The above sections led us to suggest that, first, octopus middens are consistent with octopuses acting as rate-maximizing foragers; and second, octopus response to relocation depends on habitat quality. With these initial conclusions in hand, can we now correlate prey availability and the abundance of octopuses as a rate-maximizing predator showing aggregative response to prey?
This straightforward prospect does not hold up to scrutiny. While HQI4 and octopus density were correlated with each other, neither showed significant correlation with any aspect of midden content, including diversity of items or measures of foraging success (number of items, summed length of items). Characters used in calculating HQI may be related to cover for octopuses (boulders and substrate), food availability (prey presence or absence), or both (kelp, depth, substrate), but other interpretations are possible. For example, kelp may serve as cover, as a food resource aggregating prey populations, or cover to potential prey, or may alter octopus and prey larval recruitment (Gaines and Roughgarden, 1987). While the correlation of these characters with octopus counts has previously been established (Scheel, 2002), little data exist to test mechanisms causing the correlation.
Nor did octopus population trends mirror prey population trends as measured in live prey surveys over the course of this study. Octopus densities increased and then declined over time (Fig. 20.1), while major prey densities increased throughout the study (Fig. 20.5). Furthermore, of the three predictions formulated above, the data allow us to reject both rate-maximizing forager alternatives: midden data was stable in the face of changing prey availability. Nor did octopus count parallel habitat productivity as reflected in live prey abundance. We are thus left with the alternative of octopuses as risk- or time-minimizing foragers. This behavior is consistent with the importance of shelter in octopus ecology. However, a more sophisticated approach to formulating hypotheses about foraging and habitat selection is needed.
Patch (habitat) selection models include the trade-off of foraging gain rate against (1) metabolic, (2) predation, and (3) missed-opportunity costs of foraging (Charnov, 1976; Brown, 1988). While foraging, octopuses may select prey as rate-maximizing foragers (as suggested by their selectivity for larger prey items and species). Metabolic costs are unlikely to vary among habitats. However, movement among foraging patches (Charnov, 1976) may be minimized due to high predation risk while in poor HQI4 patches, which are characterized by low shelter and greater depth. Octopus life history reflects specialization on risk-minimization strategies (Hanlon and Messenger, 1996). Octopuses display diverse cryptic, deimatic, and protean escape behaviors1. Other risk-minimizing strategies include use of dens for shelter, and transport of captured prey to the den for handling and consumption. Few data are available to examine perceived predation risk among octopuses, but within- and between-patch travel rates while foraging in habitats with varying cover could be used to examine active management of risk by octopuses. Missed opportunities costs potentially include those for reproductive activities, social activities, grooming and physiological maintenance, and den construction or maintenance. Octopuses increasingly forego foraging at the onset of maturity and reproductive activities, so that trade-off is managed by life history and unlikely to affect the submature juvenile octopuses in this study. Octopuses are solitary, and, hence, social
1 Deimatic displays function to threaten, startle or frighten, from Deimos, Greek god of terror; while protean behaviors involve changing shape, from Proteus, a shape-changing sea-god.
activities are not important. Finally, grooming and den maintenance are both performed at the den and seem to occupy relatively little time, although the role of sleep in octopus is still uncertain. These trade-offs are promising avenues for further research, but at this time seem likely to be of minor importance to octopuses relative to foraging success and risk from predators.
Patch selection by octopuses should be modeled as a function of trade-off between foraging success within depletable patches (gain rate) and predation risk. The need for this type of model is emphasized by the following considerations. First, the lack of correlation between octopus counts or HQI4, and any aspect of midden content suggests that prey abundance be excluded as a determinant of habitat selection. Second, if so, of the habitat characteristics measured (depth, substrate, boulders, kelp cover), only kelp cover varies from year-to-year. Third, we do not have data to exclude the hypothesis that kelp cover may directly affect predation rates upon octopuses; however, kelp effects on prey densities or perception of predation risk were not suggested by our analyses (i.e., measured prey density did not significantly correlate with octopus counts; nor did foraging success rate measured from midden contents, which is expected to correlate with foraging effort).
Testing of this proposed model would require detailed data on within- and between-patch movement rates, and characteristics of habitat through which octopuses forage, as well as perceived predation risk estimated as proportion of time spent in and associated with a den; prey encounter rate, measured as counts
of representative prey encountered per minute foraging; average energetic rate of gain, measured as energetic content of captured prey per minute foraging over the full duration of each foraging bout; and giving-up harvest rate (also a measure of perceived risk), measured as energetic content of captured prey per minute foraging at the end of each foraging bout. Giving-up harvest rate corresponds to a rate that just balances metabolic costs, predation costs, and missed-opportunity costs of foraging in a patch (Charnov, 1976; Brown, 1988).
Such data have been hard to collect in foraging studies, even for terrestrial organisms, and new techniques are required to obtain such data for mobile benthic marine organisms such as octopuses. Shallow, tropical water videography has shown some promise for octopuses (Forsythe and Hanlon, 1997). One approach under development by the authors for cold or deep water is the use of acoustic-positioning telemetry, already adapted for cephalopod studies (e.g., O'Dor et al., 1988; Rigby, 2004), to simultaneously track both a target animal and a submersible vehicle for collection of behavioral and habitat data as the target animal forages and seeks shelter (Shadow, Fig. 20.7). Telemetry has been successfully used for positioning information in cephalopod studies, and the addition of underwater data would substantially increase the utility of such studies in interpreting movement data.
Was this article helpful?