Conclusions
The distribution of resources on the landscape have a modest effect on determining which camp mobility strategies are most effective. This empirically based model analysis shows that the ethnographically observed Ache pattern of extremely high mobility and randomly placed campsites would not be effective if relevant prey species were concentrated in a few habitat patches within the landscape.
But in moderately heterogeneous landscapes, like their current Mbaracayu home range, the Ache mobility strategy performs very well relative to alternatives. But the most surprising result is that much greater heterogeneity in resource distributions in the environment does not favor larger camp size in our model, nor does it change camp mobility very much (2.8 mean days in camp in a highly variable clumped environment vs. 1 day in camp for the natural and relatively homogeneous environment). We did discover, however that hunters would increase their mean return rates by about a 30 % in clumped heterogeneous environments by targeting the most productive habitats in camp moves. This is an important insight. Human foragers more efficient at exploiting patchy rather than homogenous environments because they can recognize and exploit the patchiness in a manner that reduces their search time in unproductive environments. While this might seem intuitively obvious, we were surprised at the magnitude of the increase in efficiency on patchy landscapes even when the biomass of prey on the landscape was the same.Ecological hypothesis about how resource distributions through time and space affect hunter-gatherer group size and mobility are generally accepted as fact due to their logical coherence and their origins in animal behavioral ecology where they are qualitatively supported. Nevertheless here we found minimal changes in optimal group size or mobility patterns as environments contained more dispersed or more clumped prey distributions.
The environmental differences we simulated were quite substantial, from very minor differences in prey abundance across 7 vegetation types to more than 10 fold differences in prey abundance across vegetation types. We also examined a wide range of patchiness, from many small habitat patches of only a few hectares to a landscape where 52 % the habitat was found in patches of greater than 1000 ha. Despite this variation the optimal group size under all conditions was 7 hunters per camp. This group size not only maximizes the mean expected hunting return rate, but it is far superior for both maximizing return rate and simultaneously minimizing the probability of no hunted game in camp on any particular day. The best return rates in the model come from group sizes in which the benefits of cooperative hunting are maximal relative to the costs encounter rate depression from frightening game in the same search area (i.e. feeding competition).The impact of increased patchiness on mobility is also very moderate. Under natural conditions return rates are maximized by moving every day of the year (congruent with ethnographic observation), but in our simulated highly variable and patchy environment, the best performing strategy is found to have a mean camp staying time of 2.1 days. It is important to note however that our model includes no costs of movement outside that of localized depletion if the hunters remain in an already hunted area.
Finally, when habitats are more heterogeneous and clumped, targeted movement, in which camps are located in the best habitats whenever possible, increases the hunting return rate by about 30 % over a pattern in which camps are simply moved in a random direction each day. Thus targeted moves in addition to adaptive camp movement, remaining at a camp spot until prey are depleted below some threshold, would allow simulated Ache hunters to increase their mean daily hunting return rate by 35 % in a more patchy and variable environment even when it contains exactly the same number of prey animals as their natural environment containing more dispersed prey. This is quite a substantial and biologically meaningful increase in food intake simply by exploiting patterned distribution of prey even when absolute densities do not change. Human foragers, by knowing the landscape and the spatial location of better habitats, and moving to facilitate hunting in those areas, gain a substantial advantage from that knowledge.
Acknowledgments Portions of this research were funded by NSF-IPG grant BCS-1138073.