Close Quarters: Redtail and Silver Surfperch Share the Surf

Redtail surfperch, Amphistichus rhodoterus
Redtail surfperch, Amphistichus rhodoterus

Redtail surfperch, Amphistichus rhodoterus, and silver surfperch, Hyperprosopon ellipticum, can be found year round on almost any exposed sandy beach between from the Long Beach Penninnsula, Washington to Point Reyes, California USA.   In the surf-swept intertidal of this region, they are the most common members of the Amphistichinae, a six-species clade of surfperches in the family Embiotocidae.   You can see the numbers for yourself in And the King of the Surf is…   If my mention of the Amphistichinae stirs a phylogenetic chord, you can learn more this little radiation by reading our paper on the molecular phylogeny of the Amphistichinae.


We wrote the paper because we were interested in why we see so many similar, apparently closely related, species living in what, though vast, seems to be a uniform, rather featureless sandy-bottomed environment.   Before we could tackle that question we wanted to understand a little more about connections within the group; so it was a starting point.    Since then, focusing on redtails and silvers, we showed, in Surfperches Share Sandy Shores, that the sandy beaches may not be as featureless to surfperches as they appear to a human.   Some beaches have a high proportion of silvers; others don’t.   At a large scale, there is some partitioning going on.   What about at a small scale?

Silver surfperch, Hyperprosopn ellipticum
Silver surfperch, Hyperprosopon ellipticum

One of the ways we sample surfperch is by angling with hook and line while walking along the beach.   We sample continuously for an hour, take a break if we need one, and then continue sampling, moving along the beach.   Schools of surf perch are on the move too.   Thus, I look at these hour-long samples as more or less independent.   Without question, it’s a loose interpretation, but it allows me to look at things on the temporal and spatial scale I am interested in.   Here, I’m interested in whether the two most common surfperch, redtails and silvers, live together or avoid each other at small spatial scales in the surf.
The chart above shows the catch at two beaches I sampled most in 2013; 44 1-hr samples at beach 1, and 18 1-hr samples at beach 2.   The x-axis shows the all possible outcomes for a sample: none = no fish caught; rt only = redtails only; sil only = silvers only; and both = both species were taken in the sample.   The y-axis is the number of 1-hr samples.

The most common outcome at beach 1 is redtails without silvers.   Redtails with silvers is the most common outcome at beach 2.   The rarest outcome at both beaches is silvers only; I didn’t see it in 44 samples at beach 1 and only twice in 18 samples at beach 2.   I’m interested in whether the two species are found together in my samples more often, or less often, than we expect from chance alone.   We can explore this by examining the probability of encountering each each species in a sample.   I’ll use beach 1 as an example: redtails were encountered alone in 21 samples and with silvers in 11 samples.   Combined, that’s 32 samples with redtails out of 44 total samples (11 samples turned up neither species).   Thus, I encountered retails in 73% of the samples.   For silvers it works the same way.  They were not encountered alone, but they were encountered with redtails in 11 of the 44 samples; that’s 25% of the samples.

Assuming redtails and silvers are operating independently, the probability of encountering both species in a sample is: P(redtails and silvers) = P(redtails) × P(silvers).

For beach 1 then, we might expect the probability of a sample containing redtails and silvers to be .73 x .25 = .18.   I have already mentioned, the actual rate of co-occurrence was .25.   So, at beach 1, the two species co-occur in my samples slightly more often than expected.   For beach 2, P(redtails) = .72 (13/18 samples) and P(silvers) = .56 (10/18 samples).   P(redtails and silvers) is .72 x .56 = .40.  The actual rate of co-occurrence was .44, just about what we would expect under independence.

This probably isn’t the best way to estimate the expected probabilities, but I don’t know a better way.   If you know of one, please let me know.   In any case, I don’t see anything suggesting obvious partitioning.   A final caveat is that we have to be kind of careful about taking even this seemingly safe conclusion too far because the sampling method didn’t allow me to learn anything about the mechanisms causing the pattern.   Until somebody does it better, however, I’m going with the idea that, in the surf zone, redtails and silvers don’t segregate at small spatial and temporal scales.

Speaking of somebody doing it better, while I was preparing for this post, I read a great paper on co-occurence patterns by Darryl MacKinzie, Larissa Bailey and Jim Nichols.   If you are thinking about embarking on your own investigation of species co-occurrence patterns, their paper is a great place to start.


MacKenzie, D. I., L. L. Bailey, and J. D. Nichols. 2004. Investigating species co-occurrence patterns when species are detected imperfectly. J. Animal Ecol. 73:546-555.

Westphal, M. F., S. R. Morey, J. C. Uyeda, and T. J. Morgan. 2011. Molecular phylogeny of the subfamily Amphistichinae (Teleostei: Embiotocidae) reveals parallel divergent evolution of red pigmentation in two rapidly evolving lineages of sand-dwelling surfperch. J. Fish Biol. 79(2):313-330.

1 Comment

  1. Steve, this is classic ecology done well. I would be interested in seeing a discussion of any & all the confounding factors that people might throw at this. Can you factor in state of the tide as a covariate? And time of year, etc. To see if a pattern emerges. It will knock down your power but you’re not rejecting the null result, anyway. And the worst that could happen is that that you have to do some more sampling 🙂

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s