In the first part of this series I established a hypothetical diet appropriate to the area I live (Mendocino County) and the culture (i.e., non-hunter-gatherers, based on familiar domestic foods).[i] Growing food requires land, water, fertilizer and energy resources and I want to know for a given diet + population do the resources exist? I am leading myself through the following series of steps to address that question:

(1) Establishing a diet, (2) Translate this diet into land area requirements, (3) Scale the land area from an individual level to the population of Mendocino County, and (4) Compare to the actual land-base.

As a review, the diet being considered for now is given below. Perhaps it will have to be reconsidered following the initial results, which is not difficult to do the way spreadsheets work.

Farmland Classification

Classification of farmland merits a discussion. Soil nomenclature and taxonomy is complex, but by most accounts USDA's Natural Resource Conservation Service's Land Capability Classes I and II are considered "prime agricultural farmland," meaning the soils are deep and fine enough to be tilled, not highly subject to erosion when disturbed, and not severely hampered by potential seasonal inundation.[ii] These soils tend to form where alluvial deposits build up layers of sand, silt and clay in more or less even proportions. Most of the crops in the diet designed here require such prime land. Land suited for grazing may include non-prime land, that is class III and above, but the productivity of this land class is lower. Tree crops, including fruits, olives and nuts, may also be sown on non-prime land but with lower yields. Yields on non-prime land can be improved by seeding with desired species, managing livestock smartly, and fertilizing.

Area Required per Person

I will begin by taking columns 1 and 2 from the table above and calculating how much area is required for this diet for one year, i.e., a per capita land requirement given the above diet. To do this, I must apply estimated yields per area for Mendocino County of the specific crops considered. The results are as follows:

The tricky part of this step is finding decent, contemporary information about crop and livestock productivity for Mendocino County. One issue is that the local ag business is currently dominated by a single, ethanol crop, rather than a diversity of products as in the past.[iii] Previously, I explored this issue with respect to grains, and was forced to use data several decades old.[iv]

Because projected grains yields dominate the area requirements, understanding the yields per acre figure used is most important. Grains being modeled are the small grains, chiefly wheat, oats, barley and rye. They all give similar yields per acre. California is a major producer of wheat, and the yields from most growers in the Central Valley are fantastically high-6000 lbs per acre is considered normal. Why then the 1000 lbs per acre for this worksheet? The resources required to achieve 6000 lbs per acre are substantial and include: pre-planting application of herbicide, precision seed drilling at high density but even spacing, application of fertilizer at time of planting, irrigation, and additional fertilizer application.[v] From my conversations with grain farmers and UC extension agents I expect the following changes in yield absent the given input: remove irrigation and yields fall by half (3000 lbs/acre), remove artificial fertilizer application and yields fall by nearly half again (1800 lbs/acre), remove pre-planting herbicide and precision planting and yields fall once more (1100 lbs/acre). Because 100 lbs are needed to sow an acre of grains, a 1100 lb harvest nets an edible yield of 1000 lbs. So my number assumes dry-land farming methods, lack of sophisticated planting equipment, and no herbicides or artificial fertilizers. 1000 lbs per acre is also the yield from historic data when Mendocino County did grow grains, and it could be argued that soils were less depleted then. This may be considered a worst-case scenario, but given current conditions in the County it may be reasonable. The capital stock of equipment to grow high yield grains is lacking and energy constraints may limit use of agro-chemicals and fertilizers as well as irrigation pumps. To avoid yield problems with staple foods requires planning ahead, considering what inputs may still work in an energy-constrained County, and investing today in infrastructure that could last in such an environment.

For olive oil I could find state-level information only and I assumed 3 tons of olives per acre with 40 gallons per ton oil extraction.[vi] Olive trees are becoming more common within vineyard operations and small-scale commercial production has increased lately, including local processing equipment.

As explained in part 1, I use honey as the sugar source so it doesn't have a direct land area requirement. It would be good the check with local bee keepers about how many hives they believe the county can support without seasonal transportation to almond orchards. My estimate is that one hive per person would cover the honey/sugar quota. Another option would be to grow sweet sorghum, which requires summer irrigation and prime farmland.

Meat and dairy yields are especially difficult to estimate because they rely on lands of variable quality and encompass a diversity of production models. As far as I can gather, livestock here are rotated between winter pasture in the hills and summer pasture in the valleys. Valley pastures include both prime and non-prime farmland. Stockman operations sell half-grown cattle out of the county once rangeland dries in the summer. Conceivably these exported animals could be eaten when not fully grown, as is done with male dairy cows. Cow-calf operators bring animals into summer pastures in the valleys. A single cutting of hay is typically given to valley pasture before animals are placed on it, which is an important supplement during late summer and fall. In addition to cattle, sheep and goats are raised, but cattle are preferred here-perhaps based on cultural norms or because sheep are more susceptible to predation. Animals are usually culled in the fall, reducing food requirements while the herds are re-established on new grass that emerges during the rainy season (Oct-May). The Mendocino County Department of Agriculture gives estimates of the productivity of different forms of pasture land in the unit of measure called AUM, or Annual Unit Month, which is the food required to feed a 1000 lb steer for one month (or ca. 5 sheep), which works out to about 1000 lbs of forage. [vii] The county crop report doesn't give hay yields, but discussion with local ranchers suggests 5 tons per acre is probably a good yield. County soil surveys also give estimates for the rangeland productivity in terms of above ground dry weight in lbs per acre.[viii] Values are typically about 2000 lbs per acre, which would provide 1000 lbs of forage, or enough for 33, 1000 lb cows for 1 day. [ix] But this is a standing biomass figure at a particular stage of growth and doesn't indicate how a pasture responds to grazing and then re-growth. Will an acre of rangeland with 2000 lbs of above ground biomass grazed by 33 cows down to 1000 lbs biomass re-grow to 2000 lbs in 2 weeks or 5 weeks? The answer likely depends upon the quality of the land, the weather, and the season.

I simply don't have the expertise to sort out all these variables from first principles. Another way to go about this is to use total weight of animals produced and divide this by rangeland and pasture land acreage. The average live weight for cattles+calves and sheep+lambs for Mendocino County in 2005 and 2006 was 10.7 million pounds.[x] The actual weight of meat consumed is about 40% of the live weight.[xi] This gives about 4.3 million lbs of meat produced in the county. There are no feedlots here and the county does produce a lot of hay, but I am not sure this represents the meat production of county land only. If we assume it does, however, the following calculation can be made: Take the 4.3 million pounds of meat and divide by the acres of range and pasture land in use (720,000 acres) according the county crop report, to yield 6 lbs of meat per acre.

Humans may consume milk in its liquid form or as cheese. The table computes in cheese units equal to about 1.6 cups (12.8 oz) of milk per day. Dairy cows need to be fed in close proximity to the milking barns, so they are kept on highly productive irrigated or naturally sub-irrigated pasture, on fog-swept coastal plains, fed grains, silage and hay or some combination. The 1400 milking cows in Mendocino County yielded, on a per cow basis, 6.4 gallons of milk per day in 2006, or 18,800 lbs of milk per year. From the table below, the land required to produce these yields can be estimated.[xii] If we assume irrigated pasture for the forage, Mendocino County irrigated pasture produces about 9200 lbs per year. Then the other area would come from grain yields of about 1000 lbs per acre. So, one milking cow requires (8096/9200) 0.88 acres for forage and (8275/1000) 8.28 acres of dry-farmed grain, for a total of 9.16 acres per cow. (The grain area could be cut about in half if irrigated). Since each cow produces 18,800 lbs of milk, this is 2052 lbs of milk per acre. Since cheese is 1/3 the weight of milk, this yields 684 lbs of cheese per acre. The grain area could be reduced by more extensive grazing on highly productive pasture, but this would generally be prime farmland anyway and may not reduce total area required.

For eggs I have a book called "Living with Chickens" that advises a quarter pound of grain per layer per day. My logic then went as follows: assume 200 eggs per hen/year, fed 90 lbs of grain/year/bird, so 90 lbs of grain yields 200 eggs, or 0.091 acres of grain per 200 eggs, with each egg weighing about 0.1 lbs. 200 eggs/0.091acres = 2192 eggs/acre at 0.1 lbs per egg = 220 lbs of eggs/acre. Even when chickens are pastured (which I prefer) they still require these grain inputs to get high egg yields. Note that old laying hens can go in stew pots but this contribution to the meat diet isn't included.

Fruit and vegetable production yields are in line with county records for tree crops and my own experience with intensive vegetable cultivation. These yields assume summer irrigation water.

Discussion of Initial Results

I am concerned by the total of 9 acres. From my previous readings, the number "about 1 acre per person" stuck in my head (about 0.4 hectares) as to what is required for a diet familiar to people in North America and Europe.[xiii] Of course, many countries are already well below 0.4 hectares per person for cropland. They have diets with fewer animal products, less area demanding crops like grains, and more root crops with high yields and decent caloric density, such as potatoes. Will a "local diet" run up against land (and water) limits, forcing a change in the kinds of foods we can hope to eat?

The reason for the 9 acre per capita value is the inclusion of meat that depends upon low productivity rangeland and pasture. The previously cited value of 1 acre per person refers to the area of high quality land, or prime farmland. For Mendocino County, I am going to assume that meat is produced on non-prime farmland, whereas the other foods are more reliant on good tillable acreage. When the meat area is excluded, a more comfortable figure of ¾ of an acre per person is calculated for the remainder. From part 1 of this series we see that the food produced on this ¾ acre represents 96% of daily calories-basically an ovo-lacto-vegetarian diet.[xiv] Notice also how fruit and vegetables, which are the darlings of farmers markets, only use 3% of the ¾ acres (Fig. 1). True food security requires growing the high calorie crops. In this diet grains and dry beans are estimated to require 50% of prime farmland area directly, and since eggs and dairy rely on grains 90% of the ¾ acres is actually grain plus dry bean area.

Fig. 1. The acreage required to produce each class of food per person, absent meat. Note that because sugar is based on honey it is given no area.

Area Required for the Population

In 2005, the human population of Mendocino County was estimated at 88,161.[xv] I will divide land requirements into two categories: prime farmland and non-prime land. On a per capita basis for this diet the prime farmland need is 0.76 acres and the non-prime 8.33. Multiplying each per capita allocation times the population yields 66,778 prime and 734,675 non-prime acres (these are spreadsheet derived numbers that include more than 2 decimal places in the per capita figures). For the prime farmland portion, the assumed irrigated area in this diet-land model is 6803 acres (10%) and the non-irrigated 59,975 acres (90%). Land area could potentially be reduced significantly if much more area can be irrigated.

There is an additional factor that needs to be considered however. Lands that are not tilled and are subject to low intensity grazing do not have any crop rotation patterns to consider. Rangelands and pasture may be rested, animals of different species may be placed on them to balance the plant species being consumed, and desired plant species may be over-sown, but their fertility does not rely on incorporating "green manure." Fertilizers may be frequently needed for pasture if it is used for hay that is exported from the land, and less often if used to feed a resident grazing herd. The removal of animal flesh from a land-base does eventually require nutrient replacement.

If organic agricultural methods are used (part of my assumption) then the area devoted to food classes grains, sprouting seeds, vegetables and food reliant on grains will need to be rotated with nitrogen fixing legume crops. For example, if a field is sown in wheat for two years, the following year it may be a field of clover. Clover can be under-sown with grains to improve cover crop establishment time and reduce erosion. To ensure that the nitrogen being fixed by the clover doesn't all go into the clover seed, mowing or grazing is done. This creates some overlap, therefore, between the land-base devoted to animal and non-animal production. By grazing and spreading manure, animals can improve the land needed to produce crops.

What this means is that some proportion of the prime farmland is not directly involved in food production each year, thereby increasing the prime farmland area required. At the same time, the area being set aside for green manure can be used for grazing or making hay, thereby decreasing the non-prime land required. It is also possible to lightly graze animals on fields of annual winter grain crops, though opportunities for this may be limited in non-irrigated production systems since early and vigorous fall seedling establishment is not reliable. Pasture hens may be ideal for this practice since they cause less soil compaction in recently tilled land.

To sort out how many acres of cover crop are needed to fix nitrogen for a given acre of food crop, I started with the pounds of nitrogen fixed per year by clover, which is about 100 for irrigated conditions and I estimate about 60 lbs for non-irrigated land.[xvi] The nutrients removed by a crop is proportional to the yields. Using the pounds of nitrogen removed by a crop per year given the estimated yields, I was able to compute a "Green manure factor" that became a multiplier for the actual area required for that crop.[xvii] For example, if wheat removes 30 lbs of nitrogen per year from the soil when dryland farmed, then growing 1 acre of wheat will require a half acre of green manure. A farm could theoretically have 3 acres, with a rotation pattern of clover, wheat, wheat, and be balanced for nitrogen. In this situation if you multiplied the 2 acres in wheat by the "Green manure factor" of 1.5, you would get the 3 acres required to support the nutrient cycle. This is the same basic approach used by Bagdley et al. in their review paper of the yields of organic agriculture.[xviii]

The total prime farmland needed is therefore the area for crops requiring green manure rotation plus the area of those that do not, which sums to 95,401 acres.

Because about a portion of the prime land can be lightly grazed, subtracting this area from the non-prime land allocated to meat production reduces the non-prime land needed to 706,052 acres.

The adjusted per capita prime land requirements are therefore 1.08 acres of prime farmland and 8.01 acres of non-prime farmland. I am somewhat astonished that after all these complex computations we arrive at the number "about 1 acre" per person.

Beyond Nitrogen

Green manures fix nitrogen and add organic matter to the soil, but most commercial fertilizers add nitrogen, phosphorus and potassium (NPK). It may also be necessary to apply other, more minor nutrients, as well as lime to adjust pH and add calcium, less regularly. So even if the nitrogen issue is covered by green manures, what about the rest? Imports to the land may come in the form of rock phosphate, wood ash, oyster shells, green sand, and kelp meal-some of which are produced in energy intensive mining operation, and most of which currently require long-distance transport.

The amount of nutrients needed to grow the food for one person per year can be estimated by the sum of the multiplication of the area per crop by the nutrient demand per area. Local sources of nitrogen would be legume cover crops. Potassium is extracted from the soil by trees and can be used agriculturally by adding wood ashes to compost. Phosphorus may be the most difficult to gather as a field supplement.[xix] In the natural nutrient cycling of this bioregion, salmonoids would bring the richness of their flesh back to the land, essentially countering the process of water removing minerals from continental lands over time. A strong salmon run could therefore be important for future agriculture.

Another plentiful but usually overlooked local source exists for nitrogen, phosphorus, potassium and other vital minerals. The Chinese figured out thousands of years ago that human excrement (feces and urine to be precise) is a valuable source of fertilizer. Here in the U.S. we spent gobs of money trying to get rid of it, but that may have to change.

Data on the fertilizer value of a year's supply of human excrement is hard to come by. Two references I have cite a 1956 report from the World Health Organization.[xx] Table 1 from "Future Fertility" is reproduced here:

Nitrogen can become a gas and be lost during decomposition unless care is taken to balance the carbon:nitrogen (C:N) ratio of the composting process.[xxi] Human wastes have a 5:1 C:N ratio, but starting a compost pile with 35:1 is best for retaining nitrogen. Assuming most of the nitrogen and other nutrients can be efficiently retained; human waste can be recycled to yield about 50% of nitrogen needs, all phosphorus needs, and 40% of potassium needs. This topic warrants a detailed discussion on its own, but that will have to wait for another paper.

Improving Non-prime Land

Some may feel tempted to "improve" non-prime farmland and thereby increase the area available for intensive cultivation. I don't find it credible to rely on non-prime agricultural land for significant productivity of major staple foods. The required inputs per area to make a naturally non-fertile soil and poorly textured soil able to be cultivated are immense. This can be done in someone's backyard, no doubt, but on a scale to significantly contribute to grain production is not viable. However, non-prime land can be an important resource for grazing animals, some perennial crops, and through intensive modification small-scale cultivation of specialty crops or vegetables.

Research done for New York State, for example, found that the most area efficient diet was vegetarian and required 0.6 acres per person-very close to the ovo-lacto-vegetarian diet plan for Mendocino County. By adding 0.1 acre per person in the form of permanent pasture, the total number of people that could be fed in NY State was increased.[xxii] (Note: this was the thesis work of a Cornell University graduate student and so is far more thorough than my own!) While this addition of permanent pasture in NY was 17% above the tillable cropland, in this exercise for Mendocino County the non-prime acreage is 6.3 times greater. Reasons I can think of for this huge difference are much more productive pasture in upstate New York due to summer rains, and methodological flaws on my part in estimation of Mendocino County range and pastureland productivity (the amount of meat in the NY and Mendocino diets is similar).

The next essay will examine the actual land-base of Mendocino County.

Acknowledgement

Christoffer Hansen deserves thanks for reading drafts of this work. He questioned my assumptions, forcing me to explain them more clearly, and double checked the math in many places. Any remaining errors are my own.

[i] http://www.energyfarms.net/node/1489

[ii] http://soils.usda.gov/technical/handbook/contents/part622.html and http://www.dot.ca.gov/ser/vol1/sec3/community/ch23farm/chap23farm.htm#Ch24Definitions

[iii] http://www.mendowine.com/

[iv] http://www.willitseconomiclocalization.org/files/well/FoodSecurityReport.pdf

[v] http://anrcatalog.ucdavis.edu/pdf/8208.pdf

[vi] http://ucce.ucdavis.edu/files/filelibrary/2161/29131.pdf

[vii] http://www.co.mendocino.ca.us/agriculture/pdf/2006%20Crop%20Report.pdf

[viii] Soil Survey of Mendocino County, California, Western Part. http://www.ca.nrcs.usda.gov/mlra02/wmendo/ and http://soildatamart.nrcs.usda.gov/Manuscripts/CA694/0/MendocinoWP_CA.pdf; search for Mendocino County at http://soildatamart.nrcs.usda.gov/

[ix] http://www.bakewellrepro.com/sarahflackarticle.html

[x] http://www.co.mendocino.ca.us/agriculture/pdf/2006%20Crop%20Report.pdf

[xi] http://www.shutersunsetfarms.com/index_files/Page949.htm

[xii] http://www.oznet.ksu.edu/library/lvstk2/mf754.pdf

[xiii] http://dieoff.org/page40.htm

[xiv] http://www.energyfarms.net/node/1489

[xv] http://www.city-data.com/county/Mendocino_County-CA.html

[xvi] http://attra.ncat.org/attra-pub/covercrop.html

[xvii] I can't find a web-available reference for the table I used, but a similar table using non-organic agricultural demands and so generally with higher values is here: http://benson.byu.edu/Publication/Lessons/EN/Agronomy/SoilFertility.asp

[xviii] http://www.journals.cambridge.org/action/displayAbstract?fromPage=online&aid=1091304&fulltextType=RA&fileId=S1742170507001640 and related press release: http://www.ns.umich.edu/htdocs/releases/story.php?id=5936

[xix] http://www.energybulletin.net/28720.html

[xx] http://www.jenkinspublishing.com/humanure.html; with quantities in terms of pounds produced per person per year given in John Beeby "Future Fertility: Transforming Human Waste in Human Wealth" Table 1. Ecology Action.

[xxi] http://en.wikipedia.org/wiki/Denitrification

[xxii] http://www.news.cornell.edu/stories/Oct07/diets.ag.footprint.sl.html