to co-vary, at
least on the scales that are relevant to determination of community
structure.
Here we examine this assumption in the context of climate-related
environmental variation in a rocky intertidal system. Along the US
west coast, dominant space-occupying sessile invertebrates in the
rocky intertidal region include barnacles and mussels (e.g., Schoch
et al., 2006). On the Oregon coast, four of the most common of these
species are the barnacles Balanus glandula and Chthamalus dalli, and
the mussels Mytilus californianus and M. trossulus. These species cooccur, sharing rocky surfaces in the high, mid and low intertidal zones,
and when abundant, can interact strongly (Dayton, 1971; Navarrete,
1996; Berlow, 1997; Berlow, 1999; Menge, 2000, 2003). In these
“post-recruitment” processes, the mussel is the dominant competitor
in mid and low zones (e.g., Paine, 1966, 1974; Dayton, 1971; Menge,
237
2003), and in all zones B. glandula usually outcompetes C. dalli,
although when the latter is abundant relative to the former, it can
reverse the outcome of competition (Menge, 2000). A further
complexity is that B. glandula facilitates successful colonization of
mussels, thereby sealing its doom as the mussels settle among, grow
and eventually smother the barnacles (Berlow, 1997).
When recruitment is included in models of community structure,
it is assumed to be a primary determinant of the strength of
species interactions (Menge and Sutherland, 1987; Connolly and
Roughgarden, 1999). High recruitment is assumed to lead to strong
competition, and in conditions that are favorable for high levels of
predator activity, strong predation. Low recruitment is assumed to
underlie weak interaction strength. Whether or not among-species
differences in recruitment magnitude or variation in temporal
patterns of recruitment produce important alterations in interaction
strength and thereby in community structure has not been well
explored. Recently, we reported that (1) orders-of-magnitude
increases in mussel recruitment occurred in association with the
millennial transition (=1999–2000; Menge et al., 2009), but that,
depending on site and zone, (2) this change led to no change or just
small changes in community structure (Menge et al., 2010b). We
suggested that the minimal response was due in part to the absence of
comparable changes in barnacle recruitment, and thus to a lack of
change in the strength of mussel–barnacle interactions. The factors
underlying the changes or lack thereof in recruitment were, however,
not investigated. In the context of inevitable climate change and
uncertainty in the ecosystem consequences of these changes, it is
increasingly important to gain insight into the factors underlying
change in the input of recruits to ecological communities (see also
recent work in Europe by Poloczanska et al., 2008; Hawkins et al.,
2009).
Fig. 1. Map of study sites (after Menge et al., 2009).
238
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
H3. Relationships between recruitment of these species were altered
by the millennial transition.
Here our focus is on determining the relative contributions to
variation in recruitment of large- to regional-scale environmental
variation, as reflected in climate cycles and upwelling. We examine
barnacle and mussel recruitment time series in detail, and analyze the
variation we observed in relation to variation in large- to meso-scale
variation in climatic and environmental cycles: PDO, ENSO, NPGO (all
large scale), and coastal upwelling (a mesoscale factor that can be
involved in transport of larvae to shore, and thereby recruitment). Our
goals are: (1) to begin to understand the linkages between climaterelated variability and the key ecosystem process of recruitment,
(2) to determine if different species responded to these factors in
similar or different ways, and (3) to examine correlations in recruitment
between these species and how they were affected by the millennial
transition. We evaluate three hypotheses:
2. Methods
2.1. Study sites
Our study was conducted at rocky sites along the Oregon coast:
Cape Meares (CM; 45° 34′ 12″ N, 123° 58′ 12″ W), Fogarty Creek (FC;
44° 50′ 24″ N, 124° 3′ 36″ W), Boiler Bay (BB, 44° 49′ 48″ N, 124° 3′ 36″
W), Seal Rock (SR; 44° 30′ N, 124° 5″ 24″ W), Yachats Beach (YB; 44°
19′ 12″ N, 124° 7′ 12″ W), Strawberry Hill (SH; 44° 15′ N, 124° 7′ 12″
W), Tokatee Klootchman (TK; 44° 12′ N, 124° 7′ 12″ W), Cape Arago
(CA; 43° 18′ 36″ N, 124° 24′ W), Cape Blanco (CB; 42° 50′ 24″ N, 124°
34′ 12″ W), and Rocky Point (RP; 42° 43′ 12″ N, 124° 28′ 12″ W). Fig. 1
shows the locations of these sites along the Oregon coast. Data
collection began in 1989 at BB and SH and through the 1990s began at
additional sites in different years (Menge et al., 2009). Our present
analysis includes data up through 2008. For brevity we refer to 1989–
1999 data as the “1990s” and 2000–2008 data as the “2000s.”
H1. Recruitment of barnacles and mussels is sensitive to climatic
variability as expressed in the ENSO, PDO, NPGO climate indices and
the Bakun upwelling index.
B. glandula recruits * d-1
H2. Recruitment of these species responded similarly between 1989
and 2008 to these sources of variability.
400
300
200
100
0
400
300
200
100
0
400
300
200
100
0
1200
1000
800
600
400
200
0
400
300
200
100
0
400
300
200
100
0
600
500
400
300
200
100
0
200
150
100
50
0
150
Cape Meares
Fogarty Creek
Boiler Bay
Seal Rock
2000's > 1990's
Yachats Beach
2000's > 1990's
Strawberry Hill
Tokatee Klootchman
Cape Arago
Cape Blanco
100
50
0
150
Rocky Point
100
50
1/1/09
1/1/08
1/1/07
1/1/06
1/1/05
1/1/04
1/1/03
1/1/02
1/1/01
1/1/00
1/1/99
1/1/98
1/1/97
1/1/96
1/1/94
1/1/95
1/1/93
1/1/92
1/1/91
1/1/90
1/1/89
0
Fig. 2. Time series of monthly B. glandula recruitment (monthly mean number of recruits per day; error bars are not shown for clarity) at ten sites along the Oregon coast, arranged
north to south from top to bottom. Note that ordinate scales differ among sites. The dashed vertical line separates the 1990s from the 2000s. Note that B. glandula recruitment differed
between decades at only two sites, SR and SH, where it was greater in the 2000s (two-way ANOVA on ln average recruitment by site and decade; decade p b 0.05).
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
239
smaller pre-competent settlers and larger post-recruitment secondary settlers (e.g., Bayne, 1964), so most individuals are likely to have
been recruits. Although some have reported success at visually
identifying mussel species at the recruit stage (Martel et al., 1999),
we have been unable to repeat this success. Recent genetic analyses
indicate, however, that most (usually N90%) mussel recruits in our
collectors are Mytilus trossulus with the remainder consisting of
M. californianus (P. Raimondi and B. Menge, unpublished data).
Barnacle recruitment was estimated using 10 × 10 cm PVC plastic
plates that are coated on the top side with Saf-T-Walk™ (3M,
Minneapolis, MN, USA), a uniformly textured rubbery surface that
facilitates barnacle settlement (e.g., Farrell et al., 1991; Menge et al.,
1999; Witman et al., 2010). Collectors (n = 5) were established at
wave exposed locations in the mid intertidal zone at all sites and also
the high and low zones at BB and SH. Levels on the shore were similar
across all sites. At BB and SH, collectors were also established in wave
protected sites. Collector locations were the same each month, with
occasional moves to a spot a few cm away when attachment holes
needed redrilling due to erosion. The collectors were arrayed evenlyspaced across ~20–30 m of shore. They were replaced monthly, and
recruits were identified and counted under a dissecting microscope in
2.2. Recruitment
Following Connell (1985), we define recruitment as the number of
settlers surviving a period of time after metamorphosis, up to one
month (e.g., Broitman et al., 2008; Menge et al., 2009, 2010a). Rates of
recruitment of barnacles and mussels were estimated using replicated
deployments of larval collectors in the mid and low intertidal zones
(plastic mesh balls for mussels and settlement plates for barnacles;
e.g., Farrell et al., 1991; Menge, 1992; Barth et al., 2007; Witman et al.,
2010). Mussel plantigrade larvae attach to the filaments of the mesh
balls (SOS Tuffy Pads, The Clorox Company, Oakland, CA, USA) that
mimic the filamentous algal and mussel byssal thread surfaces that
constitute common settlement sites in nature (e.g., Paine, 1974). We
note that recruitment to tuffy collectors does not necessarily
reflect realized recruitment to the intertidal as a whole, since realized
recruitment is a function of the availability of the appropriate substratum.
Thus, numbers of mussels in tuffies are probably most accurately
considered “recruitment potential” and this is our meaning in this paper
when we refer to mussel “recruitment.”
Following our earlier practice (Rilov et al., 2008), for mussels we
analyzed only individuals between 230 and 280 μm. This excludes
150
Cape Meares
100
1990's > 2000's
50
0
150
Fogarty Creek
100
50
0
150
Boiler Bay
100
50
0
150
Seal Rock
100
C. dalli recruits * d-1
50
0
150
Yachats Beach
100
50
0
150
Strawberry Hill
100
50
0
150
Tokatee Klootchman
100
50
0
150
Cape Arago
100
50
0
50
40
30
20
10
0
50
40
30
20
10
0
Cape Blanco
09
08
07
06
05
04
03
02
01
00
99
98
97
96
95
93
94
92
91
90
89
Rocky Point
Fig. 3. Time series of monthly C. dalli recruitment at ten sites, arranged from north to south from top to bottom. Note that ordinate scales differ among sites. The dashed vertical line
separates the 1990s from the 2000s. Note that C. dalli recruitment differed between decades at only one site, CM, where it was greater in the 1990s (two-way ANOVA on ln average
recruitment by site and decade; decade p b 0.05).
240
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
influenced recruitment in the following month (barnacle and mussel
planktonic-larval durations are about two to four weeks; Strathmann,
1987).
the laboratory. Monthly samples were initiated in July 1989 at BB and
SH; the number of sites sampled increased steadily starting in 1993 to
a maximum of 10 sites from 2001 on. The month-long deployment of
these collectors means that the collected juveniles include both
settlers (individuals that have settled within the last 24–48 h) and
recruits (individuals that settled between ~2 and 30 days prior to
collection; Connell, 1985; Menge et al., 2010a). For barnacles, the premetamorphic cyprid phases are termed “settlers” and metamorphosed small barnacles are termed “recruits.” Analyses presented
elsewhere (Broitman et al., 2008; Menge et al., 2010a) show that
numbers of settlers and recruits are positively correlated, suggesting
that post-settlement mortality was density independent. Barnacle
recruits can be identified to species and we present data for the two
most common species, C. dalli and B. glandula. As with mussel
recruitment, densities of recruits on the uniformly-textured collector
plates reflect recruitment potential; actual recruitment density on
variably-textured rock surfaces is usually lower and more variable
than on plates (Menge et al., 2010a). For simplicity, we refer to the
numbers of juveniles counted on plate collectors as “recruits,” and to
patterns of “recruitment” of barnacles.
2.4. Data analysis
Analyses were carried out using JMP 8.0 (SAS Institute Inc, 2008),
R (http://lib.stat.cmu.edu/R/CRAN/), and MATLAB (http://www.
mathworks.com/products/matlab/). To reduce the number of climatic
and environmental variables in our analysis and achieve orthogonality
or independence of the independent variables, we used principle
component analysis (PCA; Legendre and Legendre, 1998). To investigate
the relationships between the environmental variables (climate cycles
and upwelling), expressed as principle components, and recruitment of
each species, we used three approaches: multiple linear regression (e.g.,
Legendre and Legendre, 1998), multiple quantile regression analysis
(e.g., Cade and Noon, 2003), and wavelet analysis (e.g., Torrence and
Compo, 1998). Multiple linear regression provides a coarse-scale look at
which factors are most closely associated with recruitment of each
species. By relating the mean of the response variable to explanatory
variables, classical linear multiple regression assumes that this
relationship is constant across different levels of the response variable.
However, variation in this relationship can reveal important insights
about the relative importance of each explanatory variable across
different levels of the response variable. For example, recruitment may
be unrelated to upwelling at low levels of recruitment, but strongly
related to upwelling at high levels of recruitment. This type of variation
would indicate that recruitment is (1) limited by processes unrelated to
upwelling at low levels of recruitment and (2) limited by upwelling at
high levels of recruitment. We explored this type of variability by
relating the environmental and climate PCA axes to different levels
2.3. Climate data
We obtained climate data from several online sources. ENSO data
were obtained from http://cdc.noaa.gov/ENSO/enso.mei_index, PDO
data were obtained from http://jisao.washington.edu/pdo/, and NPGO
data were obtained from http://www.ocean3d.org/npgo. The Bakun
upwelling index data were obtained from http://pfeg.noaa.gov. In all
cases, monthly averages were matched to monthly recruitment data
lagged by one month, under the assumption that ocean conditions
that affected larvae in the water column in one month were those that
120
100
140
A. CM
80
60
60
40
*
20
*
*
0
40
20
B. glandula recruits * d-1
*
0
1 2 3 4 5 6 7 8 9 10 11 12
60
1 2 3 4 5 6 7 8 9 10 11 12
350
C. BB
D. SR
300
50
250
40
200
30
*
20
10
150
100
*
*
50
0
0
1 2 3 4 5 6 7 8 9 10 11 12
100
80
50
E. SH
*
1 2 3 4 5 6 7 8 9 10 11 12
F. CA
40
60
30
40
20
20
1989-1999
2000-2008
100
80
70
B. FC
120
10
*
0
0
1 2 3 4 5 6 7 8 9 10 11 12
1 2 3 4 5 6 7 8 9 10 11 12
month of year
Fig. 4. Seasonality of recruitment of B. glandula in 1989–1999 compared to 2000–2008 at six sites with the longest time series. Asterisks indicate months in which recruitment
differed between the decades at each of the sites.
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
(i.e. quantiles) of recruitment using multiple quantile regression
analysis (Cade and Noon, 2003).
Investigation of how the magnitude and pattern of variability of
recruitment varies through time is best achieved with time series
analysis (e.g., Chatfield, 1996). Two key assumptions of classical time
series analysis (i.e. spectral analysis) are that the time series is
continuous, with no missing data, and “stationarity,” or that all moments
of the time series (i.e. mean, variance, kurtosis, etc.) do not change in
time (Chatfield, 1996; Torrence and Compo, 1998; Cazelles et al., 2008;
Rouyer et al., 2008a, 2008b). Here, the assumption of stationarity is
clearly violated: major changes in recruitment have occurred at several
scales (decadal, in relation to ENSO, etc.). Wavelet analysis is ideal for
non-stationary time series because it can quantify the dominant modes
of variability and how these modes vary in time (Torrence and Compo,
1998; Cazelles et al., 2008). Recent work has shown that wavelet
analysis can be highly robust to missing values (Cazelles et al., 2008).
When they occurred (data were missing in 19 of 233 months or 8.25%
for mussels, and 15 of 233 months or 6.44% for barnacles), gaps in our
time series were filled in using linear interpolation prior to conducting
wavelet analysis. Cubic spline interpolation gave nearly identical results.
We provide a complete overview of wavelet analysis in Appendix 1.
Briefly, we used cross-wavelet and wavelet coherence to quantify,
respectively, the covariance and the correlation between recruitment
and each environmental variable in the time-frequency domain. These
two methods provide complementary information about patterns of
covariation between pairs of time series. Cross-wavelet analysis
identifies regions where time series have high common power, whereas
wavelet coherence identifies regions where time series show correlated
(or coherent) fluctuations (Cazelles et al., 2008; Grinsted et al., 2004).
We present the wavelet coherence results in the main text and provide
the cross-wavelet analysis in Appendix 2.
3. Results
3.1. Recruitment
Previously, we reported that recruitment rates of mussels
increased dramatically in the early 2000s (time series shown in
Menge et al., 2009), and showed that on average, barnacles did not
exhibit a similar increase. These time series ran up to 2006 in Menge
et al. (2009), and have now been extended through 2008 (e.g., see
Menge et al., 2010b). The detailed time series for the barnacles
B. glandula and C. dalli show that recruitment has changed little since
the 1990s (Figs. 2 and 3). In contrast to mussel recruitment, the time
series of monthly B. glandula recruitment showed only slight interdecadal changes in abundance (see Appendix Fig. 2 in Menge et al.,
2010b). Of the six (out of ten) sites with time series that began in the
mid-1990s or earlier, two (SR and SH) had higher recruitment in the
2000s and four did not (Appendix Fig. 2 in Menge et al., 2010b). For
SR, this difference seems driven largely by the 2008 recruitment
season, in which extraordinarily high recruit densities occurred
(Fig. 2). At SH, the difference appears driven by the 2006, 2007 and
2008 recruitment seasons, all of which had denser recruitment than
all previous years except 1995 (Fig. 2). We note that the 1990s data for
CM, FC, and YB were limited in extent, so it is likely the inter-decadal
trends are better reflected by the other four sites (Fig. 2). These
differences were sufficient to make the effect of decade significant, but
overall the total variance explained was modest (two-way ANOVA;
decade R2 = 0.11). In general, however, in comparison to mussels,
changes in B. glandula recruitment after 1999 were minimal.
The barnacle C. dalli is a flatter, smaller barnacle than B. glandula
and in Oregon is not a strong facilitator of mussels (B. Menge, personal
observations). In predation-impact experiments (e.g., Dayton, 1971;
100
50
40
A. CM
80
30
60
20
40
10
20
1 2 3 4 5 6 7 8 9 10 11 12
C. dalli recruits * d-1
70
*
1 2 3 4 5 6 7 8 9 10 11 12
C. BB
60
50
40
40
30
30
20
20
10
10
0
D. SR
0
1 2 3 4 5 6 7 8 9 10 11 12
1 2 3 4 5 6 7 8 9 10 11 12
25
50
E. SH
20
F. CA
15
30
20
10
*
10
0
1989-1999
2000-2008
70
50
40
B. FC
0
0
60
241
*
5
*
0
1 2 3 4 5 6 7 8 9 10 11 12
1 2 3 4 5 6 7 8 9 10 11 12
month of year
Fig. 5. Recruitment of C. dalli by month in 1989–1999 compared to 2000–2008 at six sites with the longest time series. Error bars are 1 SE of the mean; asterisks indicate month/site
combinations when decadal means were different.
242
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Menge, 2003), C. dalli is often the most abundant species in treatments
with predators present, consistent with the view that C. dalli is an
unattractive prey (Paine, 1981). Exclusion of predators commonly
results in dense covers of B. glandula which are then overgrown by
mussels M. trossulus. Nonetheless, because C. dalli can occupy
significant amounts of space in the low intertidal at some sites and
because chthamaloids elsewhere can facilitate mussel recruitment
(Navarrete and Castilla, 1990), we also examined changes of this
barnacle across the decadal shift from the 1990s to the 2000s.
As with B. glandula, recruitment of C. dalli also did not change
between decades (Fig. 3). Comparing average rates of recruitment by
site and decade reveals inter-decadal differences at only one of six sites
with sufficiently long time series (Fig. 3; Cape Meares = CM), and this
difference was opposite to that shown by both mussel and B. glandula
recruitment (see Appendix Fig. 2 in Menge et al., 2010b). Finally,
recruitment densities of this smaller barnacle were generally much
lower than those of B. glandula (Fig. 3), suggesting a more limited
capacity to reach the high densities that are needed for effective
facilitation of mussels.
3.2. Seasonal shifts
Although inter-decadal differences in barnacle recruitment were
limited, it is possible that the seasons of recruitment differed between
decades, thus potentially altering their interaction with mussels. As
we have shown previously, mussel recruitment peaks have generally
shifted one to two months later in the season, from July/August to
September/October (Menge et al., 2009).
Overall, seasonal shifts in barnacle recruitment were slight, and
although statistically significant shifts occurred in a few cases, these
shifts seem ecologically trivial. For example, at five of six sites (CM, FC,
BB, SR, and SH), small seasonal shifts in recruitment occurred in
B. glandula, and changes at CA, though not significant, show a trend
similar to the other sites (Fig. 4, Appendix 2 Table 1). In general, the
season of recruitment at these sites has expanded, with higher
recruitment occurring either earlier (March), later (October–December),
or both, in the year.
The small barnacle C. dalli showed no inter-decadal seasonal shifts
although non-significant trends towards earlier recruitment were
evident at all six sites (Fig. 5, Appendix 2 Table 2). Thus, although
B. glandula recruitment season appears to have expanded in the
2000s, these trends are not consistent with a shift in phenology that
might have reduced the amount of barnacle substratum available for
mussel recruits. In fact, the tendency for higher barnacle recruitment
later in the year is in the same direction as that seen for mussels,
suggesting that at some sites, if anything, barnacle substratum may be
more, not less available.
3.3. Linear and quantile multiple regressions
We first performed a PCA of all 1-month lagged environmental
variables in order to (1) reduce the number of independent
environmental variables used in the multiple regressions of recruitment and (2) prevent any correlation between the independent
environmental variables from affecting the estimates of the multiple
regression coefficients (i.e., avoid the “bouncing beta” issue due to
multi-collinearity). PCA of all 1-month lagged environmental factors
(MEI, NPGO, PDO and upwelling) resulted in four axes, the first three
of which explained over 88% of the total variance in the environment
(Fig. 6, Tables 1 and 2). PCA axis 1 mainly represents MEI
(loading = 0.6) and PDO (0.59); PCA axis 2 mainly represents
upwelling (0.96); PCA axis 3 mainly represents NPGO (0.82) (Fig. 6,
Table 1).
We further limited the number of PCA axes used for each multiple
regression of recruitment by conducting forward selection on all PCA
axes using a double-stopping criterion based on α = 0.05 and the
Fig. 6. Principal component analysis of all environmental variables (i.e. MEI, NPGO, PDO
and the Bakun upwelling index) from 1989 to 2008. (A) Proportion of the total
environmental variance explained by each PCA axis (bars) and cumulative proportion
of the total variance explained (blue line). (B) Biplot of PCA analysis of the
environment. Red dots indicate site scores and labeled blue vectors indicate the
contribution of each environmental variable to the principle component axes. Axis 1
mainly represents MEI (0.598) and PDO (0.591); axis 2 mainly represents upwelling
(0.96); axis 3 mainly represents NPGO (0.82). See Tables 1 and 2 for more details about
the loadings for each axis.
adjusted R2 of the full multiple regression model containing all four PCA
axes (Blanchet et al., 2008). This resulted in the selection of PCA axes 2–3
for Mytilus spp. recruitment (Table 3) and B. glandula recruitment
(Table 4), and PCA axes 1–3 for C. dalli recruitment (Table 5).
Multiple linear regression of log-transformed recruitment against
all forward-selected environmental PCA axes showed similar effects
of upwelling (PCA axis 2) and NPGO (PCA axis 3) on recruitment
across species: recruitment tended to increase primarily with
upwelling and NPGO (Fig. 7). The strength of these relationships
were similar among species (Mytilus spp.: p b 0.0001, Adj. R2 = 0.298;
B. glandula: p b 0.0001, Adj. R2 = 0.386; C. dalli: p b 0.0001, R2 = 0.369;
see Fig. 7, Tables 3–5).
Multiple quantile regression of recruitment as a function of all
forward-selected environmental PCA axes showed interspecific
Table 1
Loadings from principal component analysis of environmental variables. Abbreviations:
MEI = Multivariate El Niño/Southern Oscillation Index; NPGO = North Pacific Gyre
Oscillation; PDO = Pacific Decadal Oscillation.
Environmental variable
PCA axis 1
PCA axis 2
PCA axis 3
PCA axis 4
MEI
NPGO
PDO
Upwelling
0.598
− 0.535
0.591
0.081
− 0.17
− 0.17
− 0.12
0.96
0.26
0.82
0.44
0.25
− 0.74
− 0.10
0.66
− 0.07
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Table 2
Proportion of environmental variation explained by principle component analysis.
Standard deviation
Proportion of variance
Cumulative proportion
PCA axis 1
PCA axis 2
PCA axis 3
PCA axis 4
1.370
0.469
0.469
1.011
0.255
0.725
0.796
0.158
0.883
0.684
0.117
1
differences in the relationship between recruitment quantiles and the
environment. Overall, B glandula recruitment was consistently and
strongly related to both climatic (NPGO) and environment (upwelling) indices (PCA axes 2 and 3), respectively; (Fig. 8), whereas the
effects of climatic and environmental indices were contingent upon
recruitment rates for both C. dalli and Mytilus spp. The effect of climate
(NPGO) on C. dalli recruitment was strongly unimodal, with low rates
of recruitment being unrelated to climate, intermediate to high rates
showing a strong positive relationship with recruitment, and the
highest rate being unrelated to climate (PCA axis 3; Fig. 9). The response
of C. dalli recruitment to the environment was saturating, with low rates
being unrelated to the environment and high rates being strongly
related to the environment. Mussel recruitment relationships to
upwelling and NPGO were similar to those for C. dalli, and thus also
show a saturating response to both climate and environmental indices
(Fig. 10). Hence, the effect of climate and the environment on
recruitment is strong and consistent across recruitment rates for B.
glandula, but contingent upon recruitment rates for both C. dalli and
mussels, with weak effects at low rates and strong effects at
intermediate and high rates. These results suggest that factors other
than upwelling, MEI, NPGO and PDO play an important role in
controlling recruitment variability at low and high rates (Cade and
Noon, 2003).
3.4. Relationship between recruitment and environmental variables
We used cross-wavelet and wavelet coherence analyses to
determine the pair-wise relationships between recruitment and
each environmental variable (see Appendix 1). Wavelet coherence
and cross-wavelet describe, respectively, the correlation and the
covariance between pairs of time series in the time-frequency domain
(Cazelles et al., 2008). Since both analyses yielded similar results, we
present the wavelet coherence results in the main text (Figs. 11–13)
and include the cross-wavelet analyses in Appendix 2 (Figs. 3–5). We
note that strong patterns of covariation in these analyses do not
necessarily imply causation (Cazelles et al., 2008). A strong association at annual frequencies between recruitment and upwelling, for
example, could simply reflect the fact that both have strong annual
cycles.
The relationships between recruitment and environmental variables clearly varied among species, with periodicity, and through time
(Figs. 11–13; Appendix 2, Figs. 3–5). For all species, most times of
coherence between environment and recruitment occurred at low
periods (~8–16 months), that is approximately on an annual scale,
particularly with PDO and upwelling. Relationships were complex
Table 3
Properties of the forward-selected environmental PCA axes used in multiple regression
analysis of log-transformed Mytilus spp. recruitment.
PCA axis 3
PCA axis 2
R2
Cumulative R2
Cumulative Adj. R2
F
p-value
0.197
0.107
0.197
0.304
0.194
0.298
56.685
35.369
0.001
0.001
Note: Forward selection based on double-stopping criterion (α = 0.05, Adj. R2 of full
model containing all four PCA axes) and 1000 permutations of the residuals under the
reduced model.
243
Table 4
Properties of the forward-selected environmental PCA axes used in multiple regression
analysis of log-transformed B. glandula recruitment.
PCA axis 2
PCA axis 3
R2
Cumulative R2
Cumulative Adj. R2
F
p-value
0.212
0.179
0.212
0.391
0.209
0.386
62.267
67.702
0.001
0.001
Note: Forward selection based on double-stopping criterion (α = 0.05, Adj. R2 of full
model containing all four PCA axes) and 1000 permutations of the residuals under the
reduced model.
however, with both in-phase and anti-phase and longer period (16–
32 months or ~2 year and 32–64 months or ~ 4–5 year) associations.
B. glandula recruitment was associated with MEI (in-phase) and
NPGO (anti-phase). These relationships occurred only briefly, from
the late 1990s to about 2000 (MEI) or 1999–2002 (NPGO), but varied
dramatically in strength, with intermittent and weak-to-moderate
levels of in-phase fluctuations with MEI, especially during the strong
1998 ENSO event and subsequent La Niña (Fig. 11B). The anti-phase
fluctuations between recruitment and NPGO overlapped the La Niña
period, extending into the early 2000s (Fig. 11C). PDO and upwelling
were much more strongly associated with recruitment of B. glandula.
PDO showed strong but fluctuating associations (in-phase and antiphase fluctuations) at 8–16 month periodicities, strong and persistent
in-phase fluctuations at ~2 year periodicities from 1993 to 2005, and
even apparent strong and persistent relationships at 4–5 year
periodicities (at least during the middle of the time series, Fig. 11D).
Upwelling showed strong associations at 8–16 month periods in the
mid 1990s and from about 2000 on, with a notable gap during the
time of the 1997–2000 ENSO/La Niña and at 16–32 month period in
the mid 1990s (Fig. 11E). Thus, overall, B. glandula recruitment varied
with all environmental measures, most persistently at annual
periodicities with PDO and upwelling, but with interesting gaps that
seem to coincide with the strong 1997–2000 ENSO/La Niña and shortlived shift in PDO.
C. dalli recruitment showed very similar patterns of variation,
with intermittent and brief in-phase fluctuations with MEI (~8–
16 months) during the 1997–2000 ENSO/La Niña, weaker and
sporadic in- and out-of-phase fluctuations with NPGO at low periods
(~8–16 months), periodically strong in- and out-of-phase fluctuations
with PDO, and more persistent in-phase fluctuations at low periods
(~8–16 months) with upwelling (Fig. 12B–E). We again note the
interesting coincidence of the strong association with El Niño/La Niña
and gaps in PDO and upwelling coherence during this time. Mytilus
spp. recruitment showed very similar patterns to those of the
barnacles (Fig. 13).
3.5. Relationship between recruitment of different species
We also used cross-wavelet and wavelet coherence analysis to
determine patterns of interspecific co-variation in recruitment
(Fig. 14 and Appendix 2, Fig. 6). In general, all species showed strong
and coherent fluctuations in recruitment at periodicities ranging up to
~two years. Recruitment of all species showed highly coherent
Table 5
Properties of the forward-selected environmental PCA axes used in multiple regression
analysis of log-transformed C. dalli recruitment.
PCA axis 2
PCA axis 3
PCA axis 1
R2
Cumulative R2
Cumulative Adj. R2
F
p-value
0.211
0.141
0.025
0.211
0.352
0.377
0.208
0.347
0.369
61.852
50.124
9.167
0.001
0.001
0.005
Note: Forward selection based on double-stopping criterion (α = 0.05, Adj. R2 of full
model containing all four PCA axes) and 1000 permutations of the residuals under the
reduced model.
244
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
(A) Intercept
0.0
0.2
0.4
0.6
0.8
1.0
0.8
1.0
0.8
1.0
(B) PCA axis 2 (Upwelling)
0.0
0.2
0.4
0.6
(C) PCA axis 3 (NPGO)
0.0
Fig. 7. Multiple linear regressions used to describe log-transformed recruitment as a
function of forward-selected environmental PCA axes 2 and 3 for (A) B. glandula
(Adjusted R2 = 0.386, p-value b 2.2 × 10− 16), (B) C. dalli (Adjusted R2 = 0.369, pvalue b 2.2 × 10− 16) and (C) Mytilus spp (Adjusted R2 = 0.300, p-value b 2.2 × 10− 16).
PCA axis 2 mainly represents upwelling (0.96) and PCA axis 3 mainly represents NPGO
(0.82). See Tables 1 and 2 for more details about the loadings for each axis.
(Fig. 14) but weak (Appendix 2, Fig. 6) fluctuations at very high
frequencies (or lower periods). All associations were in-phase, or
positively correlated. Recruitment of the barnacle species showed the
most similarities in space/time (Fig. 14B), but mussel recruitment was
generally similar in coherence with both barnacle species as well
(Fig. 14C, D).
4. Discussion
The factors underlying the often massive variations in recruitment
are notoriously difficult to discern. Recruitment can vary because of
variation in climate, in oceanographic variability (e.g., upwelling,
alongshore currents, offshore breezes, internal tidal bores, wave
regime, and coastal geomorphology), in reproductive output, in larval
availability and behavior, and in immediate post-settlement factors
such as species interactions and disturbance (Pineda, 2000). In this
analysis, we attempted to detect the relationships between largescale climatic factors as expressed in the MEI, NPGO and PDO climatic
indices and intermediate-scale variation in upwelling, and recruitment of three major component taxa of rocky intertidal communities
along the US west coast. Our results suggest that recruitment shows
0.2
0.4
0.6
Fig. 8. Quantile and linear regressions used to describe log-transformed B. glandula
recruitment (y) as a function of forward-selected environmental PCA axes 2 ðx1 Þ and 3
ðx2 Þ ð log 10 ðyÞ = β1 x1 + β2 x2 + β0 + εÞ. Red solid lines indicate linear multiple
regression estimates for each coefficient and intercept, and red dashed lines indicate
95% confidence intervals (linear multiple regression: p-value b 2.2 × 10− 16, Adjusted
R2 = 0.386). Quantile regression was used to describe how the relationship between
recruitment and forward-selected environmental PCA axes varied across quantiles of
recruitment. Black lines and dots represent estimates of each coefficient and intercept
for different quantiles τ of recruitment. Gray regions depict 90% confidence intervals
around quantile regression coefficient and intercept estimates.
surprisingly strong links to these regional to ocean-basic scale
variables, but also that the relationships are highly complicated in
space and time.
4.1. Patterns of recruitment
As with mussel recruitment (Menge et al., 2009), barnacle
recruitment varied as a function of season, site and latitude, as well as
in time (Appendix 2 Fig. 2, Figs. 2 and 3). Temporal and spatial variation
in barnacle recruitment was, however, less than that observed
for mussels. Despite these differences, all three analyses aimed at
determining the relationship between climate, regional environment
(upwelling) and recruitment led to similar results. Environmental
variation, even at large, ocean-basin scales, can explain substantial
amounts of variation in recruitment, ranging from 30 to 40%. Given the
many local sources of variation, this amount of explanatory power is
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
(B) PCA axis 1 (PDO, MEI)
(A) Intercept
0.0
0.2
245
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
(C) PCA axis2 (Upwelling) (D) PCA axis 3 (NPGO)
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Fig. 9. Quantile and linear regressions used to describe log-transformed C. dalli recruitment (y) as a function of forward-selected environmental PCA axes 1 ðx1 Þ, 2 ðx2 Þ and 3 ðx3 Þ
ð log 10 ðyÞ = β1 x1 + β2 x2 + β3 x3 + β0 + εÞ. Linear multiple regression: p b 2.2 × 10− 16, Adjusted R2 = 0.369. See Fig. 8 caption for further explanation.
surprisingly high. Thus, as expected from hypothesis H1, each taxon was
sensitive to climate-related variation, and responded similarly to these
sources of variability (hypothesis H2). However, contrary to hypothesis
H3, barnacle recruitment showed only slight shifts between the 1990s
and 2000s.
4.2. Large-scale forcing, sequential filters and the role of larval transport
in determining recruitment
Pineda (2000) modeled the process by which larvae travel to the
adult habitat as a series of filters that sequentially reduce the numbers
of larvae as they progress towards recruitment into the adult
habitat, with multiple filtering mechanisms. Our analysis suggests
that, at large scales of space and time, these three taxa respond
somewhat similarly to climate and environmental effects. At the same
time, the 30–40% of variance explained by these factors means that
N60% of the variance in recruitment is due to other factors,
presumably including those highlighted above. The local-scale,
among-site and temporal variations evident in the actual time series
for each of these species (Figs. 2 and 3, Fig. 2 in Menge et al., 2009) is
presumably a reflection of the influence of such factors.
The similarities in the responses of recruitment to climate and
environmental variability at large spatio-temporal scales documented
here and the differences observable in the time series data highlight
how some processes can vary in their role as a function of the scale
considered. For example, our analysis suggests that upwelling affects
all species strongly, at least at longer time scales (annually or multiannually). But upwelling also varies on scales of days to months
within years, with alternations between upwelling and downwelling
events at scales of days to weeks being a particularly strong shortterm scale. The role of upwelling in influencing recruitment has been
controversial, and current evidence is actually contradictory on the
role of upwelling in recruitment of these species. For B. glandula for
instance, some studies find an impact of upwelling regime, with
apparent strong (e.g., Farrell et al., 1991), moderate (Dudas et al.,
2009) or weak effects of upwelling relaxation on recruitment (Shanks,
2009), while others find no relationship with upwelling (Morgan
et al., 2009; Shanks and Shearman, 2009).
At finer, within-region, among-site scales, a host of processes must
be considered. Thus, the different patterns between mussel and
barnacle recruitment seen at these scales likely reflect both spatiotemporal and qualitative differences in larval transport mechanisms,
although differences in susceptibility to post settlement mortality
sources could also play a role. The most likely larval transport
mechanisms involved at the more local and shorter scales include
transport by short-term upwelling–downwelling fluctuations, winddriven transport including diurnal fluctuations in sea breezes,
current-related variation in larval behavior, wave forcing, and
tidally-driven change, such as internal waves (Farrell et al., 1991;
Hawkins and Hartnoll, 1982; Pineda, 1991, 1999; Shanks and Brink,
2005; Woodson et al., 2007; Rilov et al., 2008; Dudas et al., 2009;
Morgan et al., 2009; Shanks, 2009; Shanks et al., 2010; Witman et al.,
2010).
Recent studies have sharpened the focus on processes delivering
larvae across the last few hundreds of meters to shore. For example,
consistent with a growing number of examples for fish larvae
(Warner and Cowen, 2002; Swearer et al., 2002; Cowen et al.,
2006), extensive cross-shelf sampling off Bodega Bay, California,
indicates that the larvae of the majority of invertebrates develop close
to shore (b3 km; Morgan et al., 2009) rather than being swept far
offshore during upwelling as previously postulated. In another
example, spatial variation among sites in mussel recruitment was
unrelated to abundance of larvae just outside the surf zone (Rilov
et al., 2008), suggesting a key role for processes within the surf zone in
determining spatial and likely temporal variation in recruitment.
Shanks et al. (2010) found that recruitment of B. glandula to boulders
tended to be higher on dissipative than on reflective sandy beaches,
and was greater on days with higher significant wave heights,
suggesting the importance of micro-hydrodynamical processes as
larvae approach their final destination. Shanks et al. (2010) also note
that these patterns may not translate to continuous rocky shores, and
indeed, recruitment of both barnacles at our sites tends to be higher,
246
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
(A) Intercept
0.0
0.2
0.4
0.6
0.8
1.0
0.8
1.0
0.8
1.0
(B) PCA axis 2 (Upwelling)
0.0
0.2
0.4
0.6
(C) PCA axis 3 (NPGO)
0.0
0.2
0.4
0.6
Fig. 10. Quantile and linear regressions used to describe log-transformed Mytilus spp.
recruitment (y) as a function of forward-selected environmental PCA axes 2 ðx1 Þ and 3 ðx2 Þ
ð log 10 ðyÞ = β1 x1 + β2 x2 + β0 + εÞ. Linear multiple regression: p-valueb 2.2×10−16,
Adjusted R2 =0.298. See Fig. 8 caption for further explanation.
not lower, at steeper and presumably more reflective sites (e.g., SR, FC
vs. SH, TK; Appendix Fig. 2).
4.3. Interactive effects of climate patterns
The wavelet analyses suggest that recruitment rates respond in
complex ways to the different climate and environment measures. In
addition to the issues discussed above, the fact that coherences
between climate and environment, and recruitment, vary in whether
they are in-phase (positively correlated) or in anti-phase (negatively
correlated) suggests that the indices are aliasing the actual factor(s)
that larvae are responding to as oceanographic conditions fluctuate.
Pairwise comparisons suggest that during the 1989–2008 time period,
this is the case. MEI, PDO and NPGO show highly non-stationary (i.e.
variable in time) patterns of co-variation (Appendix 2 Fig. 7). For
example, MEI and NPGO showed strong and anti-phase patterns of
fluctuations at high periods (~64 months), but also in-phase moderateto-weak fluctuations at low periods (~8–16 months). Strong fluctuations occurring at high periods remained consistently anti-phase in
time, whereas weak-to-moderate fluctuations tended to shift from inphase (e.g. from 03–93 to 09–98) to anti-phase (e.g. 02–04 to 11–06) in
Fig. 11. Standardized time series and pairwise wavelet coherence of B. glandula
recruitment and monthly (B) MEI, (C) NPGO, (D) PDO and (E) upwelling from 1989 to
2008. Wavelet coherence describes the amount of correlation (or coherence) in the
fluctuations of two time series (color bar) in the time-frequency domain. Black arrows
indicate the phase angle between the fluctuations of the two time series. Arrows
pointing to the left indicate anti-phase fluctuations, whereas arrows pointing to the
right indicate in-phase fluctuations. Black contour lines indicate regions in which the
observed wavelet coherence values are statistically significantly (α = 0.05) based on
Monte Carlo randomizations applied to 1000 pairs of surrogate time series whose firstorder autoregressive coefficients match those of the original time series. Blacked-out
areas represent the “cone-of-influence” where edge effects can influence the analysis.
Period is in units of months. The x-axis indicates the time (month-year).
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Fig. 12. Standardized time series and pairwise wavelet coherence of C. dalli recruitment
and monthly (B) MEI, (C) NPGO, (D) PDO and (E) upwelling indices from 1989 to 2008.
See Fig. 11 caption for further explanation.
time. MEI and PDO showed strong and persistent in-phase fluctuations
at high periods (~64 months) and transient, weak-to-moderate inphase fluctuations at low periods (~8–12 months; e.g. from 12–95 to
05–01). Finally, PDO and NPGO showed strong and persistent antiphase fluctuations at high periods (~64 months), but transient and inphase moderate-to-weak in-phase fluctuations at low periods (~6–
16 months). Since all the climatic patterns are large, basin-scale
phenomena, the specific mechanisms underlying the apparent tele-
247
Fig. 13. Standardized time series and pairwise wavelet coherence of Mytilus spp.
recruitment and monthly (B) MEI, (C) NPGO, (D) PDO and (E) upwelling from 1989 to
2008. See Fig. 11 caption for further explanation.
connections between these large-scale patterns to regional coastal
conditions to region- and site-scale patterns of recruitment will clearly
be challenging to identify.
5. Conclusions
In the quest to determine the causes of the often puzzling variation
in patterns of recruitment of marine organisms, most attention has
been focused on processes and mechanisms that operate on shorter
248
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Those working with the recruitment datasets included B. Daley, G.
Hudson, M. Foley, J. Pamplin, C. Carlson, M. Robart, C. Cardoni, G.
Murphy, C. Pennington and M. Poole. During this work we also
received useful comment, criticism, help and encouragement from
many labmates, graduate students and postdocs too numerous to
mention by name, but to all we are grateful. S. Dudas, S. Close, A. Iles,
J. Rose, and D. Menge provided valuable comments. J. Sapp and
M. Frenock provided assistance in data extraction and management.
We thank B. Abbott for providing long-term access to our FC study
site, the “jewel” of the Oregon coast. The research was supported by
grants from the NSF, the Andrew W. Mellon Foundation, and the
David and Lucile Packard Foundation, the Gordon and Betty Moore
Foundation, and by endowment funds from the Wayne and Gladys
Valley Foundation. This is contribution number 382 from PISCO, the
Partnership for Interdisciplinary Studies of Coastal Oceans funded
primarily by the Gordon and Betty Moore Foundation and David and
Lucile Packard Foundation. [SS]
References
Fig. 14. Standardized time series and pairwise wavelet coherence of B. glandula, C. dalli
and Mytilus spp. recruitment from 1989 to 2008. See Fig. 11 caption for further
explanation.
time scales at smaller spatial scales. In our analysis, we have taken
advantage of unique recruitment time series for key components of
rocky intertidal communities along the Oregon coast in an attempt to
assess the possible role of climatic and environmental variation. We
found that such large-scale, long-term variation can explain over a third
of the variability in our time series, suggesting that much variation in
recruitment may be due to factors operating on climatic scales. Thus,
addition of these factors to comprehensive models of recruitment that
also include the smaller- and shorter-scale factors may provide better
descriptors, and therefore better predictors, of how recruitment to
marine populations will vary under future environmental scenarios,
including those occurring under climate change.
Supplementary materials related to this article can be found online
at doi:10.1016/j.jembe.2011.02.003.
Acknowledgements
We thank especially the legion of technicians and undergraduate
interns who have helped maintain these long-term field studies.
Barth, J.A., Menge, B.A., Lubchenco, J., Chan, F., Bane, J.M., Kirincich, A.R., McManus, M.A.,
Nielsen, K.J., Pierce, S.D., Washburn, L., 2007. Delayed upwelling alters nearshore
coastal ocean ecosystems in the northern California current. Proc. Natl Acad. Sci.
USA 104, 3719–3724.
Bayne, B.L., 1964. Primary and secondary settlement in Mytilus edulis L. (Mollusca).
J. Anim. Ecol. 33, 513–523.
Berlow, E.L., 1997. From canalization to contingency: historical effects in a successional
rocky intertidal community. Ecol. Monogr. 67, 435–460.
Berlow, E.L., 1999. Strong effects of weak interactors in ecological communities. Nature
398, 330–334.
Blanchet, F.G., Legendre, P., Borcard, D., 2008. Forward selection of explanatory
variables. Ecology 89, 2623–2632.
Brodeur, R.D., Ralston, S., Emmett, R.L., Trudel, M., Auth, T.D., Phillips, A.J., 2006.
Anomalous pelagic nekton abundance, distribution, and apparent recruitment in
the northern California Current in 2004 and 2005. Geophys. Res. Lett. 33, L22S08.
doi:10.1029/2006GL026614.
Broitman, B.R., Blanchette, C.A., Menge, B.A., Lubchenco, J., Krenz, C., Foley, M.M.,
Raimondi, P., Lohse, D., Gaines, S.D., 2008. Spatial and temporal variability in the
recruitment of intertidal invertebrates along the West coast of the U. S. A. Ecol.
Monogr. 78, 403–421.
Cade, B.S., Noon, B.R., 2003. A gentle introduction to quantile regression for ecologists.
Front. Ecol. Environ. 1, 412–420.
Cazelles, B., Chavez, M., Berteaux, D., Menard, F., Vik, J.O., Jenouvrier, S., Stenseth, N.C.,
2008. Wavelet analysis of ecological time series. Oecologia 156, 287–304.
Chan, F., Barth, J.A., Lubchenco, J., Kirincich, A., Weeks, H., Peterson, W.T., Menge, B.A.,
2008. Emergence of anoxia in the California Current Large Marine Ecosystem.
Science 319, 920.
Chatfield, C., 1996. The Analysis of Time Series: An Introduction, 6th edition. Chapman
and Hall/CRC, Boca Raton, Florida, USA.
Chavez, F.P., Strutton, P.G., Friederich, G.E., Feely, R.A., Feldman, G.C., Foley, D.G.,
McPhaden, M.J., 1999. Biological and chemical response of the equatorial Pacific
Ocean to the 1997–98 El Niño. Science 286, 2126–2131.
Connell, J.H., 1985. The consequences of variation in initial settlement vs. post settlement
mortality in rocky intertidal communities. J. Exp. Mar. Biol. Ecol. 93, 11–45.
Connolly, S.R., Roughgarden, J., 1999. Theory of marine communities: competition,
predation, and recruitment-dependent interaction strength. Ecol. Monogr. 69, 277–296.
Cowen, R.K., Paris, C.B., Srinivasan, A., 2006. Scaling of connectivity in marine
populations. Science 311, 522–527.
Dayton, P.K., 1971. Competition, disturbance, and community organization: the
provision and subsequent utilization of space in a rocky intertidal community.
Ecol. Monogr. 41, 351–389.
Di Lorenzo, E., Schneider, N., Cobb, K.M., Chhak, K., Franks, P.J.S., Miller, A.J., McWilliams,
J.C., Bograd, S.J., Arango, H., Curchister, E., Powell, T.M., Rivere, P., 2008. North
Pacific Gyre Oscillation links ocean climate and ecosystem change. Geophys. Res.
Lett. 35, L08607. doi:10.1029/2007GL032838.
Dudas, S.E., Grantham, B.A., Kirincich, A.R., Menge, B.A., Lubchenco, J., Barth, J.A., 2009.
Current reversals as determinants of intertidal recruitment on the central Oregon
coast. ICES J. Mar. Sci. 66, 396–407.
Farrell, T.M., Bracher, D., Roughgarden, J., 1991. Cross-shelf transport causes recruitment to
intertidal populations in central California. Limnol. Oceanogr. 36, 279–288.
Feely, R.A., Sabine, C.L., Lee, K., Berelson, W., Kleypas, J., Fabry, V.J., Millero, F.J., 2004. Impact of
anthropogenic CO2 on the CaCO3 system in the oceans. Science 305, 362–366.
Francis, R.C., Hare, S.R., 1994. Decadal-scale regime shifts in the large marine ecosystems of
the North-east Pacific: a case for historical science. Fish. Oceanogr. 3, 1–13.
Gaines, S.D., Roughgarden, J., 1985. Larval settlement rate: a leading determinant of
structure in an ecological community of the marine intertidal zone. Proc. Natl Acad.
Sci. USA 82, 3707–3711.
Glynn, P.W., 1988. El Niño-Southern Oscillation 1982–83: nearshore population,
community and ecosystem responses. Annu. Rev. Ecol. Syst. 19, 309–345.
B.A. Menge et al. / Journal of Experimental Marine Biology and Ecology 400 (2011) 236–249
Gouhier, T., Guichard, F., Menge, B.A., 2010. Ecological processes can synchronize marine
population dynamics over continental scales. Proc. Natl Acad. Sci. USA 107, 8281–8286.
Grantham, B.A., Chan, F., Nielsen, K.J., Fox, D.S., Barth, J.A., Huyer, A., Lubchenco, J.,
Menge, B.A., 2004. Upwelling-driven nearshore hypoxia signals ecosystem and
oceanographic changes in the Northeast Pacific. Nature 429, 749–754.
Grinsted, A., Moore, J.C., Jevrejeva, S., 2004. Application of the cross wavelet transform
and wavelet coherence to geophysical time series. Nonlinear Processes Geophys.
11, 561–566.
Hare, S.R., Francis, R.C., 1994. Climate change and salmon production in the Northeast
Pacific Ocean. In: Beamish, R.J. (Ed.), Climate Change and Northern Fish
Populations: Can. Spec. Publ. Fish. Aquat. Sci., 121, pp. 357–372.
Harley, C.D.G., Hughes, A.R., Hultgren, K.M., Miner, B.G., Sorte, C.J.B., Thornber, C.S.,
Rodriguez, L.F., Tomanek, L., Williams, S.L., 2006. The impacts of climate change in
coastal marine systems. Ecol. Lett. 9, 228–241.
Hawkins, S.J., Hartnoll, R.G., 1982. Settlement patterns of Semibalanus balanoides (L.) in
the Isle of Man (1977–1981). J. Exp. Mar. Biol. Ecol. 62, 271–283.
Hawkins, S.J., Sugden, H.E., Mieszkowska, N., Moore, P., Poloczanska, E., Leaper, R.,
Herbert, R.J.H., Genner, M.J., Moschella, P.S., Thompson, R.C., Jenkins, S.R.,
Southward, A.J., Burrows, M.T., 2009. Consequences of climate driven biodiversity
changes for ecosystem functioning of North European Rocky Shores. Mar. Ecol.
Prog. Ser. 396, 245–259.
Helmuth, B., Mieszkowska, N., Moore, P., Hawkins, S.J., 2006. Living on the edge of two
changing worlds: forecasting responses of rocky intertidal ecosystems to climate
change. Ann. Rev. Ecol. Evol. Syst. 37, 373–404.
Helmuth, B., Yamane, L., Lalwani, S., Matzelle, A., Tockstein, A., Gao, N., 2011. Hidden
signals of climate change in intertidal ecosystems: what (not) to expect when you are
expecting. J. Exp. Mar. Biol. Ecol. 400, 191–199 (this issue).
Holmgren, M., Scheffer, M., Ezcurra, E., Gutierrez, J.R., Mohren, G.M.J., 2001. El Niño
effects on the dynamics of terrestrial ecosystems. Trends Ecol. Evol. 16, 89–94.
Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Xiaosu, D. (Eds.),
2001. Climate Change 2001: The Scientific Basis. Cambridge University Press,
Cambridge, UK.
Johnson, C.R., Banks, S.C., Barrett, N.S., Cazassus, F., Dunstan, P.K., Edgar, G.J., Frusher, S.D.,
Gardner, C., Haddon, M., Helidoniotis, F., Hill, K.L., Holbrook, N.J., Hosie, G.W., Last, P.R.,
Ling, S.D., Melbourne-Thomas, J., Miller, K., Pecl, G.T., Richardson, A.J., Ridgway, K.R.,
Rintoul, S.R., Ritz, D.A., Ross, D.J., Sanderson, J.C., Shepherd, S.A., Slotwinski, A.,
Swadling, K.M., Taw, N., 2011. Climate change cascades: shifts in oceanography, species'
ranges and subtidal marine community dynamics in eastern Tasmania. J. Exp. Mar. Biol.
Ecol. 400, 17–32 (this issue).
Kleeman, R., Power, S.B., 2000. Modulation of ENSO variability on decadal and longer
timescales. In: Diaz, H.F., Markgraf, V. (Eds.), El Niño and the Southern Oscillation:
Multiscale Variability and Global and Regional Impacts. Cambridge University
Press, Cambridge, UK, pp. 413–442.
Kudela, R.M., Cochlan, W.P., Peterson, T.D., Trick, C.G., 2006. Impacts on phytoplankton
biomass and productivity in the Pacific Northwest during the warm ocean
conditions of 2005. Geophys. Res. Lett. 33, L22S06. doi:10.1029/2006GL026772.
Legendre, P., Legendre, L., 1998. Numerical Ecology, Second English Edition. Elsevier
Science B. V., Amsterdam.
Mann, M.E., Bradley, R.S., Hughes, M.K., 2000. Long-term variability in the El Niño/
Southern Oscillation and associated teleconnections. In: Diaz, H.F., Markgraf, V.
(Eds.), El Niño and the Southern Oscillation: Multiscale Variability and Global and
Regional Impacts. Cambridge University Press, Cambridge, UK, pp. 357–412.
Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific interdecadal
climate oscillation with impacts on salmon production. Bull. Am. Meteorol. Soc. 78,
1069–1079.
Martel, A.L., Robles, C., Beckenbach, K., Smith, M.J., 1999. Distinguishing early juveniles
of eastern Pacific mussels (Mytilus spp.) using morphology and genomic DNA.
Invertebr. Biol. 118, 149–164.
McGowan, J.A., Cayan, D.R., Dorman, L.M., 1998. Climate-ocean variability and
ecosystem response in the northeast Pacific. Science 281, 210–217.
Menge, B.A., 1992. Community regulation: under what conditions are bottom-up
factors important on rocky shores? Ecology 73, 755–765.
Menge, B.A., 2000. Recruitment vs. post-recruitment processes as determinants of barnacle
population abundance on Oregon rocky shores. Ecol. Monogr. 70, 265–288.
Menge, B.A., 2003. The overriding importance of environmental context in determining
the outcome of species deletion experiments. In: Karieva, P., Levin, S.A. (Eds.), The
Importance of Species: Perspectives on Expendability and Triage. Princeton
University Press, Princeton, NJ, pp. 16–43.
Menge, B.A., Sutherland, J.P., 1987. Community regulation: variation in disturbance,
competition, and predation in relation to environmental stress and recruitment.
Am. Nat. 130, 730–757.
Menge, B.A., Daley, B.A., Lubchenco, J., Sanford, E., Dahlhoff, E., Halpin, P.M., Hudson, G.,
Burnaford, J.L., 1999. Top–down and bottom–up regulation of New Zealand rocky
intertidal communities. Ecol. Monogr. 69, 297–330.
Menge, B.A., Chan, F., Nielsen, K.J., Di Lorenzo, E., Lubchenco, J., 2009. Climatic variation
alters supply-side ecology: impact of climate patterns on phytoplankton and
mussel recruitment. Ecol. Monogr. 79, 379–395.
Menge, B.A., Foley, M., Pamplin, J., Murphy, G., Pennington, C., 2010a. Supply-side ecology,
barnacle recruitment, and rocky intertidal community dynamics: do settlement
surface and limpet disturbance matter? J. Exp. Mar. Biol. Ecol. 392, 160–175.
Menge, B.A., Hacker, S. D., Freidenburg, T.L., Lubchenco, J., Craig, R., Rilov, G., Noble, M.,
Richmond, E., 2010b. Potential impact of climate-related changes on a rocky intertidal
ecosystem is buffered by differential responses to recruitment and species interactions.
unpublished MS.
Morgan, S.G., Fisher, J.L., Miller, S.H., McAfee, S.T., Largier, J.L., 2009. Nearshore larval retention
in a region of strong upwelling and recruitment limitation. Ecology 90, 3489–3502.
249
Navarrete, S.A., 1996. Variable predation: effects of whelks on a mid-intertidal
successional community. Ecol. Monogr. 66, 301–321.
Navarrete, S.A., Castilla, J.C., 1990. Barnacle walls as mediators of intertidal mussel
recruitment: effects of patch size on the utilization of space. Mar. Ecol. Prog. Ser. 68,
113–119.
Paine, R.T., 1966. Food web complexity and species diversity. Am. Nat. 100, 65–75.
Paine, R.T., 1974. Intertidal community structure: experimental studies on the
relationship between a dominant competitor and its principal predator. Oecologia
(Berl.) 93–120.
Paine, R.T., 1981. Barnacle ecology: is competition important? Paleobiology 7, 553–560.
Peterson, W.T., Schwing, F.B., 2003. A new climate regime in northeast Pacific
ecosystems. Geophys. Res. Lett. 30, 1896. doi:10.1029/2003GL017528.
Philippart, C.J.M., Anadón, R., Danovaroc, R., Dippnerd, J.W., Drinkwater, K.F., Hawkins,
S.J., Oguz, T., O'Sullivan, G., Reid, P.C., 2011. Impacts of climate change on European
marine ecosystems: Observations, expectations and indicators. J. Exp. Mar. Biol. Ecol.
400, 52–69 (this issue).
Pineda, J., 1991. Predictable upwelling and the shoreward transport of planktonic larvae
by internal tidal bores. Science 253, 548–551.
Pineda, J., 1999. Circulation and larval distribution in internal tidal bore warm fronts.
Limnol. Oceanogr. 44, 1400–1414.
Pineda, J., 2000. Linking larval settlement to larval transport: assumptions, potentials,
and pitfalls. Oceanogr. East Pac. 1, 84–105.
Poloczanska, E.S., Hawkins, S.J., Southward, A.J., Burrows, M.T., 2008. Modelling the response
of populations of competing species to climate change. Ecology 89, 3138–3149.
Rilov, G., Dudas, S.E., Menge, B.A., Grantham, B.A., Lubchenco, J., Schiel, D.R., 2008. The
surf zone: a semi-permeable barrier to onshore recruitment of invertebrate larvae?
J. Exp. Mar. Biol. Ecol. 361, 59–74.
Rouyer, T., Fromentin, J.M., Menard, F., Cazelles, B., Briand, K., Pianet, R., Planque, B.,
Stenseth, N.C., 2008a. Complex interplays among population dynamics,
environmental forcing, and exploitation in fisheries. Proc. Natl Acad. Sci. 105,
5420–5425.
Rouyer, T., Fromentin, J., Stenseth, N.C., Cazelles, B., 2008b. Analysing multiple time series
and extending significance testing in wavelet analysis. Mar. Ecol. Prog. Ser. 359, 11–23.
SAS Institute Inc, 2008. JMP ® 8 User Guide. SAS Institute Inc., Cary, NC.
Schiel, D.R., 2011. Biogeographic patterns and long-term changes on New Zealand coastal
reefs: non-trophic cascades from diffuse and local impacts. J. Exp. Mar. Biol. Ecol. 400,
33–51 (this issue).
Schoch, G.C., Menge, B.A., Allison, G., Kavanaugh, M., Thompson, S.A., Wood, S.A., 2006.
Fifteen degrees of separation: latitudinal gradients of rocky intertidal biota along
the California Current. Limnol. Oceanogr. 51, 2564–2585.
Schwing, F.B., Bond, N.A., Bograd, S.J., Mitchell, B.G., Alexander, S., Mantua, N.J., 2006.
Delayed coastal upwelling along the U. S. West Coast in 2005: a historical
perspective. Geophys. Res. Lett. 33, L22S01. doi:10.1029/2006GL026911.
Shanks, A.L., 2009. Barnacle settlement versus recruitment as indicators of larval
delivery. II. Time-series analysis and hypothesized delivery mechanisms. Mar. Ecol.
Prog. Ser. 385, 217–226.
Shanks, A.L., Brink, L., 2005. Upwelling, downwelling, and cross-shelf transport of
bivalve larvae: test of a hypothesis. Mar. Ecol. Prog. Ser. 302, 1–12.
Shanks, A.L., Shearman, R.K., 2009. Paradigm lost? Cross-shelf distributions of intertidal
invertebrate larvae are unaffected by upwelling or downwelling. Mar. Ecol. Prog.
Ser. 385, 189–204.
Shanks, A.L., Morgan, S.G., MacMahan, J., Reniers, Ad.J.H.M., 2010. Surf zone physical
and morphological regime as determinants of temporal and spatial variation in
larval recruitment. J. Exp. Mar. Biol. Ecol. 392, 140–150.
Southward, A.J., Hawkins, S.J., Burrows, M.T., 1995. Seventy years' observations of
changes in distribution and abundance of zooplankton and intertidal organisms in
the western English Channel in relation to rising sea temperature. J. Therm. Biol 20,
127–155.
Steele, M.A., 1997. The relative importance of processes affecting recruitment of two
temperate reef fishes. Ecology 78, 129–145.
Strathmann, M.F., 1987. Reproduction and Development of Marine Invertebrates of the
Northern Pacific Coast. University of Washington Press, Seattle.
Swearer, S., Thorrold, S., Shima, J., 2002. Evidence for self-recruitment in benthic marine
populations. Bull. Mar. Sci. 70, 251–272.
Sydeman, W.J., Bradley, R.W., Warzybok, P., Abraham, C.L., Jahncke, J., Hyrenbach, K.D.,
Kousky, V., Hipfner, J.M., Ohman, M.D., 2006. Planktivorous aukley Ptychoramphus
aleuticus responses to ocean climate, 2005: unusual atmospheric blocking?
Geophys. Res. Lett. 33, L22S09. doi:10.1029/2006GL026736.
Tilman, D., 1997. Community invasibility, recruitment limitation, and grassland
biodiversity. Ecology 78, 81–92.
Torrence, C., Compo, G.P., 1998. A practical guide to wavelet analysis. Bull. Am.
Meteorol. Soc. 79, 61–78.
Underwood, A.J., Denley, E.J., 1984. Paradigms, explanations, and generalizations in models
for the structure of intertidal communities on rocky shores. In: Strong, D.R., Simberloff,
D., Abele, L.G., Thistle, A.B. (Eds.), Ecological Communities: Conceptual Issues and the
Evidence. Princeton University Press, Princeton, N. J., pp. 151–180.
Warner, R.R., Cowen, R.K., 2002. Local retention of production in marine populations:
evidence, mechanisms, and consequences. Bull. Mar. Sci. 70, 245–249.
Witman, J.D., Brandt, M., Smith, F., 2010. Coupling between subtidal prey and
consumers along a mesoscale upwelling gradient in the Galapagos Islands. Ecol.
Monogr. 80, 153–177.
Woodson, C.B., Eerkes-Medrano, D.I., Flores-Morales, A., Foley, M.M., Henkel, S.K.,
Hessing-Lewis, M., Jacinto, D., Needles, L., Nishizaki, M.T., O'Leary, J., Ostrander, C.E.,
Pespeni, M., Schwager, K.B., Tyburczy, J.A., Weersing, K.A., Kirincich, A.R., Barth, J.A.,
McManus, M.A., Washburn, L., 2007. Local diurnal upwelling driven by sea breezes
in northern Monterey Bay. Cont. Shelf Res. 27, 2289–2302.