Climate, crop failures and hunger

Article I examined the relationships between monthly mean temperatures and pro-vincial crop yields of rye and barley in Finland over the period 1861–1913 using a correlation analysis. Overall, yields of both crops correlated positively with spring and summer (April-August) mean temperatures in large parts of the country. Warm-er tempWarm-eratures wWarm-ere consistently associated with highWarm-er crop yields, emphasising the importance of temperature as a yield-limiting factor in pre-industrial agriculture in the region. In addition to spring and summer mean temperatures, the onset of the growing season was found to have a decisive impact on the crop yields. The onset of the growing season, in turn, was partly dictated by the winter severity (temperature and snow cover). Evidence of precipitation playing a significant role in controlling crop yields in pre-industrial Finland was not found (Article I).

The crop yield sensitivity to temperature was found to vary spatially within 19th century Finland. Crop cultivation was most temperature-sensitive in the north-ern provinces, whereas the temperature linkage almost disappeared in the most south-eastern parts of the country. Additionally, the crop yields of rye correlated positively with January mean temperatures in southern Finland (Article I, Figure 2).

Although considerable spatial differences in the temperature response were found, both crops failed in large parts of the country in 1867, 1877, 1881, 1892, 1899, 1902. These years of countrywide crop failure coincided systemically with the years of low April-August mean temperatures (Figure 2). Because of the low temperatures, the ripening of the crops was delayed to a period when the first autumn night-frosts commonly occur (Article I). Thus, the low spring and summer temperatures explain the extent and severity of the frost damage on crops. In most cases, the low tem-peratures directly affected the same year’s harvest. Alternatively, when the adverse conditions affected also the quantity and quality of the seed grain, the crop response could lag one year.

Article II further analysed the climate-sensitivity of crop cultivation in medie-val Novgorod and Ladoga (1100–1500) with qualitative analysis. These results were largely in agreement with the findings of the statistical analysis of the 19th century data. Also in medieval north-west Russia, most of the severe crop failures resulted from harvest time night-frosts and these failures happened in years when adverse climatic conditions had delayed the ripening of the crops. Similar to the findings of

Article I, delayed onset of the growing season and cool summers were considerable threats to crop cultivation. Yet, in addition to low temperatures, precipitation vari-ability was also shown to influence harvest success in Novgorod (Article II).

6 7 8 9 10 11 12

-4 -3 -2 -1 0 1 2

1870 1880 1890 1900

Yield ratio Tree ring Temperature

Yield ratio and tree-ring (z-scores) April August temperature (°C)

Crop faulure

Figure 2. Finnish crop yield ratio, April-August mean temperature (averaged from Helsinki and Tornio stations), and averaged southern and northern tree-ring MXD series 1866–1905. The grey shadowing indicates years of countrywide crop failures. The close correspondence be-tween the different series illustrates how tree-ring data can be used as both climate and yield proxy data.

Source: Article I; Klingbjer & Moberg 2003; Tuomenvirta 2004; Helama et al. 2014; Matskovsky

& Helama 2014.

Note: See Article I for information on the missing values. The yield ratio was calculated with weighted arithmetic mean from the provincial crop yield data, where the weighting was deter-mined by the share of the provincial grain (rye or barley) harvest to the total harvest

3.2.2 Crop failures and hunger

The analysis carried out in Article II found that adverse climate or weather was at least a contributing factor in 90% of the recorded hunger events in medieval Novgorod and Ladoga. Similarly, the time series analysis in Article IV found that each time contemporaries witnessed hunger in early modern Ostrobothnia, the climatic condi-tions were considerably unfavourable for crop cultivation over the preceding years.

(Article IV). Thus, there is strong evidence that climate-driven crop failures underlay the recurrent hunger experienced in medieval and early modern North-East Europe.

However, the significance of adverse climate underlying hunger varied over the studied period and area (Article II, V). For example, the role of climatic factors un-derlying acute hunger was less evident in 14th and 15th century Novgorod (Article II) than in 17th century Finland (article IV).

Moreover, the situation did not evolve into a hunger crisis each time the tree-ring data or narrative sources indicated climate and weather conditions potentially unfa-vourable for crop yields (Article II, IV, V). Commonly, the most severe hunger crises

followed after at least two back-to-back harvest failures (Article II, IV). Despite the considerable climate-sensitivity of the crop yields in the region, the results suggest that the hunger crises in the north-east were almost never caused by climate-driven production failures alone. The harvest success did not solely determine the well-be-ing of the population. Consequently, the relationship between climate and crop fail-ures appears to be generic, whereas the relationship between crop failfail-ures and hun-ger was more complex in the medieval and early modern north-east.

3.2.3 Resilience to adverse climate and crop failures

Resilience, the ability to anticipate, cope, or recover from the crop failures, explains the more complex pathways from crop failure to hunger.⁸² First of all, the people on the northern margin of agriculture anticipated that climate and weather posed a risk to crop cultivation and that crop failures were likely. Consequently, they favoured high-yielding methods, like slash-and-burn cultivation, to maximize the yield in a good year (Article II, III). The contemporaries must have had some grain storages, as Article IV demonstrated that the peasants commonly faced the greatest hardships one to three years after the first crop failure event. Furthermore, the worst climate-driv-en hunger crises examined in the articles, the famines of 1127–1128 and 1230–1231 in Novgorod and 1601–1603 and 1695–1697 in Ostrobothnia, culminated during the second or third successive year of adverse climate and crop failure (Article II, IV).

Second, the wilderness, forests and waters, provided alternative – and often cru-cial – sources of nutrition. Fish in particular might have had a decisive role wheth-er grain shortage evolved to a hungwheth-er crisis (Article II, V). Moreovwheth-er, cultivation of other crops and vegetables, like millet and root vegetables, provided supplementary sources for subsistence (Article II). And lastly, people had learned to cope with fre-quent crop failures and prepared substitute foods, like birch- and pine-bark bread (Article II).

The level of resilience, however, varied within the studied peasant populations.

The poor were the most vulnerable. For example, in the 17th century Ostroboth-nia, the landowning peasants were more resilient than the tenant farmers. Article IV demonstrates that the social consequences of severe crop failures, like impoverish-ment, were considerably profound among the tenant farmers.

3.2.4 Spatial and temporal differences

Climate-driven crop failures are commonly more local than regional because of the spatial variability in climate and weather patterns, soil types, and agricultural prac-tices (Article IV). Article V discusses how the food systems differed from place to place over the north-east. For example, when considering medieval times, in many

82 Cardona et al. 2012, 75.

areas of the north-east, the daily source of nutrition came from fishing, herding and hunting.⁸³ In many areas, crop cultivation acted only as a supplement for sustenance.

Thus, only the people whose livelihood depended on crop cultivation were affected by the adverse weather and climate conditions described above. Therefore, severe crop failures, like the one in 1314, only impacted severely on part of the population (Article V).

However, over the studied time the importance of crop cultivation as the main livelihood strategy increased notably among the population. Article IV found that climate extremes, like those resulting from volcanic forcing, could cause simultane-ous crop failures and food shortages across north-east Europe. These shortages could then escalate into pan-regional subsistence crises, like after the 1601 and 1695 crop failures.

83 However, climate variability is known to influence fish and game populations. Yet, the responses are commonly slower and more moderate than the agricultural responses. See, Article V for further discus-sion.

4 DISCUSSION