© The Authors, 2026, Published by the Universidad del Zulia*Corresponding author: m.hattab@lagh-univ.dz
Keywords:
Durum wheat
Yield losses
Emergence
Mechanical harvesting
Eect of management practices on yield and seed loss in durum wheat under arid conditions
Efecto de las prácticas de manejo en el rendimiento y la pérdida de semillas en trigo duro bajo
condiciones áridas
Efeito das práticas de manejo no rendimento e na perda de sementes em trigo duro sob condições
áridas
Mourad Hattab
1*
Abderrahmane Kessaissia
2
Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331
ISSN 2477-9407
DOI: https://doi.org/10.47280/RevFacAgron(LUZ).v43.n2.XIII
Crop production
Associate editor: Dra. Rosa Razz
University of Zulia, Faculty of Agronomy
Bolivarian Republic of Venezuela
1
Department of Agronomic Sciences. Faculty of Sciences.
Laboratory of Biological and Agronomic Sciences (LSBA),
Amar Telidji University, Laghouat, Algeria.
2
National Institute for Forestry Research (NIFR).
Experimental Station for the Protection of Watersheds,
Ténès. Chlef, Algeria.
Received: 06-02-2026
Accepted: 02-05-2026
Published: 28-05-2026
Abstract
Despite irrigation expansion, durum wheat yields in Algeria
remain substantially below genetic potential. This study quantied
emergence, natural, and mechanical losses under irrigated arid
conditions in Laghouat Province, Algeria, during 2024-2025. Five
experimental elds (Vitron variety) were monitored, measuring
emergence densities, potential yield components, pre-harvest yield,
and harvested yield. Emergence losses (24-43 %) were compensated
through tillering (r = -0.996, p ≤ 0.001), though excessive densities
(> 600 seeds.m
-2
) increased losses without yield benet. Potential
yield ranged from 9.3 to 13.7 t.ha
-1
, with grain number per spike as
primary determinant (r = 0.994, p ≤ 0.001). Natural losses (12-38
%) were amplied by extreme weather (hail, 36 mm.h
-1
rainfall,
May 14, 2025). Mechanical losses (31-47 %) exceeded international
standards (3-7 %). Cumulative losses reached 54-60 % of potential,
valorizing between 40 and 46 %. Optimization must prioritize
mechanical loss reduction, legume rotations, and row seeding at
moderate density.
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331 April-June ISSN 2477-9407.
2-7 |
Resumen
A pesar de la expansión del riego, los rendimientos de trigo duro
en Argelia permanecen sustancialmente por debajo del potencial
genético. Este estudio cuanticó las pérdidas en emergencia,
naturales y mecánicas bajo condiciones áridas irrigadas en la
provincia de Laghouat, Argelia durante 2024-2025. Cinco campos
experimentales (variedad Vitron) fueron monitoreados, midiendo
densidades de emergencia, componentes del rendimiento potencial,
rendimiento pre-cosecha y rendimiento cosechado. Las pérdidas en
emergencia (24-43 %) fueron compensadas mediante ahijamiento (r =
-0,996; p ≤ 0,001), aunque densidades excesivas (> 600 semillas.m
-2
)
incrementaron las pérdidas sin benecio en rendimiento. El
rendimiento potencial varió entre 9,3 y 13,7 t.ha
-1
, siendo el número
de granos por espiga el determinante principal (r = 0,994; p ≤ 0,001).
Las pérdidas naturales (12-38 %) fueron amplicadas por condiciones
meteorológicas extremas (granizo, lluvia de 36 mm.h
-1
, 14 de mayo de
2025). Las pérdidas mecánicas (31-47 %) excedieron los estándares
internacionales (3-7 %). Las pérdidas acumuladas alcanzaron 54-
60 % del potencial, valorizando entre 40 y 46 %. La optimización
debe priorizar la reducción por pérdidas mecánicas, rotaciones con
leguminosas y siembra en línea con densidad moderada.
Palabras clave: trigo duro, pérdidas de rendimiento, emergencia,
cosecha mecánica.
Resumo
Apesar da expansão da irrigação, os rendimentos de trigo duro
na Argélia permanecem substancialmente abaixo do potencial
genético. Este estudo quanticou as perdas na emergência, naturais e
mecânicas sob condições áridas irrigadas na província de Laghouat,
Argélia durante 2024-2025. Cinco campos experimentais (variedade
Vitron) foram monitorados, medindo densidades de emergência,
componentes do rendimento potencial, rendimento pré-colheita
e rendimento colhido. As perdas na emergência (24-43 %) foram
compensadas através do alhamento (r = -0,996; p ≤ 0,001), embora
densidades excessivas (> 600 sementes.m
-2
) aumentassem as perdas
sem benefício no rendimento. O rendimento potencial variou entre
9,3 e 13,7 t.ha
-1
, sendo o número de grãos por espiga o determinante
principal (r = 0,994; p ≤ 0,001). As perdas naturais (12-38 %) foram
amplicadas por condições meteorológicas extremas (granizo, chuva
de 36 mm.h
-1
, 14 de maio de 2025). As perdas mecânicas (31-47 %)
excederam os padrões internacionais (3-7 %). As perdas acumuladas
atingiram 54-60 % do potencial, valorizando entre 40 e 46 %. A
otimização deve priorizar a redução de perdas mecânicas, rotações
com leguminosas e semeadura em linha com densidade moderada.
Palavras-chave: trigo duro, perdas de rendimento, emergência,
colheita mecânica.
Introduction
Durum wheat (Triticum durum Desf.) is strategic for food
security in Algeria. Despite considerable investments in irrigated
area extension and cultivation modernization, average national
yields remain low (1.6-4.2 t.ha
-1
in arid regions), well below genetic
potential of improved varieties (6.0-8.0 t.ha
-1
) (Merouche et al.,
2014; MADR, 2021). This gap raises questions about loss magnitude
throughout the crop cycle. Agronomic literature identies three
critical loss phases: emergence losses from poor soil-seed contact,
inadequate sowing depth, or early water stress (Zhai et al., 2018);
natural losses during maturation from biotic and abiotic factors
(Grosse-Heilmann et al., 2024); and mechanical harvest losses
linked to improper combine settings (Kutzbach, 2000). However,
integrated quantication under irrigated Saharan conditions remains
largely lacking. Arid region durum wheat systems present marked
specicities: systematic sprinkler irrigation, sandy-loam soils with
low organic matter, extreme temperatures, and empirical practices.
Field observations report considerable mechanical harvesting losses
from operator training lack, machine obsolescence, and inadequate
settings, but no precise quantitative data exists. This study quanties
durum wheat yield losses at three key stages-emergence, maturation,
and mechanical harvest—under real irrigated Saharan conditions.
Specic objectives are: (1) quantify emergence loss rates and analyze
relationships with sowing practices; (2) estimate potential yield
and identify major productivity determinants; (3) quantify natural
losses and identify vulnerability factors; (4) quantify mechanical
harvest losses; (5) establish integrated loss balance and prioritize
improvement levers.
Materials and methods
Study location
Experimentation was conducted during the 2024-2025 season
in Ben Nacer Benchohra municipality, Laghouat province, southern
Algeria. This arid region presents annual precipitation below 200
mm, summer temperatures exceeding 40 °C, high evapotranspiration,
and predominantly sandy-loam soils with low organic matter content
(Hattab et al., 2025).
Plant material
The study focused on durum wheat (Triticum durum Desf.)
variety Vitron, widely adopted by farmers in the region. This
Spanish-origin variety presents medium to high straw height, a semi-
early cycle, and good adaptation to arid and semi-arid conditions
(Ladjal and Azouzi, 2014). Certied seeds, provided by the Algerian
Interprofessional Cereals Oce (OAIC), were treated with fungicide
coating. Thousand-grain weight (TGW) and germination capacity
(GC) presented some variability depending on the batches distributed
to farmers.
Characteristics of the experimental elds
Five durum wheat elds were randomly selected in the
municipality of Ben Nacer Benchohra. This multi-site approach allows
evaluating the impact of cultural practices under real production
conditions. For each eld, agronomic and cultural characteristics were
comprehensively recorded to correlate them with the results obtained
in the evaluated variables. Weather conditions being comparable
between sites, observed variations are mainly attributable to cultural
practices.
Emergence loss estimation
Emergence loss assessment was performed at the 3-leaf stage (3
to 4 weeks after sowing, BBCH 13 according to Hack et al. (1992),
the stage at which emergence is complete and stabilized. In each eld,
10 quadrats of 1 m
2
were placed randomly avoiding edges (> 5 m).
Seedling counting was performed by successive 20 cm sections using
two parallel guide sticks. Emergence density (ED, seedlings.m
-2
)
was calculated as the average of the 10 quadrats. Viable seed sowing
density (VSSD, seeds.m
-2
) was determined from the sowing rate,
corrected by the TGW and GC of the batch used:
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Hattab et al. Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331
3-7 |
The emergence loss rate (ELR, %) was calculated according to:
Potential yield estimation
Potential yield was estimated at the dough stage (BBCH 85
according to Hack et al. (1992), corresponding to physiological
maturity where the grain has reached its maximum dry matter weight
but still retains some moisture. In each eld, 10 quadrats of 1 m
2
were randomly placed avoiding edges and atypical zones (machinery
tracks, lodging areas, etc.). Measurement protocol: (1) Comprehensive
counting of fertile spikes.m
-2
(SN); (2) Random sampling of 10
spikes.quadrat.m
-2
; (3) Manual threshing and counting of grains
from each spike; (4) Calculation of average grain number.spike.m
-2
(GNS); (5) Calculation of grain number.m
-2
(GN.m
-2
= SN × GNS);
(6) Determination of grain weight.m
-2
from actual TGW measured
on samples before harvest: GW.m
-2
= (GN.m
-2
× TGW
A
)/1000; (7)
Extrapolation of potential yield (PY) per hectare:
Natural loss estimation
Natural losses, occurring between the dough stage and the day
before harvest, result from biotic factors (diseases, pests) and abiotic
factors (violent winds, excessive desiccation, precipitation). They
were quantied by the dierence between potential yield and pre-
harvest yield. One day before scheduled harvest (BBCH 92 stage
according to Hack et al., 1992), 200 spikes were randomly collected
at 10 locations per eld (20 spikes.location
-1
). Only spikes still carried
by plants were sampled. All grains were counted to calculate the
average grain number.spike.m
-1
before harvest (GNS
BH
). The actual
TGW (TGW
A
) was determined on three samples of 1000 grains. Pre-
harvest yield (PHY) was calculated according to this equation:
Natural losses (NL) and their rate (NLR) were calculated:
Mechanical harvest loss estimation
Mechanical harvest losses, mainly due to improper combine
harvester settings, were quantied by the dierence between pre-
harvest yield and actual harvested yield. Harvested yield (HY) was
obtained by farmer declaration at the end of harvest. Farmers of
the ve studied elds having delivered their harvest to the Algerian
Interprofessional Cereals Oce (OAIC) beneted from certied
scale weighing, guaranteeing good reliability and traceability of
actual harvested yield. This method reects actually valorized yield
and local yield measurement practices.
Mechanical harvest losses (ML) and their rate (MLR) were
calculated:
Total loss balance
Total losses (TL) from physiological maturity to the end of harvest
were calculated as the sum of natural losses and mechanical losses:
The total loss rate (TLR) relative to potential yield was expressed:
This balance allows quantifying the gap between productive
potential and valorized yield, and determining the relative contribution
of each loss type.
Statistical analyses
For each variable, means, standard deviations, and coecients
of variation were calculated. One-way ANOVA tested signicant
dierences between elds, with Tukey’s HSD test (p ≤ 0.05)
identifying homogeneous groups. Pearson correlation matrix was
established between 13 variables to identify signicant relationships
and prioritize yield determinants. Analyses were performed with
XLSTAT version 2025.1 (α = 0.05).
Results and discussion
Characteristics of the ve experimental durum wheat elds
The ve selected elds present commonalities guaranteeing
comparability and dierences allowing analysis of practice impacts
(Table 1). Areas vary from 2 to 4.5 ha (average 3.06 ha) with at
topography, eliminating slope eects. The Vitron variety was
cultivated in all elds, with homogeneous seed quality (TGW 50.8 ±
0.84 g, GC 94.8 ± 0.84 %). Sowing dates were grouped (November
15-20, 2024) except Field 4 (December 4). Harvest dates ranged from
June 25 to July 15, 2025.
Uniform practices included standardized soil tillage (plowing +
disc harrowing without nishing), organic manure application, basal
mineral fertilization, mobile sprinkler irrigation, no plant protection,
and mechanical harvest. Sowing rates varied considerably (2.22-4
q.ha
-1
), translating into densities of 435-800 seeds.m
-2
. Previous crops
showed diversity: durum wheat (Fields 1, 2, 4), alfalfa (Field 3), and
fallow (Field 5). Sowing methods diered: broadcast (Fields 1, 2, 3)
versus row seeder (Fields 4, 5). Top-dressing nitrogen was applied in
four elds (1, 2, 4, 5), but not Field 3. Chemical weeding occurred in
three elds (1, 2, 5).
Durum wheat emergence losses
As shown in Table 2, emergence assessment at the 3-leaf stage
revealed average densities ranging from 310.9 (Field 5) to 430.0
seedlings.m
-2
(Fields 3, 4), though variance analysis showed no
signicant dierences (p = 0.191) due to high intra-plot variability.
Field 4 (row seeded, moderate density) displayed the most regular
emergence, while Field 1 (broadcast, high density) showed highest
variability. The apparent homogeneity despite contrasting practices is
explained by irrigation compensating for unfavorable factors. Under
irrigated conditions, constant water availability attenuates the impact
of unfavorable factors such as poor soil-seed contact or inadequate
sowing depth (Zahiryan and Hamayoun, 2020).
Absolute losses showed highly signicant dierences (p ≤ 0.001),
varying from 98.1 (Field 5) to 330.0 seeds.m
-2
(Field 3), revealing
a density-dependent relationship: excessive density generates
increased losses without proportional benet (Mahdi et al., 1998).
Field 3 (800 seeds.m
-2
) required 1.77 seeds per emerged seedling
versus 1.32 for Field 5 (435 seeds.m
-2
), representing ~92 kg.ha
-1
additional cost. Relative loss rates varied from 24.0 % to 43.4 %,
remaining high compared to agronomic standards (10-20 % under
optimal conditions, Singh et al., 2025), explained by the absence of
soil nishing operations. The combination row seeding + moderate
density (Field 4: 600 seeds.m
-2
) appears optimal.
Potential yield and its components
Potential yield at dough stage (BBCH 85) represents maximum
theoretical yield absent losses between maturity and harvest (Table
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331 April-June ISSN 2477-9407.
4-7 |
Table 1. Agronomic characteristics and cultural practices of the experimental elds with durum wheat
Parameter Field 1 Field 2 Field 3 Field 4 Field 5 Mean ± SD
Area (ha) 3.3 2.5 3.0 2.0 4.5 3.06 ± 0.96
Topography Flat Flat Flat Flat Flat -
Previous crop Durum wheat Durum wheat Alfalfa Durum wheat Fallow -
Variety Vitron Vitron Vitron Vitron Vitron -
TGW (g) 52 51 50 50 51 50.8 ± 0.84
GC (%) 95 94 95 96 94 94.8 ± 0.84
Sowing date 20/11/2024 15/11/2024 20/11/2024 04/12/2024 20/11/2024 -
Sowing rate (q.ha
-1
) 3.64 3.00 4.00 3.00 2.22 3.17 ± 0.64
Sowing density
(seeds.m
-2
)
700 588 800 600 435 625 ± 137
Soil tillage P+DH P+DH P+DH P+DH P+DH -
Sowing method Broadcast Broadcast Broadcast Row seeder Row seeder -
Organic manure Yes Yes Yes Yes Yes -
Basal fertilizer Yes Yes Yes Yes Yes -
Top-dressing N Yes Yes No Ye s Ye s -
Chemical weeding Yes Yes No No Yes -
Plant protection No No No No No -
Irrigation MSS MSS MSS MSS MSS -
Harvest CH CH
CH CH CH -
Harvest date 05/07/2025 25/06/2025 15/07/2025 13/07/2025 11/07/2025 -
TGW: Thousand-Grain Weight; GC: Germination Capacity; P+H: Plowing + Disc Harrowing; MSS: Mobile Sprinkler System; CH: Combine Harvester; SD: Standard Deviation.
Table 2. Emergence densities and loss rates in the ve studied elds.
Field
SD
(seeds.m
-2
)
VSSD
(seeds.m
-2
)
ED
(seedlings.m
-2
)
CV (%)
NEVSSD
(seeds.m
-2
)
CV (%) ELR (%) CV (%)
1 700 665 383.56 ±
166.02
a
43.28 281.44 ±
166.02
ab
59 42.32 ± 24.97
a
59
2 588 553 383.33 ±
083.14
a
21.69 169.67 ±
083.14
bc
49 30.68 ± 15.03
a
49
3 800 760 430.00 ±
141.93
a
33.00 330.00 ±
141.93
a
43 43.42 ± 18.67
a
43
4 600 576 430.00 ±
072.33
a
16.82 146.00 ±
072.33
bc
49.54 25.35 ± 12.56
a
49.54
5 435 409 310.89 ±
082.32
a
26.48 098.11 ±
082.32
c
83.9 23.99 ± 20.13
a
83.9
ANOVA (p) - - 0.191
NS
- ≤ 0.001
***
- 0.089
NS
-
SD: Sowing Density; VSSD: Viable Seed Sowing Density; ED: Emergence Density; NEVSSD: Non-emerged Viable Seed Sowing Density; ELR: Emergence Loss Rate; CV: Coecient of
Variation; NS: Not Signicant (p > 0.05); ***: Highly Signicant (p ≤ 0.001). Superscript letters indicate homogeneous groups according to Tukey’s test (HSD) at 5 % threshold.
3). Fertile spike densities varied minimally (438-473 spikes.m
-2
)
with no signicant dierences (p = 0.954), contrasting sharply
with emergence density heterogeneity. This results from tillering
compensation (indices 1.08-1.41). Field 5, with lowest emergence
(311 seedlings.m
-2
), compensated through intense tillering (1.41) to
reach 438 spikes.m
-2
, while Fields 3-4 limited tillering (1.10, 1.08).
This phenotypic plasticity is well documented: low-density stands
favor secondary tiller emission thanks to lower competition for space,
light, and resources (Kondic et al., 2017; Liu et al., 2023).
Grain number.spike
-1
showed highly signicant dierences (p
≤ 0.001), varying from 38 (Field 2) to 53 grains.spike
-1
(Field 5).
Field 3 (alfalfa previous crop) displayed excellent spike fertility
(52), suggesting favorable nitrogen nutrition eects (Peoples et
al.
, 2009; Ni et al., 2026). Field 5 presented highest grain number.
spike
-1
(53), illustrating classic compensation (Slafer et al., 2014).
Grain number.m
-2
showed signicant dierences (p = 0.026), with
Field 3 clearly standing out (24264 grains.m
-2
), combining high spike
density and maximum grain number.spike
-1
. Potential yield showed
very signicant dierences (p = 0.006), varying from 9.29 (Field 2) to
13.65 t.ha
-1
(Field 3). Field 3 validated the favorable alfalfa previous
crop eect, reaching 13.65 t.ha
-1
without top-dressing nitrogen. Field
5 achieved comparable yield (13.40 t.ha
-1
) through dierent strategy:
low emergence → intense tillering → high spike fertility → maximum
TGW (57.97 g). Grain number.spike
-1
emerged as the major potential
yield determinant under these conditions.
Natural losses
Natural losses occur between dough stage and pre-harvest, from
biotic and abiotic factors. On May 14, 2025, exceptional weather
occurred: heavy hail and 36 mm.h
-1
intense rain, extreme for a region
averaging < 200 mm annually. This violent episode during critical
maturation caused mechanical shattering, lodging, and potential
sprouting (Bucheli et al., 2024). Spatial hailstorm heterogeneity
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Hattab et al. Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331
5-7 |
Table 3. Potential yield components of the ve studied elds.
Field SN.m
-2
CV (%) TI GNS CV (%) GN.m
-2
CV
(%)
TGW
A
(g)
CV
(%)
PY (t.ha
-1
)
CV
(%)
1 468 ± 114
a
24.36 1.22 41 ± 11.00
b
26.83 18935 ± 6137
ab
32.41 54.21 ± 0.77
c
1.42 10.26 ± 3.32
ab
32.41
2 462 ± 127
a
27.49 1.20 38 ± 04.00
b
10.53 17117 ± 4183
b
24.44 54.28 ± 0.43
c
0.79 09.29 ± 2.27
b
24.44
3 473 ± 117
a
24.74 1.10 52 ± 07.00
a
13.46 24264 ± 6292
a
25.93 56.29 ± 0.13
b
0.23 13.65 ± 3.54
a
25.93
4 464 ± 300
a
64.66 1.08 46 ± 05.00
ab
10.87 21057 ± 2357
ab
11.19 55.85 ± 0.21
b
0.38 11.76 ± 1.31
ab
11.20
5 438 ± 109
a
24.89 1.41 53 ± 04.00
a
07.55 23126 ± 6328
ab
27.36 57.97 ± 0.65
a
1.21 13.40 ± 3.66
a
27.36
ANOVA (p) 0.954
NS
-
-
≤ 0.001*** - 0.026* - ≤ 0.001*** - 0.006** -
SN: Average Spike Number; TI: Tillering Index; GNS: Average Grain Number.spike
-1
; GN: Average Grain Number; CV: Coecient of Variation; PY: Potential Yield; TGW
A
: Actual Thousand-
Grain Weight (1 day before harvest); NS: Not Signicant (p > 0.05); *: Signicant (p ≤ 0.05); **: Very Signicant (p ≤ 0.01); ***: Highly Signicant (p ≤ 0.001). Superscript letters indicate
homogeneous groups according to Tukey’s test (HSD) at 5 % threshold.
Table 4. Pre-harvest yield components and natural losses.
Field GNS
BH
GN.m
-2
BH
CV
(%)
TGW
A
(g)
CV
(%)
PHY (t.ha
-1
)
CV
(%)
NL (t.ha
-1
) CV (%) NLR (%) CV (%)
1 25 11688 ± 2847
c
24.36 54.21 ± 0.77
c
1.42 06.33 ± 1.54
c
24.35 3.92 ± 1.54
a
39.28 38.27 ± 15.04
a
39.30
2 27 12461 ± 3421
c
27.45 54.28 ± 0.43
c
0.79 06.76 ± 1.85
c
27.45 2.52 ± 1.85
a
73.49 27.20 ± 19.98
ab
73.46
3 45 21285 ± 5281
a
24.81 56.29 ± 0.13
b
0.23 11.98 ± 2.97
a
24.81 1.67 ± 2.97
a
177.28 12.28 ± 21.77
b
177.28
4 33 15296 ± 1002
bc
06.55 55.85 ± 0.21
b
0.38 08.54 ± 0.56
bc
06.55 3.21 ± 0.56
a
17.41 27.36 ± 04.76
ab
17.40
5 40 17500 ± 4365
ab
24.94 57.97 ± 0.65
a
1.21 10.14 ± 2.53
ab
24.94 3.26 ± 2.53
a
77.58 24.33 ± 18.87
ab
77.56
ANOVA (p) - ≤ 0.001*** - ≤ 0.001*** - ≤ 0.001*** - 0.165
NS
- 0.031* -
GNS
BH
: Average Grain Number.spike
-1
Before Harvest; GN.m
-2
BH
: Average Grain Number.m
-2
Before Harvest; TGW
A
: Actual Thousand-Grain Weight; PHY: Pre-Harvest Yield; NL: Natural Losses;
NLR: Natural Losses Rate; CV: Coecient of Variation; NS: Not Signicant (p > 0.05); *: Signicant (p ≤ 0.05); ***: Highly Signicant (p ≤ 0.001). Superscript letters indicate homogeneous
groups according to Tukey’s test (HSD) at 5 % threshold.
Natural loss rates showed signicant dierences (p = 0.031),
varying from 12.28 % (Field 3) to 38.27 % (Field 1). Field 3’s
remarkable resilience (12.28 %) is attributable to alfalfa protective
eects (soil structure, root anchorage, lodging resistance). Field
1’s extreme vulnerability (38.27 %) resulted from phenological
heterogeneity (CV 43.3 % emergence), lodging sensitivity from
broadcast high-density seeding (700 seeds.m
-2
), and possible maximum
hail exposure. Observed rates (12-38 %) must be interpreted within
the May 14 extreme event context and don’t represent «normal»
losses but illustrate vulnerability to extreme climatic hazards.
Mechanical harvest losses and valorized yield
Mechanical losses occur during combine harvesting. Actually
harvested yield (HY), obtained via OAIC certied weighing, varied
from 41 (Field 1) to 6.0 t.ha
-1
(Field 3) (Table 5). Field 3 maintained
dominance with 6.0 t.ha
-1
, from exceptional potential (13.65 t.ha
-1
) and
low natural loss rate (12.28 %). Fields 1-2 showed lowest yields (4.1,
4.3 t.ha
-1
) from cumulative moderate potential, high natural losses
(38.27 %, 27.20 %), and signicant mechanical losses (31.89 %,
31.38 %). Field 1 ultimately valorized only 40 % of initial potential.
Absolute mechanical losses showed highly signicant dierences
(p ≤ 0.001), varying from 2.23 to 5.98 t.ha
-1
. Field 3 presents an
apparent paradox: highest harvested yield yet most signicant absolute
losses (5.98 t.ha
-1
), explained by exceptionally high pre-harvest yield
(Zhou and Vilar-Zanon, 2025) partly explains high loss variability.
Comparative analysis revealed widespread grain number.spike
-1
decreases (Table 4), varying from 7 grains.spike
-1
(Field 3, 13.5 %) to
16 grains.spike
-1
(Field 1, 39 %). Field 1 recorded severe degradation
(41 to 25 grains.spike
-1
), causing yield collapse to 6.33 t.ha
-1
versus
10.26 t.ha
-1
potential. Field 3, despite highest potential (13.65 t.ha
-1
),
maintained 11.98 t.ha
-1
pre-harvest yield, suggesting better structural
resistance from alfalfa previous crop (Peoples et al., 2009).
(11.98 t.ha
-1
). This demonstrates absolute mechanical losses increase
proportionally to available yield (Kutzbach, 2000). Mechanical loss
rates varied from 31.38 % to 47.40 %, far exceeding agronomic
standards (3 % Delchev and Trendalov, 2013; 2.5 % Al-Sammarraie
and Alhadithi, 2021; 6.83 %, Abdulla et al., 2025). Field 3 presented
highest mechanical rate (47.40 %), possibly from very high pre-
harvest yield exceeding combine capacity, generating clogging losses
(Kutzbach, 2000), or improper settings in dense productive stands.
Although harvest occurred under dry conditions, the May 14 episode
had lasting repercussions: partial lodging generating increased losses
from cutting diculty, cutter bar clogging, and ground grain losses
(Dahiya et al., 2018; Yadav et al., 2024). High losses also result
from improper combine settings, operator training lack, and machine
obsolescence. Cumulative natural and mechanical losses lead to
dramatic total loss rates of 54-60 %, valorizing only 40-46 % of
potential. Priority improvements include: operator training programs
on optimal combine settings; nely adapting machine parameters to
eld yield; renewing obsolete combines; adapting cutting height and
speed in lodging cases; determining optimal harvest date balancing
natural and mechanical loss reduction.
Correlations between studied variables
Correlation analysis between 13 variables reveals key relationships
(Table 6). Sowing density correlated with spike number.m
-2
(r
= 0.933, p ≤ 0.05) and emergence loss rate (r = 0.897, p ≤ 0.05),
conrming excessive densities (> 600 seeds.m
-2
) generate increased
losses without benet. Tillering index and emergence density showed
almost perfect negative correlation (r = -0.996, p ≤ 0.001), conrming
phenotypic compensation explaining why elds reached similar spike
densities (438-473 spikes.m
-2
) despite contrasting emergence (311-
430 seedlings.m
-2
).
Grain number.spike
-1
emerged as major yield determinant (r =
0.994, p ≤ 0.001), with strong correlations to grain number.m
-2
(r =
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331 April-June ISSN 2477-9407.
6-7 |
Table 5. Harvested yields and mechanical harvest losses.
Field PY (t.ha
-1
) PHY (t.ha
-1
) HY (t.ha
-1
) ML (t.ha
-1
) CV (%) MLR (%) CV (%) TL (t.ha
-1
)
1 10.26 06.33 4.1 2.23 ± 1.54
b
69.00 31.89 ± 15.70
a
49.23 6.16
2 09.29 06.76 4.3 2.46 ± 1.85
b
75.37 31.38 ± 21.09
a
67.21 4.99
3 13.65 11.98 6.0 5.98 ± 2.97
a
49.71 47.40 ± 11.52
a
24.30 7.65
4 11.76 08.54 5.0 3.54 ± 0.56
ab
15.81 41.25 ± 03.78
a
09.16 6.76
5 13.40 10.14 5.6 4.54 ± 2.53
ab
55.67 41.26 ± 16.27
a
39.43 7.80
ANOVA (p) 0.006** ≤ 0.001*** - ≤ 0.001*** - 0.092
NS
- -
PY: Potential Yield; PHY: Pre-Harvest Yield; HY: Harvested Yield; ML: Mechanical Losses; MLR: Mechanical Loss Rate; TL: Total Losses; CV: Coecient of Variation; NS: Not Signicant
(p > 0.05); **: Very Signicant (p ≤ 0.01); ***: Highly Signicant (p ≤ 0.001). Superscript letters indicate homogeneous groups according to Tukey’s test (HSD) at 5 % threshold.
Table 6. Correlation between studied variables.
Variables SD ED ELR TI SN GNS GN TGW
A
PY PHY NLR MLR HY
SD
1
0.783
0.897
*
-0.739
0.933
*
-0.100 0.096 -0.459 -0.002 0.154 -0.245 0.196 0.044
ED
1
0.436
-0.996
***
0.898
*
-0.177 0.006 -0.429 -0.077 0.085 -0.246 0.253 0.018
ELR
1
-0.378 0.746 -0.117 0.039 -0.452 -0.045 0.055 -0.078 0.006 -0.058
TI
1 -0.880
*
0.222 0.047 0.451 0.126 -0.043 0.227 -0.215 0.019
SN
1
-0.370 -0.176 -0.669 -0.271 -0.101 -0.070 -0.005 -0.199
GNS
1 0.979
**
0.928
*
0.994
***
0.919
*
-0.673
0.898
*
0.944
*
GN
1
0.838
0.995
***
0.947
*
-0.718
0.948
*
0.952
*
TGW
A
1 0.888
*
0.779 -0.546 0.743 0.841
PY
1 0.941
*
-0.707
0.932
*
0.956
*
PHY
1 -0.902
*
0.960
**
0.993
***
NLR
1
-0.833
-0.878
*
MLR
1 0.959
**
HY
1
* p≤0.05; ** p≤0.01; *** p≤0.001. SD: Sowing Density; ED: Emergence Density; ELR: Emergence Loss Rate; TI: Tillering Index; SN: Spike Number.m
-2
; GNS: Grain Number/Spike; GN: Grain
Number.m
-2
; TGW
A
: Actual Thousand-Grain Weight; PY: Potential Yield; PHY: Pre-Harvest Yield; NLR: Natural Loss Rate; MLR: Mechanical Loss Rate; HY: Harvested Yield.
0.979, p ≤ 0.01), TGW
A
(r = 0.928, p ≤ 0.05), and pre-harvest yield (r
= 0.919, p ≤ 0.05). Grain number.m
-2
correlated almost perfectly with
potential yield (r = 0.995, p ≤ 0.001), mechanical losses (r = 0.948,
p≤0.05), and harvested yield (r = 0.952, p ≤ 0.05). This hierarchy
(GNS > GN.m
-2
> TGW
A
) conrms grain number as the main lever.
Potential yield correlated strongly with pre-harvest yield (r = 0.941,
p≤0.05) and absolute mechanical losses (r = 0.932, p ≤ 0.05), showing
high-yield elds generate more losses due to large biomass exceeding
combine capacity. Pre-harvest yield best predicts harvested yield (r =
0.993, p ≤ 0.001). Natural and mechanical loss rates weren’t correlated
(r = -0.833), indicating independent determinants. Emergence loss
rate wasn’t correlated with any yield variable, conrming tillering
compensation neutralized its impact—emergence optimization
targets economic eciency rather than yield improvement.
Conclusion
This study reveals catastrophic cumulative losses of 54-60 %
of durum wheat potential yield in Laghouat Province (2024-2025).
Emergence losses (24-43 %) are compensated by tillering, though
excessive densities (> 600 seeds.m
-2
) waste seeds. Grain number.spike
-1
is the major determinant. Natural losses (12-38 %) were amplied by
extreme weather, showing 38.27 % vulnerability in monoculture versus
12.28 % after alfalfa. Mechanical losses (31-47 %) exceed international
standards (3-7 %) and constitute the major bottleneck. Optimization
must prioritize legume rotations, row seeding at 500-600 seeds.m
-2
,
mechanical loss reduction through operator training, and climate
adaptation strategies for Saharan cereal food security.
Literature cited
Abdulla, Z. B., Mohammed, S. H., Youns, A. F., & Isaak, M. (2025). Measuring
Losses of Wheat Harvester Units at Dierent Speeds and Cutting Heights.
Kirkuk University Journal for Agricultural Sciences, 240-248. https://doi.
org.10.58928/ku25.16426
Al-Sammarraie, M. A. J., & Alhadithi, S. A. (2021). Determination of Grain Losses
on Combine Harvester. Journal of Scientic and Engineering Research,
8(1), 196-202. https://doi.org/10.5281/zenodo.10552653
Bucheli, J., Visse-Mansiaux, M., Herrera, J., Haner, L. L., Tack, J., & Finger, R.
(2024). Precipitation causes quality losses of large economic relevance in
wheat production. Q Open, 4(1), qoae008. https://doi.org/10.1093/qopen/
qoae008
Dahiya, S., Kumar, S., Chaudhary, H., & Chaudhary, C. (2018). Lodging:
Signicance and preventive measures for increasing crop production.
International Journal of Chemical Studies, 6(1), 700-705. https://www.
chemijournal.com/archives/2018/vol6issue1/PartJ/5-6-322-652.pdf
Delchev, N., & Trendalov, K. (2013). Method for rapid determination of the
percentage rate of grain losses by the combine harvester according to its
parameters. Agricultural Science and Technology, 5(1), 62-64. https://
www.cabidigitallibrary.org/doi/pdf/10.5555/20133136521
Grosse-Heilmann, M., Cristiano, E., Deidda, R., & Viola F. (2024). Durum wheat
productivity to day and tomorrow: A review of inuencing factors and
climatec hange eects. Resources, Environment and Sustainability, 17,
100170. https://doi.org/10.1016/j.resenv.2024.100170
Hack, H., Bleiholder, H., Buhr, L., Meier, U., Schnock-Fricke, U., Weber, E., &
Witzenberger, A. (1992). Einheitliche Codierung der phänologischen
Entwicklungsstadien mono- und dikotyler Panzen - Erweiterte BBCH-
Skala, Allgemein. Nachrichtenbl. Deut. Panzenschutzd, 44(12), 265-
270. https://api.agrometeo.ch/storage/uploads/bbch-skala_deutsch.pdf
Hattab, M., Kessaissia, A., & Boughar, S. (2025). Situation de la luzerne pérenne
dans une région agropastorale aride. Fourrages, 261, 19-29. https://doi.
org/10.64256/FOU2449
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Hattab et al. Rev. Fac. Agron. (LUZ). 2026, 43(2): e264331
7-7 |
Kondic, D., Bajic, M., Hajder, D., & Bosancic, B. (2017). The Rate of Productive
Tillers per Plant of Winter Wheat (Triticum aestivum L.) Cultivars under
Dierent Sowing Densities. Agro-knowledge Journal, 17(4), 345-357.
https://doi.org/10.7251/AGREN1604345K
Kutzbach, H. D. (2000). Trends in Power and Machinery. Journal of Agricultural
Engineering Research, 76(3), 237-247. https://doi.org/10.1006/
jaer.2000.0574
Ladjal, I., & Azouzi, B. (2014). Etude du comportement varietal de cinq varietes de
ble dur sous l’eet des deux doses de semis dierentes en environnement
semi-aride de Djelfa. Sciences and Technologie, 40, 36-42. https://revue.
umc.edu.dz/c/article/view/1689/1809
Liu, P., Yin, B., Liu, X., Gu, L., Guo, J., Yang, M., & Zhen, W. (2023). Optimizing
plant spatial competition can change phytohormone content and promote
tillering, thereby improving wheat yield. Frontiers in Plant Science, 14,
1147711. https://doi.org/10.3389/fpls.2023.1147711
MADR, Ministère de l’Agriculture et du Développement Rural. (2021). Direction
des Statistiques Agricoles et des Systèmes d’Information. Supercies
et productions. Alger. Algérie. 87 p. https://madr.gov.dz/wp-content/
uploads/2022/04/SERIE-B-2019.pdf
Mahdi, L., Bell, C. J., & Ryan, J. (1998). Establishment and yield of wheat
(Triticum turgidum L.) after early sowing at various depths in a semi-
arid Mediterranean environment. Field Crops Research, 58(3), 187-196.
https://doi.org/10.1016/S0378-4290(98)00094-X
Merouche, A., Debaeke, P., Messahel, M., & Kelkouli, M. (2014). Response of
durum wheat varieties to water in semi-arid Algeria. African Journal
of Agricultural Research, 9(38), 2880-2893. https://doi.org/10.5897/
AJAR2013.7318
Ni, H., Chu, H. K., Zhang, Z. X., Lai, X. F., Yang, X. L., & Shen, Y. Y. (2026).
Sustainable alfalfa-winter wheat rotation maintains yield and soil water
balance under precipitation variation. European Journal of Agronomy,
172, 127824. https://doi.org/10.1016/j.eja.2025.127824
Peoples, M. B., Brockwell, J., Herridge, D. F., Rochester, I. J., Alves, B. J. R.,
Urquiaga, S., Boddey, R. M., Dakora, F. D., Bhattarai, S., Maskey, S.
L., Sampet, C., Rerkasem, B., Khans, D. F., Hauggaard-Nielsen, H., &
Jensen, E. S. (2009). The contributions of nitrogen-xing crop legumes to
the productivity of agricultural systems. Symbiosis, 48, 1-17. https://doi.
org/10.1007/BF03179980
Singh, B., Mukherjee, A., Norton, R., & Elshikha, D. E. (2025). Preliminary
Evaluation of Wheat, Barley and Triticale Varieties at the Maricopa
Agricultural Center for Early to Mid-Season 2024–2025. The University of
Arizona Cooperative Extension, 6 p. https://hdl.handle.net/10150/678672
Slafer, G. A., Savin, R., & Sadras, V. O. (2014). Coarse and ne regulation of
wheat yield components in response to genotype and environment. Field
Crops Research, 157, 71–83. https://doi.org/10.1016/j.fcr.2013.12.004
Yadav, K., Kumar, M., Gulaiya, S., Singh, N., Singh, S., Salar, A., Joshi, M., Pal,
S. S., & Singh, P. P. (2024). Adverse impacts of lodging and strategies
for management in cereal crops: a comprehensive review. Plant Archives,
24(2), 495-503. https://doi.org/10.51470/PLANTARCHIVES.2024.v24.
no.2.069
Zahiryan, R., & Hamayoun, H. (2020). Eect of irrigation frequency on growth
and yield of wheat (Triticum aestivum L. var. Chonte 1) under Kabul agro-
climatic conditions, Afghanistan. e-planet, 18(1), 22-28. https://e-planet.
co.in/images/Publication/vol-18-1/eect_of_irrigation.pdf
Zhai, Y., Wu, Q., Chen, G., Zhang, H., Yin, X., & Chen, F. (2018). Broadcasting
Winter Wheat Can Increase Grain Yield without Reducing the Kernels
per Spike and the Kernel Weight. Sustainability, 10(12), 4858. https://doi.
org/10.3390/su10124858
Zhou, N., & Vilar-Zanon, J. L. (2025). Climate change and crop insurance:
geographical heterogeneity in hailstorm risk for wine grapes in Spain.
European Actuarial Journal. https://doi.org/10.1007/s13385-025-00419-6