Sorghum and Finger Millet Cultivation during the Aksumite Period: Insights from Ethnoarchaeological Modelling and Microbotanical Analysis

Abel Ruiz-Giralt; Alemseged Beldados; Stefano Biagetti; Francesca D’Agostini; A. Catherine D’Andrea; Yemane Meresa; Carla Lancelotti

doi:10.5334/jcaa.132

Figures & Tables

Elevation map of the study area showing the location of the sites included in the model.

Workflow used in this study. Grey boxes represent previously published data. Yellow boxes contain new data. Blue shapes represent datasets. Processing data steps in R software 4.2.2 (R Core Team 2021) are red rectangles. Green shapes represent the results of R processing.

Table 1

Definitions of agricultural practices considered in this study.

VARIABLE	DEFINITION	REFERENCES
Casual Agriculture	Slight or sporadic cultivation of food or other plants incidental to a primary dependence upon other subsistence practice	Murdock 1981: 98
Extensive Agriculture	Or shifting cultivation, as where new fields are cleared annually, cultivated for a year or two, and then allowed to revert to forest or brush for a long fallow period	Murdock 1981: 98
Intensive Agriculture	On permanent fields, utilizing fertilization by compost or animal manure, crop rotation, or other techniques so that fallowing is either unnecessary or is confined to relatively short periods	Murdock 1981: 98
Rainfed Agriculture	Water is provided by rainfall alone (directly or as run-off), cultivation occurs far from any permanent water sources and without any water harvesting	Lancelotti et al. 2019: 1027
Décrue Agriculture	Or floodplain cultivation, as where water is provided by natural inundation, typically from major river systems.	Lancelotti et al. 2019: 1027
Irrigated Agriculture	Water is provided to crops at regular intervals throughout the growing season by human intervention	Lancelotti et al. 2019: 1027

Table 2

Performance measures of model fitting and cross-validation using different classification thresholds (Acc. = Accuracy, Prec. = Precision, Rec. = Recall, F1 = F1-Score).

MODEL	METHOD	MODEL FITTING				CROSS-VALIDATION
MODEL	METHOD	ACC.	PREC.	REC.	F1	ACC.	PREC.	REC.	F1
Crop choice	Sens = Spec	0.77	0.84	0.75	0.80	0.63	0.89	0.93	0.76
	MaxSens + Spec	0.81	0.94	0.73	0.82	0.54	0.67	0.52	0.67
	MaxPCC	0.85	0.85	0.92	0.88	0.89	0.89	1.00	0.94
	PredPrev = Obs	0.82	0.86	0.84	0.85	0.89	0.89	1.00	0.94
	ObsPrev	0.76	0.93	0.64	0.76	0.54	0.93	0.52	0.67
	MinROCdist	0.81	0.91	0.74	0.82	0.54	0.93	0.52	0.67
FM cultivation	Sens = Spec	0.95	0.95	0.95	0.95	0.49	0.48	0.49	0.48
	MaxSens + Spec	0.95	0.95	0.95	0.95	0.49	0.48	0.49	0.48
	MaxPCC	0.95	0.95	0.95	0.95	0.49	0.48	0.49	0.48
	PredPrev = Obs	0.95	0.95	0.95	0.95	0.49	0.48	0.49	0.48
	ObsPrev	0.62	0.75	0.35	0.48	0.4	0	0	0
	MinROCdist	0.95	0.95	0.95	0.95	0.49	0.48	0.49	0.48
SB cultivation	Sens = Spec	0.8	0.67	0.78	0.72	0.73	0.6	0.67	0.63
	MaxSens + Spec	0.78	0.63	0.85	0.72	0.71	0.55	0.7	0.62
	MaxPCC	0.89	0.85	0.84	0.84	0.98	0.98	0.98	0.98
	PredPrev = Obs	0.86	0.79	0.79	0.79	0.93	0.93	0.85	0.89
	ObsPrev	0.63	0.44	0.35	0.39	0.41	0.03	0.02	0.02
	MinROCdist	0.79	0.64	0.81	0.72	0.67	0.5	0.57	0.53

Table 3

Summary of the results of the model application to 16 archaeological sites in percentage. The full results are available as Supplementary Information, Table S1 (FM = Finger millet, SB = Sorghum, CAS = Casual, EXT = Extensive, INT = Intensive, RF = Rainfed agriculture, DEC = Décrue/Flood agriculture, IRR = Irrigated agriculture). Note that length of growing cycle was not used for FM as it was not retained by the model as a significant explanatory variable.

MODEL		1 KM	5 KM	10 KM	25 KM	50 KM	100 KM	200 KM
Crop choice	FM	93.75	93.75	87.5	87.5	93.75	62.5	93.75
Crop choice	SB	18.75	31.25	43.75	43.75	87.5	100	100
FM cultivation	EXT-RF	91.67	92.86	92.31	92.86	66.67	90.91	93.33
	INT-RF	8.33	7.69	7.69	7.14	33.33	9.09	6.67
	INT-IRR	0	0	0	0	0	0	0
SB cultivation (90 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	0	0	0	0	0	0	0
	INT-RF	0	0	0	0	0	0	0
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	100	100	85.71	100	100	93.75	100
SB cultivation (120 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	0	0	0	0	0	0	0
	INT-RF	33.33	33.33	71.43	75	64.29	75	100
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	66.67	80	28.57	25.0	28.57	12.5	0
SB cultivation (150 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	0	0	0	0	0	0	0
	INT-RF	100	100	100	100	100	100	100
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	0	0	0	0	0	0	0
SB cultivation (180 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	0	0	28.57	25	0	0	0
	INT-RF	100	100	71.43	62.5	85.71	93.75	100
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	0	0	0	0	0	0	0
SB cultivation (210 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	33.33	40	71.43	62.5	78.57	100	100
	INT-RF	66.67	66.67	28.57	37.5	28.57	18.75	0
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	0	0	0	0	0	0	0
SB cultivation (240 days)	CAS-RF	0	0	0	0	0	0	0
	EXT-RF	100	100	100	100	100	100	100
	INT-RF	33.33	20	0	12.5	0	0	0
	INT-DEC	0	0	0	0	0	0	0
	INT-IRR	0	0	0	0	0	0	0

Table 4

Summary of results from single-celled phytolith analysis grouped by taxonomic categories.

ARCHAEOLOGICAL CONTEXT		INDETERMINATE HERBACEOUS		C4							C3	PALMS AND HERBS			WEEDS
ARCHAEOLOGICAL CONTEXT	N SAMPLES	POACEAE/CYPERACEAE (%)	POACEAE (%)	ANDROPOGONEAE (%)	CF. ARISTIDOIDEAE (%)	PANICOIDEAE/CHLORIDOIDEAE (%)	PANICOIDEAE/CHLORIDOIDEAE (%)	PANICOIDEAE (%)	CHLORIDOIDEAE (%)	CF. CHLORIDOIDEAE (%)	POOIDEAE (%)	POOIDEAE (%)	ARECACEAE (%)	ZINGIBERALES (%)	CF. ZINGIBERALES (%)	CYANOTIS SP. (%)	COMMELINA SP. (%)	COMMELINACEAE (%)	CYPERACEAE (%)	NON DIAGNOSTIC (%)
Late Pre-Aksumite	4	29.4	28	0	0	7.4	1.4	0.2	9.3	0.3	4.3	0.2	1	0	0.1	0	0	0	2.6	15.6
PAA transition	10	24.3	31.7	0.1	0.04	7.7	1.1	0.3	10.2	0.2	4.4	0.3	1.4	0.2	0	0.2	0	0	2	15.9
Early Aksumite	9	24.7	35.1	0.04	0.04	8.1	1.2	0.1	9.2	0.1	3.4	0.2	1.4	0	0.04	0	0	0	1.6	14.8
Middle Aksumite	9	28.5	30.5	0	0	10.1	1.1	0.2	7.7	0.4	3.5	0.3	1	0	0	0.4	0.04	0.04	1.2	14.8
Late Aksumite	14	31.1	26.9	0	0	10	0.4	0.1	7.5	0.2	4.2	0.2	0.6	0.03	0.03	0.1	0	0	2	16.9
Total Ona Adi	46	27.7	30.4	0.02	0.02	9	0.9	0.2	8.6	0.2	4	0.2	1	0.04	0.02	0.1	0.01	0.01	1.8	15.7

Plot of the probability of Ona Adi’s archaeological phytolith samples to be derived from a wellwatered crop-phytolith assemblage.

Sorghum and Finger Millet Cultivation during the Aksumite Period: Insights from Ethnoarchaeological Modelling and Microbotanical Analysis

Figures & Tables

Figure 1

Figure 2

Table 1

Table 2

Table 3

Table 4

Figure 3

Paradigm

My account