At what rate are herring eggs lost following spawn events, and what biological and physical factors are driving these rates?

Because herring are mobile and aggregate, estimating their population size can be very difficult. Consequently, once spawned, herring eggs are used to back calculate the number of reproductive adults that spawned them. Yet, egg counts can be compromised due to various sources of mortality or loss that occur between egg deposition and egg surveys.

To estimate the population abundance of herring, spawned eggs are measured every year with dive surveys, which typically consist of a one-time survey per spawn location. However, there is often a time lag (1-17 days) between spawn events and the dive team assessment when egg loss can occur due to wave action, predation and other biophysical factors. Egg loss prior to spawn surveys can result in underestimates of the actual number of eggs originally spawned, leading to underestimates of spawning population abundances.

Measuring herring egg loss rates after spawning is therefore critical. In fact it is a key number with high uncertainty in the current Department of Fisheries and Oceans (DFO) stock assessment model.  To understand egg loss, SFU Masters student Britt Keeling, under the direction of marine ecologists Margot Hessing-Lewis and Anne Salomon, conducted two linked studies.

In the spring of 2011, we conducted an experiment to determine the quantity and rate of herring roe consumption by marine predators. We based our experimental design on the traditional Spawn on Kelp (SOK) fishery by enclosing standardized pieces of spawn-on-kelp in predator exclusion cages that were suspended at different depths throughout the water column (see graphic). We then monitored egg abundance over a two-week period.. 

In 2012, we monitored herring egg loss following a spawn event for up to 22 days at 9 study sites within permanently placed quadrats. Divers returned to monitor the same quadrat every 3 to 5 days throughout the egg incubation period (18 to 22 days), visiting each site a total of 5 times. To determine potential herring egg predators, we also conducted dive surveys of fish and benthic invertebrates at all survey locations.


1)   Our 2012 observational dive surveys estimated egg loss rates from 59-75% over 6.8 days, which is the average lag time between spawn event and herring egg survey by DFO.  This estimate of egg loss rate is more than double the value currently used to estimate herring spawning stock biomass for Canada’s Pacific herring fishery. This means that a large proportion of eggs could be uncounted, which could lead to uncertainty in current stock assessments.

2)   Egg loss rates varied 4.5 fold among sites, and spawn area was found to be a key factor positively influencing egg loss rates. This high site-level variability is important to account for when estimating herring abundance and calculating uncertainty in stock assessments.

3)   Experimental evidence from the 2011 study suggests that both egg predation and habitat type are also strong drivers of egg loss with the majority of eggs being consumer on the sea floor compared to those suspended in the water column.


Photo: Experimental design used to test the effects of habitat and predation type on egg loss rate. Sixty sub-samples of Spawn-On-Kelp (SOK) were assigned to 4 predator treatments and submerged to 3 different habitats (subsurface, midwater, benthic i.e. sea floor). This design was replicated 5 times across a bay in Spiller Channel, near Bella Bella, BC. 


Photo: Dive team, Yago, Brittany, Margot and Ryan prepare for a herring egg survey.


How are rockfish affected by herring spawning events?


While it is generally understood that many marine species rely on herring spawn, much of the research related to this topic has focused on seabirds.   Linkages between herring spawn and subtidal species are much less understood. In 2012, Drs. Margot Hessing-Lewis and Anne Salomon focused research in Spiller Channel in the Central Coast to ask how rockfish were affected by herring spawning events. 


Stomach contents were identified and measured from Quillback and Copper Rockfish (Sebastes maliger and caurinus). Contents were grouped by category (herring roe, invertebrate, fish, other, and empty). Tissues were also analyzed for trends in nitrogen and carbon stable isotope values, which can track changes in food sources.




1)   We found that the percentage of fish in rockfish diets switched from 30% fish tissue before a spawn event to 34% herring roe after a spawn event, and that this shift peaked at 2-3 weeks post spawn. Thus, herring provide a subsidy to coastal food webs by moving from offshore to nearshore waters for annual spawning events.


2)   The dietary switch following spawn events is most pronounced in female rockfish, where stomach contents contained an average of 29% herring roe in the month following a spawn event. This may be due to differences in energy requirements for pregnant females. Chemical markers in fish tissues (muscle, heart, liver, gonads) also showed evidence that herring were eaten and were used to grow rockfish tissue.


This provides evidence that herring fuel the production of subtidal reef fish, many of which are commercially and culturally important to coastal communities along the northwest coast of North America. Specifically, herring are one of the principle means of transferring energy from tiny plankton to larger predatory fish, marine mammals, and seabirds.  Consequently, population fluctuations in herring and other forage fish can have extensive ripple effects across temperate marine ecosystems. Any ecosystem-based fisheries management approach in BC waters must consider the critical role of Pacific herring in fueling entire marine food webs.


Switch of diet in rockfish using herring spawn time

Dietary switch in rockfish spp. prey composition following a spawn event at one site in Spiller Channel (2011). 



Thursday, January 23, 2014