Having argued recently that Athens needed a great deal of grain imported, let me shift gears and argue the opposite.


Athens was situated in an area of Greece that is not particularly productive. While there are parts of Attica that are good for growing grain, the amount of land that is arable is limited. Based on a 1961 census of land in Attica, Peter Garnsey guessed that somewhere between 35-40 percent of the land of Attica was cultivable. Michael Whitby accepts this percentage with the proviso that some of that cultivable land was not suited to grain. However, according to Robert Sallares, some of the land included as arable in the 1961 census was so poor it had only been capable of producing a cereal crop once in the five years prior to the survey. He concluded that 35 percent of Attica available for grain was a maximum, so 30 percent seems more likely. The actual area sown would have varied from year to year in response to a number of factors.

Not the least of those factors is the issue of the fallow regime followed. The debate revolves around the question of the predominance of a two- or three-year cropping cycle. Garnsey argues that the two-year system cannot be proven and therefore the assumption that only 50 percent of the cultivable land was available for grain is incorrect. In classical Attica, the common type of farms would have been small-scale, intensive, mixed farms, which would allow for a higher percentage of the land being available for grains and would have made the need for a two-year fallow system less severe. Sallares, however, notes that in Attica “a not insignificant proportion of arable land is still left fallow.” He continues by stating that Attic soil received too little rainfall for growing leguminous crops that were a part of the three-year crop rotation favored by Garnsey. Garnsey also notes this fact. A two year crop rotation provided the best chance of long term cropping in Attica and according to Whitby, Sallares’ arguments suggest that grain was grown on arable land in Attica either one year in two or one year in three with a probable preponderance towards one year in two. Sallares notes that the cereal yields of early modern Greece on the eve of the introduction of chemical fertilizers were the result of the practice of biennial fallow. Still, Salares and Whitby are unable to completely discount the possibility of a three-year fallow system. That crops were highly intermixed in Attica seems a very likely possibility. Likewise, the lack of hard evidence supporting a two-year cycle tends to strengthen Garnsey’s belief that more than 50 percent of the available land was under cultivation.

Other factors contributed to a low percentage of land in Attica being used to grow grain. Rich landowners might not devote large tracts of land to the production of grain, since it was not the most profitable of crops. They might also dedicate portions of their land to the production of crops that went to sustain their horses and other livestock. Other arable land would have been used for gardens. Further, cultivable land could have been kept out of cultivation either through neglect or inability to work it as seen in Xenophon (Economics, 20.22)

Sallares does not believe that any arable land was likely left unsown, noting that in the classical period, there was practically no cultivable land in Attica left unoccupied. Taken singly, these amount to very little, but cumulatively, they deserve mention. If Sallares is correct and 30 percent of Attic land was cultivable for grain, and if the dominant form of fallow was a two year cycle, than a maximum of 15 percent of the lands in Attica could have been used to grow grain in a single year, and Whitby notes that it was probably between 10-15 percent.

This percent was the lowest Garnsey argued and based other calculations on 20 percent and 25 percent arguing that there was a three-year rotation and that vines and olives could occupy hill slopes and poorer arable land. Garnsey also noted that there are remains of terracing, which indicates that cultivation in antiquity was more extensive than it is today. Garnsey’s arguments have merit, and despite Sallares’ arguments to the contrary, I find that there is reason to raise the estimate of arable land under cultivation for the growth of grain to be higher than the 15 percent Whitby argues for. I accept Garnsey’s figures as being more likely, that is that 20-25 percent of Attica’s arable land was used each year for the growing of grain.

The last question that must be asked concerning the grain production of Attica is how much yield was returned on the Attic grain crop and therefore how many people could be fed from the grain crops of a normal year in Attica. There is very little actual evidence concerning this. The Eleusis First Fruits inscription stands alone in regards to yields for grains in Attica. The inscription is dated to 329/8 and has been used to work out how much wheat and barley was produced in Attica and various dependent territories in that year. Garnsey uses the inscription in order to show that the acreage under grain cultivation in Attica must be higher than is often believed. The Eleusis First Fruits inscription must indicate a very poor year in grain cultivation for Attica because according to Garnsey’s calculations, a yield of 8 hectolitres per hectare (the bottom end of a regular year of production according to Jarde) in wheat and 16 hectolitres per hectare in barley would mean that Attica had only 5.4 percent of its land under cultivation for grain. This is in comparison to Lemnos, which at 8 and 16 hectolitres per hectare of wheat and barley respectively produced enough to account for 24.9 percent of its land. In order for Attica to reach a similar total of land, its yields must be reduced to approximately 2.25 and 4.25 hectolitres per hectare for wheat and barley. No one argues that the percent of Attic land under cultivation for grain is any where near as low as 5.4 percent, nor would they agree that in a normal year, Attica has yields as low as 2.25 and 4.25 respectively, therefore, the Eleusis First Fruits inscription must indicate a very poor year for grain in Attica.

Average grain yields in Attica may well have been slightly lower than in other areas of Greece, but this would have been due to the cultivation of marginal land. As Sallares noted, almost no cultivable land was left unused, and so the overall amount of arable land would have produced a lower average yield than other areas. While wheat yields may have been somewhat lower than the norm in Attica, the yield of barley should have been somewhat higher than normal. Theophrastus said that Attic soils were excellent for producing barley. “At Athens the barley produces more meal than anywhere else, since it is an excellent land for that crop.” Clearly Athens was able to produce a good crop of barley. In addition to the light soil production of barley, the plains of Marathon and Mesogeia have rich soil with a relatively deep plough-zone that make for good wheat production. In contrast to the numbers that the Eleusis First Fruits inscription seems to indicate, Garnsey proposes the following numbers for Attic production: 8 hectolitres per hectare of wheat and 12 hectolitres per hectare of barley. These productivity numbers are on the low side, especially for the barley. Jarde guessed the normal returns for wheat and barley in Greece was in the range of 8-12 hectolitres per hectare for wheat and 16-20 hectolitres per hectare of barley. On s'éloignera sans doute peu de la vérité en supposant pour la Grèce antique un rendement de 8 à 12 hectolitres à l'hectare pour le froment, de 16 à 20 pour l'orge. While wheat productivity was probably at the lower end, Attica’s good reputation for growing barley makes me inclined to raise Garnsey’s estimates to 18 hl/ha, right in the middle of Jarde’s estimates. Given these numbers for production, if only 20 percent of Attic land was under grain production, Attica would have been able to produce 4,446,720 kg of wheat and 33,333,120 kg of barley.

Garnsey estimated that a generous consumption rate would be 175 kg of grain per person, per year. Whitby notes that Garnsey’s estimate does not differentiate between wheat and barley consumption. While 175 kg of wheat per year may well suffice, 175 kg of barley would not. Since a person’s total caloric intake would most likely not come solely from grain consumption in ancient Attica, I accept that 175 kg of wheat per person would be an adequate amount of wheat per year. The research of L. Foxhall and H. A. Forbes would seem to support this assumption. They estimated that approximately 212-237 kg of wheat per year was the average consumption, but believed the numbers to be too high. However, the weight of barley consumed per person per year must be raised, and 360 kg of barley per person per year seems an adequate amount. Thus, Attic produced barley could have fed 92,600 people while Attic produced wheat would have fed an additional 25,500 residents. It should be remembered that these numbers represent a normal year in Attica and are weighted to the lower end of production. The actual number of residents fed could be increased by an additional 14,000-27,000 (by increasing either the acreage cultivated or the average production numbers). But maintaining our most pessimistic numbers, Attica could feed approximately 120,000 residents, without any external aid. Garnsey also argued that an inner ring of dependant states would have fed an additional 20-25,000 Athenians bringing our total up to approximately 140,000 Athenians fed on local grain each year.

The last variable that needs to be investigated is the actual population of Attica during the time in question. Just as the search for grain production estimates, the search for population figures is based on best guesses and assumptions. The numbers vary widely. Whitby believes that the population of Attica in the mid-fourth century was approximately 250,000-300,000. These numbers are significantly higher than Garnsey’s estimates of approximately 150,000-200,000 in the mid-fourth century. As we are dealing with the previous century, neither of these numbers is relevant except as comparisons. Garnsey believed that following the Persian Wars (480), the Attic population was between 120,000 and 150,000. This number rose to a high of about 250,000 immediately preceding the Peloponnesian War. Gomme’s numbers are similar to Garnsey’s. Gomme guessed that the population of Attica in 480 would have been approximately 140,000. At the beginning of our period then (479 BC), the population of Attica could not have been much greater than 140,000, and by the end of our period (440), it might possibly have reached approximately 200,000. As we have already seen, Attica was capable of producing enough grain to feed approximately 120,000-140-000 of these residents. The dire need for imported grain to Athens argued by some may not be the case.

So if the possibility of the need for large scale imports of grain to Athens is not necessarily supported by the evidence for Attica, can it be indicated by Athenian activities in securing grain from elsewhere?

Athenian actions following the formation of the Delian League did not protect the Black Sea grain route, nor is there any other indication that Athens was even interested in the Black Sea grain during the first half of the fifth century B.C.
According to Thucydides, the first action that the Athenians undertook upon the formation of the Delian League was to siege the town of Eion. They captured the town and made slaves of its inhabitants. In addition to driving out the Persians, the conquest of Eion provided raw materials for the League, especially wood for the fleet. Following this, the Athenians turned to the island of Scyros. Scyros is in the Aegean and would be directly on the route from the Black Sea. While Scyros would undoubtedly have been an important island to control in order to protect the Black Sea grain route, this does not appear to be the case here. Plutarch tells us the reason behind the League attack on Scyros in his Life of Cimon (Plutarch, Cimon, 8.3). The Dolopes living on Scyros were disrupting the trade of the Aegean states; those states were part of the Delian League. The League was made up of island and coastal towns whose economies were largely dependant on trade. It is easy to see why the League would have been quite happy to use its power to remove this non-Persian threat.

The next recorded Athenian action was against Carystus, a city of Euboea. Once again the League was using its might against a Greek state. Thucydides gives no reason for this attack, but Buckley posits that the reason may be implied in the historian’s text (Thucydides 1.98.3). Buckley believes that once again the League is showing its flexibility in going beyond its original charter. He notes that Carystus was a coastal town in Euboea that would have been enjoying all the benefits, both political and commercial, that the League was providing. It was, however, benefiting from the League’s actions without actually helping to pay for those actions. The city surrendered on terms, and those terms likely included Carystus becoming a phoros-paying member of the League. Buckley goes on to say that Athens was once again acting out of personal gain because Carystus, like Scyros, was ideally situated along the corn route from the Hellespont. While it is possible Athens used these opportunities to protect the grain route, I think it far more likely that both Scyros and Carystus were conquered by Athens and her allies for the reasons previously stated, and not because of their position along the grain route, which as we have seen was probably of minor importance at this time.

The next action the Athenians undertook was the restoration of Naxos to the League. While we do not know why Naxos decided to revolt from the League, Plutarch and Thucydides both agree that the main problem within the League became the unwillingness of the allies to provide men and ships for the war against the Persians. Athens was maintaining its hegemony of the League and in this action reduced Naxos to a subject state. This action is, to my mind, an extension of the policy that forced Carystus into the League. Naxos was not to enjoy the benefits of the League without remaining a part of the League.

The Athenians and League members then fought the battles of the Eurymedon in Pamphylia. Both a land and a sea battle were fought, the Athenians being victorious over the Persian forces (Cimon 12.1). At this point, the Phoenician fleet was destroyed and the threat of a Persian resurgence in the Aegean must have seemed small indeed to the members of the Delian League. What follows this episode is the siege and reduction of Thasos, an episode that shows the Athenians to be self-seeking as they succeeded in stripping from Thasos a mine and the markets under the control of the Thasians. From this point on, the Pentecontaetia details Athenian actions down to approximately 432. Every action of the Athenians after the Battles of the Eurymedon River were against other Greeks, and focused either on the mainland, the Peloponnese, or in subduing rebellious allies. The only exception to this is the Athenian deployment of a large fleet to Egypt in about 460 B.C. The Athenians were invited to help liberate Egypt from Persian control and this is the only action that Thucydides reported in which Athenian forces engaged Persian forces. In the end, the Athenian forces were defeated.

What is clear from the rather limited account of the Pentecontaetia is that there is no action that is known in which the Athenians can be seen to be definitively protecting the grain route from the Black Sea. Buckley believes that the capture and incorporation of Scyros and Carystus into the League represent Athenian attempts to secure the corn route, but other more valid reasons existed for these actions. We must abandon the belief that Athenian foreign policy was based upon a desire to protect the Black Sea corn route.

There is no evidence to support Athenian interest in grain from the Black Sea region until the beginning of the Peloponnesian War. Even then we have no evidence of trade for grain in the Black Sea on behalf of Athens, despite the account in Plutarch concerning Pericles’ expedition to the Black Sea. During the Archaic period, there is no reason to believe that Athens sought grain from the Black Sea notwithstanding Herodotus’ statement that merchant vessels were seen sailing out of the Pontus in 494/3. Herodotus does not say the ships were going to Athens, or that they were carrying grain. Herodotus does tell us that the Greeks continued to trade with Egypt even during the Persian Wars and that another avenue of grain for the Greek mainland was Sicily. Gelon, the tyrant of Syracuse promised to provide grain for the entire Greek army for the duration of the conflict with the Persians. Grain from the Black Sea is not mentioned. Given the lack of any evidence to the contrary, Garnsey has stated that a regular grain trade with the Black Sea in the archaic period can only be considered a “figment of the imagination.”

So when did Athens become dependant on the Black Sea grain route? It is possible that Athens lost a major source of grain with the defeat of its forces in Egypt c. 454 B.C. Such a loss would have had considerable effect on the rapidly expanding population of Athens. Antony Keen agrees. He finds that the inclusion of a number of Euxine states in the tribute assessment of 425 an indication of Athenian interest in the Pontic grain route at least in the mid-5th century. It is likely that following the outbreak of the Peloponnesian War in 432, and the annual Spartan invasion of Attica, Athens was no longer able to provide much, if any, of its required food supply. Pericles’ expedition to the Pontus supposedly occurred just a few years prior to the outbreak of the war, years in which the coming of the war was likely foreseeable. The policy that Pericles advised Athens to follow during the war would seem to indicate that he understood that the bulk of Athenian food would necessarily be imported in the following years (Thuc. 1.143.5). This admission that the land must for the present time be forgotten indicates that the Athenians must now rely on foreign imports for all their grain. Clearly Athenian grain imports before the Peloponnesian War, while possibly significant, did not make up the bulk of grain consumed in Attica. To think otherwise would be to consider the Spartans ignorant of the amount of grain Attica produced. They plainly believed that the devastation of the Attic grain crops would force Athens into either surrender or battle. It was their stated strategy in order to win the war quickly.

The Athenian need for imported grain probably existed from before the Persian Wars. That Athens imported the majority of its grain at this time, however, is unlikely. Given the potential of grain production in Attica and the probable population figures following the Persian Wars down till the mid 5th century, Athens would have been able to feed a large proportion of its residents on its own. The need for ever greater amounts of imported grain grew, of course, but not until the middle of the century, and possibly later (the outbreak of the Peloponnesian War) did Athens depend on foreign imports for a majority of its grain. As noted above, Attica could have produced enough grain in a normal year to provide for approximately 120,000 to 140,000 of its 200,000 residents at mid century. In addition, no evidence indicates that Athens was importing food in large quantities before the beginning of the Peloponnesian War.

That Athens aided Corcyra against Corinth in order to maintain the grain routes from Capua and Sicilia is speculation. We know Gelon offered to provide grain during the war against Persia, but as far as I know we have no further suggestion that Athens herself purchased grain from the west. Indeed, Kagan (I believe) argues that the reduction of Sicily by Athens was attempted in order to take away a supply of grain from the Spartans. Clearly, then, the grain could not also have been going to Athens.