Deep ocean communities impacted by changing climate over 24 y in the abyssal northeast Pacific Ocean

Results

Surface ocean chlorophyll a, net primary production (NPP), and export flux (EF) estimated from satellite ocean color over a 100-km-radius circle around Station M (SI Materials and Methods), exhibited a distinct seasonal pattern that was highest in summer and fall (Fig. 1 A–C). Over the measured period from 1997 through 2012, average NPP was significantly higher from 2003 to 2012 than in years prior (Mann–Whitney U test, P < 0.001), with peaks more prevalent from 2006 onward. EF estimates were higher in 2012 than in any previous year (Fig. 1C). Sinking particulate organic carbon (POC) flux exhibited seasonal, interannual, and episodic rate variations over the entire time series from 1989 through autumn 2012 (Fig. 1D). Two major peaks in POC flux occurred over the last 18 mo of the time series, one from June to August 2011 and another between March and May 2012. The peak POC flux in spring/summer 2011 was the highest recorded over the 24-y time series (Fig. 1D), with the predominant constituent being the diatom Rhizosolenia (Rhizosolenia setigera and Rhizosolenia styliformis). The peak in POC flux in spring 2012 (Fig. 1D) included a relatively high abundance of tunics and fecal material from salps, Salpa spp., which were tentatively identified from high-definition video in the upper 600 m of the water column with a remotely operated vehicle in June 2012 (SI Materials and Methods). This spring peak was soon followed by another major sedimentation event to the sea floor in September 2012, which consisted of a combination of salp fecal pellets and phytodetritus (Fig. 1E). This pulse corresponded with strong peaks in surface chlorophyll, NPP, and EF (Fig. 1 A–C).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Long time-series measurements of surface-ocean and abyssal processes from July 1989 through November 2012 at Station M in the northeast Pacific Ocean on a monthly basis. Blue bars highlight the timing of peaks in NPP from 2010 to 2012. Dashed lines on x axes indicate data gaps longer than 1 mo. (A) Chlorophyll a concentration within 100-km-radius circle around Station M. (B) Net primary production within 100-km-radius circle around Station M. (C) Satellite estimated EF in a 100-km-radius circle around Station M, calculated for a nominal depth of 100 m. (Dataset S1). (D) POC flux. (E) Phytodetrital (black) and salp (red) aggregate percent cover on the sea floor (daily averages). Benthic Rover fluorescence images relative brightness value (dashed line). (F) Monthly averages of SCOC measured seasonally from 1989 until 2011 and daily from 2011 through 2012 (oxygen consumption has been converted to milligrams of carbon assuming a respiratory quotient of 0.85).

The daily presence of detrital aggregates on the sea floor did not exceed 15% coverage over the period from 1990 to 2007. Detrital aggregates showed distinct seasonal variation, reaching the highest peaks in the summer of 2002 (< 15% cover), fall of 1994 (< 14% cover), and summer and fall of 2001 (<8% cover) (Fig. 1E). The highest sea-floor coverage by detrital aggregates measured throughout the 24-y time series occurred between March and August 2012, when salp detritus ranged from <1% cover in early March to a high of 98% cover on 1 July (Fig. 1E). This was the only measurable deposition event of salps observed during the entire time series. Following this salp pulse, phytodetrital aggregates combined with some salp detritus formed another major deposition event beginning in late August and peaking in mid-September. This pulse covered up to 61% of the sea floor (Fig. 1E), the largest primarily phytodetrital aggregate peak recorded during the time series. Highs in relative sea-floor fluorescence accompanied both of these depositional events (Fig. 1E). The pulse of Rhizosolenia detected by the sediment traps in 2011 was not apparent in the time-lapse photographs because there were no large aggregations with dissimilar coloration from the background sediment.

Once on the sea floor, the phytodetritus and salp detritus became a food source for the organisms living in and on the surface sediments and overlying water. The utilization of organic carbon on the sea floor, estimated from measurements of sediment community oxygen consumption (SCOC), exhibited a strong periodicity, highest in summer and fall and lowest in winter months (Fig. 1F). Over the entire time series, SCOC fluctuated substantially from year to year in synchrony with POC flux but with less variance. However, more recently there was a major increase in SCOC, reaching a high of 25 mg C m⁻²⋅d⁻¹ and coinciding with the peak in the deposition of salp detritus in spring 2012 (Fig. 1 E and F). This peak was 54% higher than any pre-2011 measure of SCOC. SCOC decreased after the salp aggregate peak, only to increase again corresponding to a secondary peak in September of a detrital aggregate deposition that clogged the sediment traps (Fig. 1 E and F), although not eliciting the same magnitude in response. There was no increase in SCOC during the major phytodetrital (Rhizosolenia) peak in POC flux the previous year (2011). However, with only seasonal SCOC measurements until November 2011, any increase might not have been detected. With SCOC measured daily starting in November 2011, it was possible to document the rapid response of SCOC to the pulses of salps or detrital aggregates on the sea floor in 2012 (Fig. 1 E and F and Fig. S1). SCOC spiked from 12 mg C m⁻²⋅d⁻¹ in February 2012 to its peak of 25 mg C m⁻²⋅d⁻¹ in June 2012, and then declined to 15 mg C m⁻²⋅d⁻¹ 2 mo later.

We compared the organic carbon supply as POC flux to the estimated utilization by the sediment community (SCOC). The highest POC fluxes and SCOC occurred in the last 2 y of the time series (Fig. 2A). The ratio of POC flux to SCOC increased above unity nine-times over the time series (n = 55 monthly estimates), four of which occurred in 2011 and 2012. POC flux:SCOC reached a high of 1.6 in spring 2012 (Fig. 2B). These results suggest that food utilization by the benthic community at Station M tracks POC flux (Spearman-rank correlation, r_s = 0.67, P < 0.001) (Fig. S2). However, the hypothesis that there is insufficient food to sustain the benthic community at Station M on multiyear to decadal time scales still cannot be rejected based on these most recent POC flux and SCOC measurements alone.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Comparisons of food supply as POC and sea floor AOC with food utilization by the benthic community (SCOC) at Station M from July 1989 through November 2012. Dashed lines on x axes indicate data gaps exceeding 1 mo. (A) POC flux (line) and SCOC (dots). (B) Ratio of POC flux to SCOC. (C) Monthly estimated organic carbon flux of detrital aggregates (AOC) on the seafloor expressed in milligrams of C m⁻²⋅d⁻¹. Phytodetrital AOC flux (black lines) minimum (dashed line) and maximum (solid line) based on the “August” model and the “September” model of Smith et al. using residence times of 45 h and 95 h. (16) (SI Materials and Methods). Salp AOC flux (red lines) maximum (solid line) using a residence time of 109 h, and minimum (dashed line) using a residence time of 365 h. (D) COC flux, the higher of POC flux or AOC flux for each time period, maximum (black line) and minimum (red line) estimates. (E) The ratio of COC flux minimum estimates [(salp + AOC fluxes) or POC] to SCOC (triangles), the ratio of COC flux maximum estimates [(salp + AOC fluxes) or POC] to SCOC (circles), dashed line indicates unity (COC flux:SCOC = 1).

A more comprehensive view of carbon supply to and utilization on the sea floor considers input from detrital aggregates as well as POC flux, because POC flux might not include larger aggregates or salp tunics blocked by the small apertures in the sediment trap baffle (SI Materials and Methods). To calculate aggregate organic carbon (AOC) flux, the rate at which detrital aggregates supply organic carbon to sediments, we converted the measured coverage of detrital aggregates on the sea floor (Fig. 1E) to carbon content per area coverage, combined with a visible decay rate component. Separate observation-based conversions were generated for salp and phytodetrital aggregates, respectively (16, 17) (Fig. 2C and SI Materials and Methods). Combined phytodetrital and salp AOC fluxes ranged from ≤30 (minimum estimate) to 99 (maximum estimate) mg C m⁻²⋅d⁻¹ in June 2012 during the salp deposition and up to 81 (minimum estimate) to 318 (maximum estimate) mg C m⁻²⋅d⁻¹ during the subsequent phytodetrital aggregate deposition in September 2012 (Fig. 2C). Because AOC flux measured from photographs and POC flux measured with sediment traps are simply alternative methods of measuring POC arriving on the sea floor, we chose the highest estimate from either method (SI Materials and Methods) and defined this value as the combined organic carbon (COC) flux (Fig. 2D).

Recent pulses of high COC flux at Station M exceeded utilization (SCOC) at the greatest ratios measured in 24 y (Fig. 2E). The monthly COC:SCOC peaked at 4.1 (minimum estimate) to 16.1 (maximum estimate) in September 2012, suggesting that a single monthly surplus alone could supply food to sediment communities for many months or even in excess of a year. The average COC:SCOC for the majority of the time series (1989–2007) was 0.57, whereas the 2008–2012 COC:SCOC averaged 2.5 using the maximum COC estimate and 1.1 using the minimum COC estimate. The combined magnitude of the COC flux suggests that the deficit seen in food supply during the earlier portion of the time series could be met by episodic pulses of particulate organic matter, such as those in 2011 and 2012. Based on these recent large food pulses to the seafloor, we would reject the original hypothesis that there is insufficient food supply to meet the benthic community utilization over the 24-y time series. However, this interpretation infers that there were previous pulses of high food supply before our study or possibly missed during periods of no data collection.