related to body size, indicating that bottom-up forcing mechanisms - - PDF document

Linking phytoplankton phenology to pink salmon productivity along a north-south gradient Michael J. Malick1, Sean P. Cox1, Franz J. Mueter2, and Randall M. Peterman1

1 School of Resource and Environmental Management, Simon Fraser University, Burnaby, British

Columbia, Canada

2 School of Fisheries and Ocean Sciences, University of Alaska Fairbanks, Juneau, Alaska, USA

For Pacific salmon (Oncorhynchus spp.), the first year of ocean residence is widely viewed as a critical period that can strongly influence productivity (i.e., adult recruits per spawner) of a stock (Parker 1968). Evidence suggests that salmon mortality during the early marine life stage is inversely related to body size, indicating that bottom-up forcing mechanisms that affect prey resources may strongly influence stock productivity (McGurk 1996). Several bottom-up mechanisms, e.g., the optimal-stability window and match-mismatch hypotheses, have been proposed to explain productivity variation in marine fish stocks, including salmon (Cushing 1990; Gargett 1997). A central feature of these mechanisms is that they assume a strong link between phytoplankton dynamics (e.g., phenology or total biomass) and salmon productivity. Although lower-trophic-level processes such as initiation date of the spring bloom are correlated with yield and productivity of certain marine fish populations (Platt et al. 2003), relationships between phytoplankton dynamics and salmon productivity are largely untested beyond a few local-scale studies.

−160 −150 −140 −130 45 50 55 60

Longitude (°W) Latitude (°N)

• ●
• ●
• ●
• 1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 N 500 km at 55°N

• Fig. 1: Study area indicating the grid cells used to compute the bloom initiation

date and mean chl-a concentrations (green squares) and the locations of ocean-entry points for the pink salmon stocks (solid black dots).

In this study, we asked how well the spring bloom initiation date and average spring/summer chlorophyll- a (chl-a) concentrations (estimated using SeaWiFS and MODIS-Aqua satellite data) can explain spatial and inter-annual variabil- ity in productivity of 27 pink salmon (O. gorbuscha) stocks in the Northeast Pa- cific Ocean (Fig. 1). Es- tablishing a plausible mech- anistic link between phy- toplankton dynamics and salmon productivity first requires evidence that the two processes operate at similar spatial scales; that is, spatial covariation of lower-trophic-level processes should approximately match the spatial scale of covariation observed in the salmon productivity data that they are being used to explain (Koenig 1999). To determine the extent of spatial synchrony in (1) pink salmon stock productivities, (2) spring-bloom initiation dates, and (3) spring/summer chl-a concentrations, we fit smooth non- parametric covariance functions between pairwise correlation coefficients (which were computed between time series of productivity or between grid cells for the phytoplankton variables – Fig. 1) and the distance separating correlated grid cells or ocean-entry points of salmon stocks. The spatial 1

covariance functions were summarized using the 50% correlation scale, i.e., the distance at which the covariance function falls to 50% of its observed maximum value, which provides a useful metric

• f how much the correlation declines with increasing distance.

−1.0 −0.5 0.0 0.5 1.0

(a) Pink salmon productivity

• ●
• ●
• ●
• ●
• Correlation coefficient

500 1000 1500 2000 2500 3000 −1.0 −0.5 0.0 0.5 1.0

(b) Spring bloom initiation date

• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ● ●
• ● ● ●
• ●
• ●
• ●
• ●
• ●
• ●
• ● ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ● ● ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ● ● ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ●
• ● ●
• ●
• ●
• Distance (km)
• Fig. 2: Pairwise correlations for (a) salmon productivity,

and (b) spring bloom initiation date as a function of distance between location of data pairs. Solid curves represent the estimated smooth nonparametric covariance function with the 95% confidence band shown as grey shaded region. Solid vertical lines indicate the 50% correlation scale.

We then used hierarchical Ricker spawner- recruit models to estimate both regional and stock-specific effects of spring/summer chl-a and the spring bloom initiation date on pink salmon productivity, while also accounting for within-stock density-dependence. Because the chl-a and bloom initiation variables were mod- erately correlated, the hierarchical models were fit separately for the bloom initiation date and spring/summer chl-a. For both variables, we also fit the hierarchical models separately for a southern stock group (stocks 1-9 in Fig. 1) and a northern stock group (stocks 10-27 in Fig. 1), because exploratory analyses suggested poten- tial differences in the effects of these variables between northern and southern stock groupings. For both the spatial and hierarchical models, the analysis was limited to salmon ocean-entry years 1998-2010. We found that the extent of spatial syn- chrony for all three variables was quite similar, with positive spatial covariation being strongest at the regional scale (0-800 km) and no covari- ation beyond 1500 km (Fig. 2). The estimated 50% correlation scale for pink salmon was 261 km (95% CI = 148-628 km; Fig. 2a), which was slightly smaller than for the spring bloom initi- ation date (367 km; 95% CI = 235-776 km; Fig. 2b), although there was considerable overlap in confidence intervals between these two variables. For chl-a concentrations, covariation decreased steeply with increasing distance over spatial scales of 0-500 km for all months (April-September). The 50% correlation scale was highest (~380 to 430 km) in the spring (April-May) and declined to about 250 km during summer months (June-September), which was similar to the estimated 50% correlation scale for salmon productivity. Spring bloom timing had a significant region-wide effect on salmon productivity for both northern (Alaska) and southern (British Columbia) populations, although the regional effects were

• pposite in sign (Fig. 3). An early spring bloom was associated with higher productivity for northern

populations and lower productivity for southern populations. This result contrasts with those for spring/summer chl-a concentrations, where the regional effect for both northern and southern stock groupings was negative, implying reduced salmon productivity when chl-a concentrations are higher. Although it is not clear why a lower spring/summer chl-a concentration would be associated with higher salmon productivity, a plausible explanation is top-down grazing pressure of zooplankton 2

(a primary food resource for juvenile pink salmon) on phytoplankton standing stock. Regardless

• f the mechanism, the spring bloom initiation date was consistently a better predictor of salmon

productivity than mean chl-a concentration.

1 3 5 7 9 11 13 15 17 19 21 23 25 27 −0.2 −0.1 0.0 0.1 0.2

Spring bloom coefficient Stock number

• Fig. 3: Estimated effects of the spring bloom initiation date on pink salmon pro-

ductivity from the best-fit hierarchical model. Thick solid lines show the regional effect with 95% confidence band (grey shaded region), whereas open circles indicate stock-specific effects. Stock number corresponds to numbers in Fig. 1.

Our results suggest that the phenology of biological

• ceanographic

processes can strongly influence salmon productivity and that spring bloom phenol-

• gy is more important for

higher trophic level species such as pink salmon than the standing stock of phyto-

• plankton. This conclusion

has important implications as the climate warms. It is generally recognized that a warming climate will lead to an earlier onset of spring conditions, including earlier timing of peak zooplankton biomass and outmigration of pink salmon (Edwards and Richardson 2004). Phenological mismatches could occur across trophic levels if separate processes do not change in synchrony, potentially leading to northward latitudinal shifts in peak productivity of pink salmon stocks. Cushing, D.H. 1990. Plankton production and year-class strength in fish populations: An update of the match/mismatch hypothesis. Adv. Mar. Biol. 26: 249–293. Edwards, M., and Richardson, A.J. 2004. Impact of climate change on marine pelagic phenology and trophic mismatch. Nature 430: 881–884. Gargett, A.E. 1997. The optimal stability “window”: A mechanism underlying decadal fluctuations in North Pacific salmon stocks? Fish. Oceanogr. 6: 109–117. Koenig, W.D. 1999. Spatial autocorrelation of ecological phenomena. Trends Ecol. Evol. 14: 22–26. McGurk, M.D. 1996. Allometry of marine mortality of Pacific salmon. Fish. Bull. 94: 77–88. Parker, R.R. 1968. Marine mortality schedules of pink salmon of the Bella Coola River, central British Columbia. J. Fish. Res. Board Can. 25: 757–794. Platt, T., Fuentes-Yaco, C., and Frank, K.T. 2003. Spring algal bloom and larval fish survival. Nature 423: 398–399. Keywords: pink salmon; spring bloom; phytoplankton; productivity; spatial scales Preference: Oral presentation 3