ANG 6930 – Hunter-Gatherers

K. E. Sassaman

Fall 2001

Class 3 – Subsistence Models

· As we have seen, the primary emphasis of evolutionary ecology in studies of hunter-gatherers is on subsistence

· Subsistence is regarded as proxy for reproductive success

· Longstanding assumption (without data) was that H-Gs relied most heavily on meat, and added plants to diet only as meat became short (either in evolutionary terms, with end of Ice Age, or in microeconomic terms, when game became depleted or otherwise unavailable)

· Then, with Man the Hunter symposium, emphasis shifted to importance of plants, with some data brought to bear which suggested plants were more economical than meat

· Either perspective tended to overlook tremendous range of variation in H-G diets worldwide

Range of subsistence variation

Kelly’s data on 123 H-G populations shows

· Hunting varies for 10 to 90%

· Gathering varies from 0 to 85%

· Fishing varies from 0 to 80%

· Only one-fourth of the 123 listed by Kelly (1995:Table 3.1) have diets consisting of more than 40% hunted resources (i.e., game)

(SEE GRAPHS)

· About half (n=38) of those with 40% or less hunted resources have diets consisting of 41% or more fish, if they fish at all (16 with no fish at all).

· Groups depending on plants for more than 40% of their diet total 36, only 29% of the total sample of 123 groups

How can we account for the general differences?

Kelley introduces two variables to show how H-G diet is systematically related to environmental variables

· Effective Temperature (ET): quotient derived from the mean temperatures of the coldest and warmest months of the year

· Primary production (PP): annual net above-ground plant production (g/m2/yr), a product of effective precipitation and solar radiation

· Through multiple linear regression, Kelly suggests that ET and PP can predict account for about 56% of the variation in percent of plant food in diet

· These variables do not, however, predict the predict the percentage of hunted resources in diet

· He goes on to demonstrate that fishing and trade account for the lack of correlation (see Figure 3-1).

Optimal-Foraging Models

All involve:

· a goal, usually maximizing foraging efficiency (maximize food gathered per unit of time)

· a currency, usually energy (calories)

· a set of constraints, such as maximum amount of time available to forage; number of able-bodied foragers

· and a set of options, a suite of food choices and ways to go about getting them and using them

Variants include linear programming, diet breadth model, and patch choice model, all nicely summarized by Kelly (1995:73-97)

Let’s take a close look at two classic papers on the subject: Lee 1969 and Hawkes et al. 1982

!Kung Bushman Subsistence: An Input-Output Analysis (1969), Richard B. Lee

Not an optimal foraging model per se, but consistent with the tenets of OPT and clearly an example of integrating ecological and economic theory & method

Frames the study as an evolutionary one in that Lee considers the !Kung to have an “elementary” form of economy: where the relation between production and consumption is immediate in space and time (domestic economy)

“extreme isolation and a marginal environment have been responsible for the persistence of this form” (Lee 1969:50)

Lee provides census data on composition of domestic units, noting particularly the ratio of effective producers to dependents

Percent of effective producers ranges from 33.3 to 90.9%

(14 “living units” ranging from 9 to 29 each, total of 248 people; effectives are 61.3% of total)

Northern Kalahari hot, wet summer (Nov. to March), a cool, dry winter (April to August, and hot, dry spring (September-October)

Distribution of water is the most critical limiting factor affecting food choice.

Eight permanent water holes

Eighty-five plant species and 54 animal species were listed by the !Kung as edible

!Kung ranked their 85 edible plants on basis of taste, nutritional value, abundance, and ease of collecting; mongongo is number one ranked plant food

Given the tethering effect of water, the !Kung prefer to collect and eat the highest ranked foods that are available at the least distance from permanent water

During dry season, the !Kung establish more-or-less permanent camps at permanent water and “eat their way out of it.”

(SEE COST CURVE, p. 60).

When roundtrip distance exceeds 12 miles (threshold of overnight trip), the !Kung begin to add less desirable, but closer resources (such as bitter melon, acacia gum, and heart of ivory palm)

During wet season, they move camp once the 6-mile-round-trip threshold, never having to add the less desirable resources

INPUT-OUTPUT ANALYSIS

Work Effort (Input): calculated the total man-days of work and divided this by total man-days of consumption to come up with Index of Subsistence Effort (ranged from .11 to .31 over the four weeks of data collection. Turns out to be relatively “light” work load, with per capita effort at about .23, that is, 23 days of work for 100 days on consumption

Caloric Levels (Output): with such a light work load, is energy output adequate?

Collected data on three major resources: mongongo (33% of diet), meat (37% of diet), and other plant foods (30% of diet)

Total per capita caloric output for these three resources is 2140 calories

Average daily caloric requirements for !Kung estimated at 1975 calories

Thus, output is more than adequate

These results became basis for Sahlins “Original Affluent Society”

Why Hunters Gather: Optimal Foraging and the Ache of Eastern Paraguay (1982) by Kristen Hawkes, Kim Hill, and James F. O’Connell

Frames article as debate on the dietary worth of meat

Lee arguing that it is added to diet only when plant foods are inadequate

Harris arguing that meat resources are the more efficient, and that plants were added only when game became scarce

Ache have a high proportion of meat in diet

They hunt 33 species of game, and also eat at least 10 species of reptile and amphibians, more than 15 species of fish, and lots of birds. They collect more than 40 species of plants, the most important being the palm.

Hawkes and Hill followed Ache on 7 foraging trips covering 61 foraging days and 58 hunting days. Totaled 1570 consumer days, 457 women-gathering days, 794 man-gathering days, and 674 man-hunting days.

The average daily per capita intake of calories during the 7 trips was 3610 cal. Eighty percent of this came from game animals. Given the high return, it would seem that emphasis on game is profitable. So why gather plants?

Optimality

Resources can be ranked according to the ratio of returns they provide (calories) to the cost (handling time) or acquiring and processing them (once they have been encountered; we’ll deal with encounter issues in patch choice model)

Model suggest that returns will be maximized if foragers take those resources for which this ratio is equal to or higher than the average returns they get for foraging in general and if they ignore all potential resources for which this ratio is lower than their average returns.

Resource _i is included if: E/T £ E_i/h_i

Where E = total calories acquired foraging

T = foraging time

E_i = calories available in a unit of resource j

h_i = handling time per unit of resource j

Resource is excluded if: E/T > E_a/h_a

Resources can be ranked according to ratio of calories to handling time; note that this says nothing about the actual percentage of items in diet; that is, a high-ranked resource may be encountered only rarely, but when they are, they are taken

Hawkes et al. rank the resources by E_i/h_i

(SEE TABLE 3, p. 389)

These ranks are then plotted against average returns for foraging in general, which was derived by total calories of top ranked foods (collared peccary [232 x 1950 = 452,400] and deer [300 x 819 = 245,700) and dividing by total foraging search time (3673 hours) plus 1024 hours carrying and 16 hours handling. This gives a calories/hour value of 148 for top ranked resources.

(SEE FIGURE 1, p. 390)

Continue with second-ranked resources, adding total calories from pacas and coatis are added to numerator, and handling time for these resources are added to denominator…(result is 405 cal/hr)… and so on….

Diet Breadth model: foragers adds encountered resources to the diet until the return rate for the nth resource (E_i/h_i ) and its handling cost is less than the overall return from foraging

The Ache should not take any resource with a return rate of below about 870 cal/hr (of course, this curve changes with different rates of encounter)

Note that model does not predict actual diet composition, only that resources in the optimal mix will be taken when encountered; if all encountered simultaneously, they are taken in rank

Problem: the data used to generate optimality curve is same used to generate unit costs/returns, so it is really just a description

In order to make this more useful, we would have to have data on foods that are edible but not taken, or….

Use the model to predict changes in diet breadth under different circumstances (really an archaeological tool)

Other major problem is the assumption of a fine-grained encounter (that resources are taken in proportion to occurrence and they are randomly distributed.

To overcome this last bias, the Patch-Choice Model is useful