Ti le
1
A e seed and in o ma ion ci cula ion ne wo ks ela ed o he c op di e si y and esilience
2
o smallholde a ms? A s udy p o ocol
3
Au ho s
4
Vanesse Labey ie*1,2, Julie e Ma iel*1,2, Ka ina Flo ide Olodo1,2, Da id And es Be nal Hoyo1,2,
5
Ga ance Danne 1,2, Océane Cobelli1,2, Flo ence Lebeaupin1,2, Ewen Menguy1,2, Wissal Msellek3,
6
Si akiniaina Cla a Rake abakoly4, Jus ine S u z1,2, Ndeye Fa ou Faye5, Samba Sy5, Malamine J. Badji5,
7
Salama El Fa ehi3, Younès Hmimsa3, Ve ohani a Ra idison4, Josoa Rand iamalala6, Adeline Ba naud7,
8
S éphanie M. Ca iè e8, F édé ique Jankowski1,2, Ch is ian Lecle c9,10, Delphine Rena d11, Isabelle
9
Sanchez12, I ene Teixido -Toneu13, Nicolas Ve zelen12
10
*Vanesse Labey ie and Julie e Ma iel a e bo h i s and co esponding au ho s.
11
1 CIRAD, UMR SENS, F-34398, Mon pellie , F ance.
12
2 SENS, Uni Mon pellie , CIRAD, IRD, UPVM, Mon pellie , F ance.
13
3 TEDAEEP Team Resea ch, Abdelmalek Essaadi Uni e si y, Polydisciplina y Facul y o La ache – (UAE-FPL),
14
La ache, 92000, Mo occo.
15
4 Facul é des Sciences, Uni e si é d’An anana i o, BP 906 An anana i o 101, Madagasca .
16
5 Ins i u Sénégalais de Reche ches Ag icoles, Bu eau d'Analyses Mac oéconomiques, Daka , Sénégal.
17
6 Uni e si é d'An anana i o, École Supé ieu e des Sciences Ag onomiques, BP 175, Anka so, An anana i o 101,
18
Madagasca .
19
7 DIADE, Uni e si y o Mon pellie , IRD, CIRAD, Mon pellie , F ance.
20
8 IRD UMR SENS, CIRAD, IRD, UPVM, Uni Mon pellie , Mon pellie , F ance.
21
9 CIRAD, UMR AGAP Ins i u , F-34398, Mon pellie , F ance.
22
10 UMR AGAP Ins i u , Uni Mon pellie , CIRAD, INRAE, Ins i u Ag o, F-34398 Mon pellie , F ance.
23
11 CEFE, Uni Mon pellie , CNRS, EPHE, IRD, Mon pellie , F ance.
24
12 MISTEA, INRAE, Ins i u Ag o, Uni e si y o Mon pellie , Mon pellie , F ance.
25
13 IMBE, Aix Ma seille Uni e si y, A ignon Uni e si y, CNRS, IRD, Ma seille, F ance.
26
27
Abs ac
28
Smallholde a ming is p edominan in egions ha a e pa icula ly subjec o clima e change
29
wo ldwide, especially d ylands, and pa icula ly su e om an inc ease in clima ic shocks. In his
30
con ex , c op di e si y is conside ed as an impo an asse o he esilience o smallholde s who ha e
31
limi ed inancial and ma e ial capi als, in as uc u es and echnologies. Howe e , knowledge is limi ed
32
abou he p ocesses enhancing smallholde s’ access o c op di e si y. Seed and associa ed in o ma ion
33
ci cula ion pa e ns a e conside ed as a majo d i e o c op di e si y a ailabili y and access, al hough
34
which ci cula ion pa e ns enhance he a ailabili y o a la ge di e si y pool a he collec i e le el and
35
i s access o a me s emains ma e o deba e.
36
The esea ch p o ocol p esen ed he e, designed in he ARISER p ojec (Access o c op di e si y and
37
small a ms’ esilience o clima e a iabili y in A ican d ylands: The ole o seed and in o ma ion
38
ne wo ks) unded by he Eu opean Resea ch Council, is de eloped o add ess hese c i ical gaps. This
39
p o ocol was designed o collec ing compa able longi udinal and panel da a ac oss di e se con ex s o
40
assess: (i) di e ences in a ms’ seed and associa ed in o ma ion ne wo ks, and hei socioeconomic
41
d i e s; (ii) whe he hese di e ences in a m’s ne wo ks ela e o di e ences in hei le el o c op
42
di e si y and empo al u no e , (iii) whe he hey ela e o plu iannual indica o s o a ms’ esilience
43
o shocks, mainly he empo al s abili y o c op p oduc ion and households' abili y o mee hei ood
44
and inancial needs o e ime.
45
ARISER p o ocol is he i s o his kind, co e ing ou consecu i e yea s and 2000 households in h ee
46
di e en al hough compa able s udy si es. We hus belie e ha sha ing his p o ocol and he challenges
47
we me along i s design, es and implemen a ion can be o in e es o he design o u u e esea ch in
48
his domain and wo h o be sha ed wi h ou esea ch communi y. This p o ocol is expec ed o imp o e
49
he gene al unde s anding o how seed and ela ed in o ma ion ne wo ks a ec smallholde s’ abili y o
50
use c op di e si y in he ace o dis u bances, and o which ex en his is media ed by socioeconomic
51
inequali ies. Such unde s anding is needed o imp o e decision-making in he seed sec o and be e
52
add ess he speci ic challenges ha smallholde s ace in adap ing o global changes.
53
Keywo ds
54
Resea ch p o ocol, c op di e si y, seed ci cula ion, esilience, smallholde s
55
56
57
58
59
60
In oduc ion
61
Smallholde a ming is he dominan o m o ag icul u e in a la ge pa o he wo ld, including A ica,
62
Sou h-Eas Asia, La in Ame ica, and opical Islands (Sambe g e al. 2016). Smallholde a me s
63
gene ally cul i a e small a eas and ha e limi ed access o in as uc u e and o echnologies including
64
mechaniza ion, chemical inpu s, and i iga ion because o hei low inancial capi al. Despi e hose
65
limi ed asse s, hey p oduce a subs an ial pa o he wo ld’s ood while ha ing limi ed en i onmen al
66
impac s as hey mos ly p ac ice ag oecological cul i a ion. Recen es ima ions indica e ha a ms o
67
less han i e ha p oduce an es ima ed hal o he wo ld’s ood calo ies (Riccia di e al. 2018). Small
68
a ms hus can play a key ole in sus aining ood secu i y while p ese ing na u al esou ces and
69
biodi e si y in he ace o global challenges such as adap a ion o clima e change (Al ie i and Nicholls
70
2017).
71
Smallholde s la gely ely on c op di e si y o hei li elihoods, as i is one o he main esou ces hey
72
can mobilize o ensu e p oduc ion wi hou subs an ial inancial in es men (Ce igni and Mo is 2016).
73
A la ge pa o smallholde li es in a eas whe e he a iabili y o he biophysical and socioeconomic
74
en i onmen is high. Fo ins ance, smallholde ag icul u e is dominan in he d ylands, whe e clima ic
75
a iabili y is high and is expec ed o inc ease d ama ically in he u u e wi h majo impac s on
76
ag icul u e (Knox e al. 2012, Po e e al. 2017, Co ell e al. 2019). In hese a eas, cul i a ing a
77
di e si y o c op species and a ie ies is equen ly men ioned as a way o cope wi h he highly a iable
78
en i onmen al condi ions as i s eng hens households’ capaci y o mee hei ood and inancial needs
79
in he ace o pe u ba ions (Al ie i and Nicholls 2017, Ve nooy 2022). Access o c op di e si y and o
80
he in o ma ion associa ed o he cha ac e is ics o he di e en c ops is expec ed o be c ucial o he
81
esilience o smallholde s o en i onmen al a iabili y o wo main easons. Fi s , he cul i a ion o
82
di e en c op species and a ie ies allows o bu e en i onmen al a ia ions o di e en kind,
83
especially clima ic o economic ones (Rena d and Tilman 2021, Lo eau e al. 2021, Di Falco and Cha as
84
2008, Pascual, Jackson and D ucke 2013). Second, he capaci y o a ms o change c op species o
85
a ie ies is a key o adap ing o changing condi ions (Coope e al. 2008, Magesa e al. 2023, Labey ie
86
e al. 2021).
87
To be able o di e si y o change hei c ops, a me s need access o adap ed plan ing ma e ial, as seeds,
88
cu ings, o o he p opagules. To selec adap ed species and a ie ies, hei own knowledge as well as
89
ex e nal in o ma ion on c op cha ac e is ics a e c ucial. Such in o ma ion encompasses o ins ance
90
c op a ie ies’ g ow h-cycle leng h ha condi ion hei abili y o wi hs and d ough , o hei cu en
91
p ice on local ma ke . Howe e , despi e hei ob ious impo ance in smallholde s’ esilience o
92
en i onmen al a iabili y, he p ocesses d i ing hei access o c op di e si y and in o ma ion emain
93
la gely igno ed in ag icul u al policies and de elopmen p og ams (Coomes e al. 2015). Ra he ,
94
de elopmen p og ams in A ica and wo ldwide suppo he specializa ion and uni o miza ion o c ops
95
om he plo o he egional le el, which inc eases he ulne abili y o smallholde a ming o clima e
96
a iabili y and has majo implica ions o ood secu i y a he local, na ional and global le els (Khou y
97
e al. 2014, Rena d and Tilman 2019). This uni o miza ion is pa icula ly sus ained by he cen aliza ion
98
o seed and in o ma ion ci cula ion, which also limi s oppo uni ies o de elop local ag oecological
99
inno a ions and adap a ions (Vanloque en and Ba e 2017, Coolsae 2015).
100
Smallholde s ace he sha ed challenge o gaining g ea e ecogni ion o he plu ali y o hei seed
101
sys ems, and bene i ing om go e nance o seed sys ems ha is be e sui ed o hei needs (Coomes e
102
al. 2015, Wes engen, Dalle and Mulesa 2023). Seed sys ems a e de ined as “combined ac i i ies o
103
ac o s, making use o plan ma e ials and knowledge, ha oge he a e necessa y o suppling seeds o
104
a me s” (C omwell e al. 1993; Louwaa s e Manicad 2022). Smallholde seed sys ems in ol e a ange
105
o s akeholde s and ins i u ions o di e en kinds: a me s and local seed selle s on weekly ma ke s and
106
pee s (McGui e e Spe ling 2016). Ye , in mos coun ies whe e smallholde a ming is p edominan ,
107
na ional public policies, legal amewo ks, and de elopmen p og ams led by in e na ional agencies
108
end o ocus on cen alizing seed p oduc ion and dis ibu ion (Scoones and Thompson 2011, Louwaa s,
109
de Boe and Edeme 2013). These ini ia i es aim, on one hand, o s eng hen s a e p og ams o b eeding
110
and dissemina ing new a ie ies, and on he o he hand, o de elop he p i a e seed sec o , bo h o which
111
p ima ily dis ibu e ce i ied seeds om imp o ed a ie ies o a limi ed numbe o c op species (Mulesa
112
e al. 2021). This aises majo conce ns among schola s and a me s' o ganiza ions, who a gue ha hese
113
ac ions do no adequa ely add ess he speci ic needs o smallholde s in e ms o c op seed di e si y
114
(Coomes e al. 2015, Pescha d and Rande ia 2020). Calls especially mul iply on he necessi y o
115
conside he plu ali y o seed sys ems, and a o he eme gence o new o ganiza ion modali ies be ween
116
ac o s, accoun ing o local o ganiza ion modali ies, o enhancing smallholde access o seed di e si y
117
(Mulesa e al. 2021, Be he e al. 2020). De e mining which seed sys em o ganiza ion is be e sui ed
118
in his pe spec i e is a key poin o deba e (Wes egen, Dalle and Mulesa 2023, Louwaa s and Manicad
119
2022).
120
The ci cula ion o seed and associa ed in o ma ion has especially been iden i ied as po en ially key o
121
he main enance and dynamics o c op di e si y in a ms (Labey ie e al. 2021b). Subs an ial esea ch
122
has ocused on analyzing seed ci cula ion pa e ns in smallholde ag icul u e, and o a lesse ex en
123
hose o associa ed in o ma ion. Ye , a hough ul unde s anding o how hese pa e ns ela e o
124
a ms’ c op di e si y and ela ed esilience ou comes is lacking. Social ne wo k analysis
125
(Wasse man and Faus 1994) has been applied o analyze seed ci cula ion in smallholde communi ies,
126
and o a lesse ex en , he ci cula ion o associa ed in o ma ion (Box 1). Howe e , he ex en o which
127
di e en seed and in o ma ion ci cula ion pa e ns ela e o a m-le el ou comes—such as c op
128
di e si y, empo al u no e , and he main enance o key unc ions unde pe u ba ions— emains
129
poo ly unde s ood (Labey ie e al. 2021b).
130
The esea ch p o ocol p esen ed in his a icle was designed o enhance ou unde s anding o how seed
131
and in o ma ion ci cula ion pa e ns ela e o c op di e si y le el and empo al u no e in small a ms,
132
and o indica o s o a ms’ esilience o pe u ba ions. Addi ionally, i seeks o unde s and which
133
socioeconomic ac o s enhance o limi a me s capaci y o access c op di e si y h ough hese
134
ne wo ks. De eloped wi hin he ARISER p ojec (Access o C op Di e si y and Small Fa ms’
135
Resilience o Clima e Va iabili y in A ican D ylands: The Role o Seed and In o ma ion Ne wo ks)
136
unded by he Eu opean Resea ch Council (S a ing G an 2022-2027), his p o ocol was designed o
137
collec ing compa able longi udinal and panel da a ac oss di e se con ex s o assess: (i) di e ences in
138
a ms’ seed and associa ed in o ma ion ne wo ks, and hei socioeconomic d i e s; (ii) whe he hese
139
di e ences in a m’s ne wo ks ela e o di e ences in hei le el o c op di e si y and empo al
140
u no e , (iii) whe he hey ela e o plu iannual indica o s o a ms’ esilience o shocks, mainly he
141
empo al s abili y o c op p oduc ion and households' abili y o mee hei ood and inancial needs o e
142
ime.
143
This a icle aims a sha ing his esea ch p o ocol wi h he esea ch communi y on seed sys ems, and
144
p o ide eedback on i s implemen a ion du ing he ARISER p ojec . This uni ied and adap able p o ocol
145
is expec ed o b ing a me hodological con ibu ion o esea ch on smallholde seed sys ems, helping o
146
d aw gene al conclusions on he s udied ela ionships. We expec ha compa ing esul s ob ained
147
h ough a s anda dized me hod ac oss h ee s udy si es, loca ed in di e en egions bu in simila
148
ag oecosys ems, will imp o e he gene al unde s anding o hese linkages and ha publishing his
149
p o ocol will s imula e he collec ion o compa able da a ac oss a ange o con ex s.
150
151
152
153
154
155
156
157
158
159
160
161
162

163
Box 1. Applica ions o Social Ne wo k Analysis o he analysis o seed and associa ed in o ma ion
164
ci cula ion
165
166
Hypo heses, esea ch s a egy and e hics
167
Concep ual amewo k
168
The gene al objec i e o his p o ocol is o es ablish he i s sys ema ical assessmen o he ela ionship
169
be ween seed and in o ma ion ci cula ion pa e ns, c op di e si y and indica o s o a m’s esilience in
170
A ican d yland a eas, subjec o high clima e a iabili y. Fo his pu pose, i combines concep and
171
me hods om social and ag icul u al sciences, especially: (i) social ne wo k analysis, ha s udies
172
in e ac ions be ween social ac o s (i.e. indi iduals o o ganiza ions) and how hey a e in luenced by
173
ac o s’ socioeconomic cha ac e is ics, as well as how hey a ec ac o s’ access o esou ces (Bo ga i
174
2009); (ii) ag oecology, ha s udies linkages be ween c op di e si y and he empo al s abili y o
175
p oduc ion as well as a ms’ li elihood and ou comes (B a o-Peña e Yode 2024; Rena d e Tilman
176
2021).
177
This p o ocol aims a es ing h ee gene al hypo heses, (de ailed he ea e ): (1) Fa ms di e in hei seed
178
and in o ma ion sou cing ne wo ks because o socioeconomic dispa i ies; (2) These di e ences in
179
ne wo ks ela e o di e ences in a m’s c op di e si y and empo al u no e ; (3) This is associa ed o
180
di e ences in a ms’ capaci y o main ain key unc ions, such as c op p oduc ion and households’ needs
181
co e age, in he ace o pe u ba ions (i.e. esilience).
182
(1) P e ious esea ch in u al a eas indica es ha he p ope ies o he social ne wo ks o some
183
ca ego ies o households limi hei access o esou ces (Rockenbauch e Sakdapol ak 2017). The
184
ca ego ies especially include: (i) less weal hy households (e.g. Bai d e G ay 2014), and households
185
whose heads belong o he ollowing ca ego ies: (ii) e hnic mino i ies o mig an s (e.g. Hoang e al.
186
2018), (iii) women (e.g. To kelsson 2007), (i ) young people (e.g. Cassidy e Ba nes 2012). Thus, he
187
p o ocol aims a es ing whe he a m and a me s cha ac e is ics in e ms o economic s a us,
188
iden i y and da e o se lemen , gende and age a e ela ed o di e ences in seed and in o ma ion
189
ne wo k indica o s, such as he numbe and di e si y o connec ions, and posi ion in he
190
communi y ne wo k (e.g. cen e s pe iphe y).
191
(2) Social ne wo k esea ch has la gely heo ized and es ed how he s uc u e and composi ion o
192
communi ies’ social ne wo ks a ec he ichness and di e si y o esou ces a ailable o hei membe s
193
(e.g. Pu nam 1995), and how an indi idual’s pe sonal social ne wo k and posi ion in he whole ne wo k
194
in luences access o esou ces (e.g. Lin e al. 2001; Bu 2002) (Figu e 1). Ye , bo h collec i e and
195
indi idual le els we e a ely ea ed oge he in Social Ne wo k Analysis (Agneessens e Koskinen
196
2016). Following he mul ile el social ne wo k amewo k p oposed by hese au ho s, we hypo hesize
197
ha he ne wo k pa e ns a bo h communi y (e.g., illage) and a ms’ le els a ec a m c op di e si y
198
and i s empo al dynamics as: (i) he pool o c op di e si y a ailable a he collec i e le el and i s
199
dynamics depend on he communi y seed ne wo k pa e ns, and (ii) he access o a ms o his pool
200
depends on hei posi ion in he communi y ne wo k and on pa e ns o hei own ne wo k. Based on
201
his li e a u e, he p o ocol aims a es ing whe he he c op di e si y le el and he c op u no e in
202
a ms o e ime a e ela ed o seed and in o ma ion ne wo k cha ac e is ics a he communi y
203
and a m le els. A he communi y le el, indica o s encompass he densi y o connec ions in he
204
communi y ne wo k, he di e si y o connec ions, and global s uc u al pa e ns (e.g. dis ibu ed, co e-
205
pe iphe y, o clus e s uc u e.). A he a m le el, indica o s co e he deg ee (numbe o connec ions),
206
connec ions di e si y, and cen ali y in he whole ne wo k.
207
(3) Acco ding o he “insu ance hypo hesis” o which inc easing e idence eme ge om ag oecology
208
li e a u e, main aining a highe le el o c op di e si y o e ime is expec ed o insu e mo e s able c op
209
p oduc ion o e ime, especially in con ex s whe e in e -annual clima e a iabili y is high (Rena d e
210
Tilman 2021). Mo e b oadly, s udies in di e en con ex s epo he con ibu ion o c op di e si y o
211
he main enance o ood a ailabili y and income in he ace o shocks, al hough esul s appea s ongly
212
con ex dependen (B a o-Peña e Yode 2024). We hus expec ha a ms ha ing an enhanced access
213
o c op di e si y would be mo e esilien o shocks in smallholde a ming con ex s whe e clima e
214
a iabili y is high. The p o ocol hus aims a es ing whe he a ms p esen ing seed and in o ma ion
215
ne wo ks pa e ns associa ed o a highe c op di e si y and u no e also p esen a highe and
216
mo e s able c op p oduc ion o e ime, as well as household’s needs co e age.
217
Figu e 1. Main ne wo k pa e ns a he communi y and a m le el expec ed o a ec c op di e si y le el
218
and c op u no e in a ms. The nodes ep esen he social uni s (e.g., he a ms), and he ies ep esen
219
hei in e ac ions (e.g., seed di usion).
220
221
Resea ch s a egy
222
To es hese hypo heses, longi udinal and panel a m su eys we e designed in ARISER p ojec o
223
collec compa able ield da a in h ee d yland a eas in No h Mo occo, cen al Senegal and Sou h-
224
Wes e n Madagasca (Box 2), whe e a me s a e mos ly smallholde s who p ac ice ain ed cul i a ion
225
o ce eals and legumes and ace an inc easing in e -annual ain all a iabili y. Co e ing s udy si es in
226
di e en al hough compa able con ex allows b inging mo e gene al esul s ha go beyond hose o
227
case s udies. This p o ocol is dedica ed o smallholde seed sys ems ha sha e common cha ac e is ics
228
and challenges. Al hough i was applied o smallholde ag icul u al sys ems loca ed in d ylands in he
229
ame o ARISER p ojec , i was hough o be applicable o o he clima ic and ag onomic con ex s as
230
a as smallholde a ming is conce ned.
231
These s udy a eas we e selec ed because hey a e ep esen a i e o biodi e si y-based small-scale
232
a ming ha is p ac iced by he la ge majo i y o a me s h oughou A ican d ylands, ha ing limi ed
233
access o ag icul u al echnologies and in as uc u es (Lowde e al. 2016; Riccia di e al. 2018). They
234
we e also selec ed because hey a e loca ed in a eas p one o ecu en ain all de ici , which allows
235
in es iga ing he linkages be ween seed and in o ma ion ne wo k pa e ns, c op di e si y and i s
236
ou comes (i.e. he s abili y p oduc ion, household needs co e age). Fa me s in he h ee s udy si es
237
combine he p oduc ion o ood c ops o sel -consump ion and local sale, and he p oduc ion o cash
238
c ops o sale on u ban ma ke s o o expo . The seed sys ems om which hey sou ce hei seeds
239
combine a di e si y o ac o s: a me s, go e nmen al and non-go e nmen al u al de elopmen
240
o ganiza ions, and p i a e ac o s. Al hough such a coexis ence o ac o s was ound in he h ee s udy
241
si es, he sha e o seeds ob ained by a me s om each ac o ype di e be ween si es. The h ee
242
coun ies whe e he s udy akes place also p esen simila challenges ega ding smallholde seed sys em
243
go e nance. Public policies and in e na ional de elopmen agencies suppo he cen alized seed sec o
244
ha in ol es s a e and p i a e ac o s o he dissemina ion o ce i ied seeds om a ew species and
245
a ie ies. Gene ally speaking, hese ac ions poo ly ma ch he speci ic needs o smallholde s, mainly
246
because species and a ie ies dis ibu ed a e ew and a ely adap ed o he di e si y o local g owing
247
condi ions and uses, and because seeds a e cos ly (Almekinde s and Louwaa s 2002).
248
In each coun y, s udy si es include en communi ies loca ed in he same commune (i.e. smalles
249
adminis a i e uni ). This allows es ing o he e ec o di e en communi y ne wo k pa e ns wi hin
250
a same social-ecological con ex . Longi udinal da a collec ion is being conduc ed o e ou consecu i e
251
yea s (2023-2026) o cap u e in e -annual a ia ions and es he hypo hesis conce ning he empo al
252
dynamics o c op di e si y and he s abili y o c op p oduc ion and household’s needs ul illmen .
253
Su eys a e conduc ed a e he las ha es o he ag icul u al yea , du ing pe iods whe e a me s a e
254
a ailable.
255
Wi hin ARISER p ojec , he da a collec ed h ough he a m su ey p o ocol desc ibed in his a icle,
256
mainly dedica ed o quan i a i e analysis, a e combined wi h: (i) su eys among seed selle s on local
257
ma ke s, (ii) quali a i e da a collec ed h ough a a ie y o app oaches such as e hnog aphic su eys,
258
pa icipan obse a ions, se ious games and pa icipa o y wo kshops, (iii) ex ensi e li e a u e e iew,
259
including “g ey li e a u e” and policy documen s. Da a collec ed h ough hese a ious app oaches
260
Du ing he aining, collec i ely de ining he mos co ec wo ding and ansla ion possible o each
423
ques ion wi h in e iewe s was ins umen al. This was equi ed o a oid as much as possible
424
misunde s anding and di e ences in su eys implemen a ion be ween in e iewe s, as each indi idual
425
conduc s he su eys based on his.he own pas expe iences and wi h his. he own pos u e and app oach.
426
The aining p o ided o local in e iewe s was he e o e buil up in a pa icipa i e and collabo a i e
427
way, in o de o in ol e he in e iewe s as much as possible in he s udy and enable hem o sha e hei
428
knowledge and skills. Despi e his common aining o all in e iewe s in ARISER s udy, we
429
acknowledge ha in e iewe s may ha e misunde s ood he objec i e o some ques ions and may also
430
pay less e o han ained esea che s in ge ing he accu a e answe om a me s. Assessmen o
431
in e iewe ’s bias is hus being implemen ed p io o he da a analysis.
432
Once his i s s age o su ey p epa a ion had been comple ed, a m su eys s a ed in 2023 and a e
433
epea ed each yea wi h he same households o ou yea s, so he a m household da a collec ion
434
p o ocol ends in 2027. A pa icula challenge in his longi udinal su ey was o limi pa icipan s
435
sa u a ion and disengagemen as we asked epea edly simila ques ions o ou consecu i e yea s. We
436
payed pa icula a en ion o ain he in e iewe s eam o explain he eason o ha , and his was
437
de ailed in he le e o in o ma ion o he P io In o med Consen P ocedu e. O ganizing wo kshops
438
o p esen and deba e esul s wi h su ey pa icipan s was also key o main ain hei in ol emen , and
439
aising discussions abou hei own in e es in ela ion wi h he opic o he p ojec .
440
Fa ms and pa icipan s ec ui men s a egy
441
Fi s , an exhaus i e sampling o a ms in ol ed in seed and in o ma ion ci cula ion in each communi y
442
was seek, in he objec i e o conduc comple e ne wo k analysis a he communi y le el (Wasse man
443
and Faus 1994). Comple e ne wo k allows o analyze he pa e ns o seed ci cula ion be ween
444
communi y membe s and he posi ion o a ms in his ne wo k. Howe e , we encoun e some issues o
445
mee his objec i e o di e en easons. Fi s ly, i was no possible o make su e ha all he a ms we e
446
su eyed and o es ima e he pa icipa ion a e because ge ing access o p eexis ing census o a ms
447
was no possible, and local au ho i ies we e eluc an ha esea che s conduc hei own census. This
448
p oblem was aced in all he ield si es o ARISER p ojec . Secondly, i is equen ha in e iews could
449
no be done in some a ms because he heads a e no p esen due o seasonal mig a ion, o because hey
450
e use o pa icipa e in he su ey.
451
Second, we decided in each si e who should be in e iewed in each a m based on li e a u e e iew,
452
obse a ions done du ing p elimina y su eys, and discussions wi h he local collabo a o s. C op
453
managemen p ac ices a he a m le el a e he esul o complex decision-making p ocesses and ac ions
454
in ol ing di e en household membe s. S udies gene ally in e iew a m heads, conside ed as ha ing
455
a good knowledge o he di e en ace s o c op managemen a he a m le el, bu some s udies indica e
456
ha his s a egy does p o ide an accu a e pic u e o he whole seed po olio and sou cing channels a
457

he a m le el (Wencélius e al. 2016). We ini ially aimed a in e iewing bo h he male and emale
458
a m heads, as bo h can g ow di e en c ops. In he case o polygamous households, i was ag eed ha
459
he i s spouse gene ally ge s a good knowledge o wha he o he s we e doing, al hough we ag ee his
460
app oach p esen s some limi s. Howe e , we we e cons ained o e iew his objec i e o di e en
461
easons. Fi s , in mos cases i was di icul o ind bo h he man and he women a ailable o he
462
in e iew wi hin households. This was especially he case in Madagasca , whe e we decided o
463
in e iew one o he o he depending on hei a ailabili y, as bo h ge a good gene al knowledge o wha
464
is cul i a ed in he a m. In Senegal, men equen ly mig a e o ind a job du ing he d y season whe e
465
he in e iews we e conduc ed, which aises p oblems o in e iew hem. In many cases, women we e
466
also e y busy a he ime o he in e iew o ood p epa a ion and house wo k, and i was no easy o
467
con ince hem o pa icipa e. We encoun e di icul ies conduc ing in e iews wi h women because o
468
cul u al habi s, pa icula ly in Senegal and Mo occo. In some cases, men do no le he in e iewe s
469
conduc su eys wi h hei wi es because hey a gue ha hey knew all ha was done in he a m. Men
470
could e en eel o ended ha we asked o in e iew women on his opic. This po en ially impac ed ou
471
abili y o documen he seed managemen o some c ops (e.g. hose cul i a ed in home ga dens), whe e
472
women appea o play an impo an ole.
473
Fa ms’ su eys con en
474
The a m su eys (see he su ey ques ionnai e in S1) a e conduc ed on a yea ly base and ake place
475
a e he main ha es pe iod. In ARISER p ojec , hey co e he a ge c ops (ce eals, legumes, and
476
ube s) cul i a ed du ing he yea p eceding he su ey (s a ing om he i s sowing pe iod). Fo
477
ins ance, he su eys conduc ed in 2023 co e ed: he pe iod be ween No embe 2022 and Augus 2023
478
in Madagasca , be ween May 2022 and Decembe 2022 in Senegal, and be ween Oc obe 2021 and
479
Sep embe 2022 in Mo occo.
480
The a m su eys a e based on ques ionnai es s uc u ed in wo main sec ions (sec ions D and C o
481
Figu e 2) ha aim a collec ing gene al da a conce ning he cha ac e is ics o he households, a ms and
482
ields ha a e ele an o unde s and c op and seed managemen , and le el o c op p oduc ion. Indeed,
483
his p o ocol app ehends c op di e si y (species and a ie ies) managemen as embedded in he gene al
484
household unc ioning and adop s a sys emic pe spec i e o i s analysis. This implies accoun ing o
485
he e ec s o he di e en managemen scales (i.e. household, a m, ield) wi hin a ms and he di e en
486
aspec s o a ms’ unc ioning and o a me s’ decision-making conce ning c ops and seeds. In his
487
pe spec i e, he ques ionnai es a e designed o collec da a co e ing he di e en managemen scales.
488
In ARISER p ojec , we decided o collec de ailed da a on households’ socioeconomic cha ac e is ics
489
and ield managemen only he i s yea o he su ey o cha ac e ize he a ming sys em, bu his was
490
no epea ed he ollowing yea s o limi he ime spen du ing he su ey.
491
492
Figu e 2: O ganiza ion o he ques ionnai e o a m su eys wi h he di e en o ganiza ional le els a which
493
da a we e collec ed and he ypes o da a collec ed ( o mo e de ails, see he su ey ques ionnai e in S1).
494
495
(i) Fa m’s cha ac e is ics: A he household scale (sec ion D, Figu e 2 and S1), he
496
ques ionnai e co e s he socioeconomic cha ac e is ics, including household composi ion,
497
economic and li elihood ac i i ies, and indica o s o weal h. Cha ac e is ics o he
498
indi iduals chie ly in ol ed in c op managemen (household heads) a e also documen ed,
499
including educa ion le el, age, gende , geog aphical o igin and mo he ongue. I also
500
includes some ques ions on households’ capaci y o mee hei needs o ood and income,
501
and he p oblems hey aced o c op p oduc ion, as well as hei app ecia ion conce ning
502
access o seed and in o ma ion. A he ield scale (sec ion C.1), he ques ionnai e co e s
503
cha ac e is ics ha a e impo an o c op di e si y managemen , such as modali ies o
504
access o land, ag icul u al ools and a m asse s, da a on he ag oecological condi ions (soil
505
ype and opog aphy) and on a ming p ac ices (e.g. use o inpu s, ype o soil p epa a ion,
506
ype o c ops’ associa ion).
507
This sec ion o he ques ionnai e is designed o o e come se e al issues ha we e
508
encoun e ed du ing he es s age, and we ended in de ining indica o s and o mula ing
509
ques ions ha limi hese issues. Collec ing accu a e da a on sensi i e aspec s such as he
510
economic s a us and asse s o he households, he land a ea, and he ag icul u al p oduc ion,
511
is especially challenging. In some con ex s, esponden s a e eluc an o p o ide his ype
512
o in o ma ion ei he because o local no ms and aboos, o because hey ea ha such
513
in o ma ion could be used by he au ho i ies o con ol hem, and o egula e hei access o
514
s a e suppo sys ems and subsidies. In some cases, esponden s don’ know he su ace
515
hey cul i a e, which is a classical p oblem aced in his ype o su ey, and es ima ing
516
hese su aces based on he ime spen p epa ing he soil o sowing can be e y challenging.
517
518
(ii) Seed baske s: A e hese gene al sec ions, he ques ionnai e goes in o sec ions dedica ed
519
o he seed baske and associa ed in o ma ion, ha co e s all he seed lo s cul i a ed by he
520
household o e he las yea o ce eal, legume and ube c ops. A seed lo is he se o
521
seeds o a gi en a me ’s named a ie y he.she epo ed ob aining om he same sou ce on
522
he same da e. Fo documen ing seed baske s while limi ing memo y issues, esponden s
523
we e i s asked o lis he species o legumes, ce eals o ube c ops hey ha e g own o e
524
he pas yea (sec ion C.2). A p ede ined lis o species o hese c ops was used o
525
sys ema ically check o he p esence o absence o each in he a m. Then (sec ion C.2.a),
526
o each c op species, a me s we e asked o lis all he a ie ies ha ha e been g own in
527
he a m o e he las yea based on he name hey use o iden i y hem (he ea e e e ed
528
o as “named a ie ies”). The a ie ies named by a me s can co e di e en ypes o
529
biological en i ies, hey can be land aces, i.e. a se o indi idual plan s p esen ing simila
530
mo phological cha ac e is ics acco ding o a me s and hence iden i ied using he same
531
name by hem bu ha can co e a subs an ial mo phological and gene ic di e si y (Loue e
532
2000, Villa e al. 2006), bu also some imes DHS egis e ed a ie ies (Dis inc ,
533
Homogeneous and S able), o c op popula ions ha esul om a mix be ween land aces
534
and DHS a ie ies ( e e ed as ‘c eolized a ie ies’ in he li e a u e). A p ede ined lis o
535
a ie ies names is sugges ed o a me s based on p elimina y su eys, bu he op ion o add
536
new names is possible. Cha ac e is ics o he a ie ies a e eco ded, wi h a ocus on hei
537
o igin acco ding o he a me , i.e. whe he hey a e: (i) local a me s’ a ie ies, conside ed
538
by he in e iewees o ha e been g own in he a ea be o e hey s a ed a ming, (ii)
539
in oduced a ie ies ha appea in he a ea a e hey s a ed a ming, in which case we
540
documen ed when possible i a me s hink hey we e dis ibu ed h ough o icial channels
541
such as p i a e seed companies o de elopmen o ganiza ions, o i hey a e a ie ies
542
in oduced om o he egions, and (iii) a ie ies whose o igin is unknown. We also
543
es ima ed he ancien ly o he a ie y in he s udy a ea by asking esponden s whe he hei
544
pa en s we e al eady g owing his a ie y. We documen he quan i y sowed and ha es ed
545
o each a ie y, he use o des ina ion o he ha es p oduc s (e.g. household ood sel -
546
consump ion, animal ood, g ain o seed sale), he a me 's le el o sa is ac ion wi h he
547
ha es and any p oblems encoun e ed ha a ec ed p oduc ion le els. We also asked
548
a me s abou he easons o cul i a ing his pa icula a ie y.
549
Fo each named a ie y (sec ion C.2), we ask a me s o lis and desc ibe he di e en seed
550
lo s (i.e. all he seed o he same a ie y ob ained om he same sou ce on he same da e
551
by he household) hey sow o e he pas yea , and o desc ibe each seed lo sou ce and he
552
social ie in ol ed (sec ions C.2.a and C.2.c). The seed lo s we e di e en ia ed depending
553
on whe he esponden s epo ed hey had been sel -p oduced on- a m o ob ained om
554
ou side, in case i was sel -p oduced, we asked when he seed lo was sou ced ou side he
555
a m o he las ime (Figu e 3).
556
Then, we eco d he ex e nal sou ces whe e he seeds we e las ob ained based on a
557
ca ego iza ion om p elimina y su eys ha de ined i e main ca ego ies o sou ce.
558
Ce ain ca ego ies ha e been added and adap ed depending on he s udy si e, based on he
559
su eys conduc ed among seed dis ibu o s. To co e he a ious less equen ly ci ed
560
sou ces, we lea e he possibili y o a me o ci e o he seed sou ces (ca ego y “o he
561
sou ce”). Addi ional in o ma ion is hen collec ed conce ning he seed sou ce, including i s
562
geog aphical loca ion, i s gende and i s name in case i is an indi idual loca ed in he
563
esea ch a ea, and i s loca ion and name in case i is an o ganiza ion o a p ojec . In case
564
he seed ansac ion in ol es wo indi iduals, he social ie be ween he seed ecei e and
565
p o ide is desc ibed as ollows: (i) no p e-exis ing ie, (ii) acquain ance (i.e. hey know
566
each o he bu don’ ha e a close social ela ionship), (iii) iendship, (i ) neighbo ing (o
567
house o ields), ( ) amilial ie (kinship). In case o a kinship ie, we use a s anda dized
568
way o coding p ecisely he ela ionship o be able o iden i y i s p ecise na u e. Las ,
569
in o ma ion is collec ed conce ning he easons o seeking seeds om his speci ic seed
570
sou ce, wi h p ede ined op ions sugges ed (see ques ionnai e in S1) and he possibili y o
571
add new ones. We also collec in o ma ion on he na u e o he ansac ion, i.e. i i is
572
mone ized, in he o m o gi s o exchanges, o inhe i ed.
573
The sec ion C.2.b aims a documen ing he sou ces h ough which a me s ge in o ma ion
574
conce ning each c op a ie ies and ha mo i a ed hem o s a hei cul i a ion. In his
575
pe spec i e, a me s a e in i ed o desc ibe how hey go he idea o g ow his pa icula
576
named a ie y. Fi s , we eco d i his is based on a me s’ own expe ience, o on ex e nal
577
ad ice o obse a ions. I i is ex e nal, we documen he cha ac e is ics o he in o ma ion
578
sou ce and social ie in ol ed, ha a e desc ibed in he same way as o he seed
579
ansac ions.
580
Fo bo h seed and in o ma ion sou cing, he name gene a o app oach is applied by asking
581
he in e iewees o ci e eely he name o seed/in o ma ion p o ide s and we collec o he
582
in o ma ion o make possible hei iden i ica ion a-pos e io i (age, gende , p ecise
583
loca ion). The a-pos e io i iden i ica ion o hese indi iduals is made in collabo a ion wi h
584
he esponsible o he su eys in each coun y, when necessa y. Applying a os e app oach
585
is usually ecommended o collec ing such da a in a sys ema ic way (Wasse man and Faus
586
1994), based on a p ede ined lis o households. This was no possible in he di e en s udy
587
si es co e ed by ARISER p ojec s as no census exis ed o local au ho i ies we e no allowed
588
o ansmi hem, and hey we e also eluc an o allow he esea ch eam o conduc a
589
census. The name-gene a o app oach can lead o e o s in iden i ying a-pos e io i he
590
indi iduals om whom he a me s ob ained seeds and in o ma ion. This is especially
591
challenging when homonymy is equen , which equi es collec ing p ecise in o ma ion on
592
he loca ion, age and spouse o he p o ide . Fu he mo e, collec ing da a on he p ecise
593
iden i y o seed o in o ma ion p o ide s is e y sensi i e, and in e iewees can be eluc an
594
o p o ide his in o ma ion. Accu a ely collec ing his ype o da a equi es us , which is
595
di icul o a ain in such la ge-scale su eys whe e he in e iewe s s ay a limi ed ime in
596
he communi ies. Fa me s’ memo y is also one limi , as some e en s o seed o in o ma ion
597
can be easily o go en.
598
599
Figu e 3: De ail o he di e en da a collec ed h ough he ques ionnai e sec ions on seed and in o ma ion
600
sou cing.
601
602
Ques ionnai e design and digi al implemen a ion
603
In ARISER s udy, KoboToolbox so wa e and he KoboCollec applica ion we e used o acili a e la ge-
604
scale su eys and educe he isk o in e iewe s’ e o s, by elimina ing pape -based da a collec ion and
605
indi idual da a en y in o Excel sp eadshee s. The so wa e o e s he possibili y o build a su ey o m
606
wi h di e en esponse modali ies (e.g. in ege o decimal, single choice, mul iple choice, ee ex ),
607
and whose s uc u e can include ques ion nes ing, epe i ion and loops. The su ey o m can be buil
608
o line using he xlsx language. This allows he a iable names and modali ies o be de ined a he same
609

ime as he ques ions a e o mula ed and he possible answe s a e lis ed. Thus, each ques ion
610
co esponds o a a iable, and each possible answe co esponds o a modali y. So, when a su ey is
611
ca ied ou , i.e. a o m is illed in, i is sen o he so wa e se e and he da a collec ed is di ec ly
612
eco ded and s o ed in an Excel sp eadshee acco ding o he p e-de ined a iable and modali y names.
613
Fo ield su eys, he KoboCollec applica ion enables he o m o be downloaded on o a able .
614
Da a managemen and analysis
615
Da a managemen
616
ARISER Da a Managemen Plan desc ibes he da a managemen li e cycle o he da a o be collec ed,
617
p ocessed and s o ed p ojec . As he p ojec p ocesses pe sonal da a, measu es ha e been implemen ed
618
o ensu e adequa e p o ec ion : (i) aise awa eness among use s (bo h in e nal and ex e nal o he p ojec )
619
wo king wi h pe sonal da a o he isks associa ed wi h p i acy issues, (ii) es ablish p ocedu es o
620
p ocessing, sha ing and s o age pe sonal da a in line wi h he Gene al Da a P o ec ion Regula ion, (iii)
621
use secu e In o ma ion Technologies solu ions wi h use access limi ed o he da a hey need and use
622
enc yp ed elec onic de ices (ha dwa e, compu e , phones), (i ) Include special clauses applying o
623
pe sonal da a in p o ide s’ con ac s. Pe sonal da a allowing indi iduals iden i ica ion a e only accessed
624
by he eam membe s, who a e con ac ually commi ed o comply wi h he p i acy ules.
625
Once submi ed, he ques ionnai es a e s o ed on CIRAD Kobo oolbox pla o m ha is loca ed on a
626
se e loca ed in he Eu opean Union. Da a a e hen downloaded and s o ed secu ely on enc yp ed
627
elec onic de ices. A he end o he s udy, da a will be a chi ed in a secu e manne o a limi ed du a ion
628
o 30 yea s. To comply wi h he FAIR ules, non-sensi i e o sec e da a will be made publicly a ailable
629
on he da a e se o CIRAD, and only ully anonymized o agg ega ed will be made a ailable so ha
630
indi iduals’ e-iden i ica ion will no be possible.
631
Da a Analysis o e iew
632
Quan i a i e da a analysis is being conduc ed o es he main hypo heses o he p ojec , building on he
633
di e en heo e ical and me hodological amewo ks desc ibed in he concep ual amewo k. We
634
p esen he e a gene al desc ip ion o he me hods en isioned, bu he de ailed desc ip ion o me hods
635
e ec i ely implemen ed o ea he di e en esea ch ques ions will be p o ided in he di e en a icles
636
ela ed o each ques ion. An o e iew o he esea ch ques ions, hypo hesis and analysis conduc ed is
637
p o ided in S2 (S udy Design able).
638
Fi s , a se o analysis is being conduc ed o es ablish a ypology o a me s’ seed and in o ma ion
639
sou cing ne wo ks, and assess he socio-economic ac o s associa ed o hese di e en g oups o
640
p ac ices. In his objec i e, we combine clus e ing and indica o -based app oaches o gain a be e
641
assessmen , ollowing he classical pe sonal ne wo k analysis me hods (Wasse man and Faus 1994,
642
McCa y e al. 2019). Mul i a ia e analysis (PCA) and classi ica ion algo i hms such as HAC
643
(hie a chical ascendan classi ica ion) a e being espec i ely applied o he ma ices o a ms’ seed o
644
in o ma ion sou ces. Ne wo k me ics a e being compu ed a he a m le el o measu e he ollowing
645
indica o s o pe sonal (ego) ne wo ks: (i) he numbe o di ec seed o in o ma ion sou cing ies a a m
646
has (indeg ee, F eeman 1979), (ii) he di e si y o ies de i ed om Shannon en opy ha has he
647
ad an age o being obus o missing da a and almos insensi i e o ne wo k size (Blü hgen 2006), (iii)
648
he geog aphical ex en o ne wo ks, based on he maximum geog aphical dis ance o ies. Gene alized
649
Linea Models a e being used o es o he co ela ions ei he be ween a m’s membe ship o he
650
di e en seed o in o ma ion ne wo k g oups o be ween a ms’ ne wo k measu es on one hand, and on
651
he o he hand, a ms’ socioeconomic a iables such as he gende , age, and e hnici y o he house heads
652
o he economic s a us o he household.
653
Second, a speci ic se o analysis is being conduc ed o assess gende di e ences in hei epo s o seed
654
baske s and sou cing. This analysis is only be conduc ed on da a collec ed in Senegal and Madagasca
655
as in e iews could no be conduc ed wi h women in Mo occo. Fi s , analysis is being conduc ed o
656
compa e he epo s o men a m heads VS women conce ning he whole a m seed baske and sou cing.
657
This compa ison be ween he popula ion o ma ied men and women in e iewed conce ning he a m
658
seed baske s will allow o assess whe he some gende bias exis s, by es ing i men and women di e
659
in he equency o epo ed c op cul i a ed, seed lo s sel -p oduc ion, and ype and cha ac e is ics o
660
seed sou ces. Second, we es whe he women’s seed baske s and sou ces a e accu a ely epo ed by
661
men when hey a e in e iewed abou his aspec a he a m le el. In his objec i e, we compa e he
662
answe s o men and women in ol ed in seed sou cing wi hin he same household o assess he ma ch
663
o hei epo s conce ning he c op cul i a ed by women. This is be conduc ed only o Senegal da a,
664
whe e a speci ic su ey is done wi h women in a subsample o households.
665
Thi d, he ela ionship be ween a ms’ seed and in o ma ion sou cing ne wo ks on he one hand, and
666
he c op di e si y le el in a ms and he c op u no e o e ime on he o he hand, is being es ed.
667
Measu es o he le el o c op di e si y on each a m and in each yea a e compu ed based on he numbe
668
o species and a ie ies. Then, he u no e o c op species and a ie ies o e ime is compu ed using
669
B ay-Cu is dissimila i y o compa e c op composi ion a he a m le el be ween pai s o consecu i e
670
yea s (Magu an and Hende son 2010). These measu es will be compu ed sepa a ely a he species le el
671
and a he a ie al le el o each species. Each measu e is agg ega ed o e ime in o a summa y mul i-
672
annual index, compu ed as he a io o hei mean o e he i e yea s o he s udy o hei s anda d
673
de ia ion (Rena d and Tilman 2019). We build on ecen ly de eloped mul ile el linea modeling
674
app oaches o es he espec i e e ec o communi y e sus a m ne wo k s uc u e on c op di e si y
675
and u no e measu es (Agneessens and Koskinen 2016). Fa m ne wo k measu es is compu ed as
676
desc ibed abo e, and ou communi y ne wo k measu es a e compu ed: (i) global ne wo k di e si y
677
based on bo h he abundance o each ac o ca ego y and he p obabili ies o in e ac ion be ween hese
678
ca ego ies (Ohlman e al. 2019). Ac o s ca ego ies a e de ined depending on hei na u e - indi idual,
679
go e nmen al, non-go e nmen al o p i a e o ganiza ions - hei geog aphical loca ion and hei
680
socioeconomic a ibu e, (ii) he densi y o he ies in he ne wo k (Wasse man and Faus 1994).
681
Depending on ou capaci y o es ima e global measu es o ne wo k s uc u e, syn he ic measu es
682
e lec ing how close a ne wo k is o a co e-pe iphe y s uc u e may be compu ed based on he ou pu s
683
o S ochas ic Block Models (Gonzalez e al. 2020). To deal wi h he ne wo k’s in e dependence issue,
684
ixed and andom communi y-le el e ec s a e included, as well as an au oco ela ion componen
685
(Snijde s and Boske s 2012). Cha ac e is ics o a ms (e.g. li elihood ac i i ies), communi ies (e.g.
686
size), and si es (e.g. wea he da a), a e included as co a ia es in he model o con ol o hei e ec .
687
Las , es s a e being conduc ed o assess o wha ex en he cha ac e is ics o a ms’ seed and
688
in o ma ion sou cing ne wo ks ela e o p oxies o esilience ou comes a he a m le el, namely mo e
689
s able ag icul u al p oduc ion o e ime, and an enhanced capaci y o co e hei ood and inancial
690
needs. Based on he analysis p esen ed abo e, we expec o iden i y g oup o a ms p esen ing pa icula
691
ne wo k cha ac e is ics ela ed o a highe le el o c op di e si y and c op u no e . We es i he a ms
692
p esen ing hese connec i i y cha ac e is ics ha e a mo e s able ag icul u al p oduc ion o e ime and
693
be e co e hei needs o e se e al yea s as compa ed o o he a ms. An index o c op p oduc ion is
694
being compu ed a he a m le el o each yea as he agg ega ion o he quan i ies ha es ed o each
695
c op species con e ed in ene gy (calo ies). I is hen be agg ega ed o e ime in o a summa y mul i-
696
annual index, compu ed as o c op di e si y. We es i he mean p oduc ion di e s be ween a ms
697
g oups using a iance analysis, and i needs co e age di e s using adap ed ank es s.
698
Challenges and pe spec i es
699
Th ough he p o ocol desc ibed, we expec o imp o e he gene al unde s anding o he ole o seed and
700
in o ma ion sou cing p ac ices in smallholde a me s’ access o c op di e si y, and o wha ex en his
701
con ibu es o hei esilience o pe u ba ions. Collec ing compa able da a wi hin a common amewo k
702
and ield se ing o e di e en s udy si es is a key challenge o p og ess in he gene al unde s anding o
703
he p ocesses s udied. As de ailed in he in oduc ion, se e al case-s udies in es iga ed he ela ionship
704
be ween c op di e si y and seed sou cing p ac ices, leading o di e ging conclusions. The di e ences
705
in me hods and con ex s o hese s udies s ongly limi ed ou capaci y o iden i y he o igin o hese
706
di e ences and hei implica ions. Conduc ing la ge scale longi udinal su eys is hence needed o
707
complemen case s udies and b ing gene al conclusions, bu i aises pa icula challenges which we
708
belie e i is impo an o sha e wi h ou esea ch communi y. La ge panel and longi udinal su eys on
709
smallholde seed and in o ma ion ne wo ks, co e ing housands o esponden s and epea ed o se e al
710
yea s, a e no common in smallholde seed sys ems esea ch. ARISER p o ocol is he i s o his kind,
711
co e ing ou consecu i e yea s and 2000 households in h ee di e en al hough compa able s udy si es.
712
The quan i a i e app oach based on his p o ocol p esen s some s eng hs, bu also some weaknesses,
713
as discussed abo e. We hus belie e ha sha ing his p o ocol and he challenges we me along i s
714
design, es and implemen a ion can be o in e es o he design o u u e esea ch in his domain and
715
wo h o be sha ed wi h ou esea ch communi y. These aspec s should also be conside ed when
716
examining he esul s and conclusions ha can be d awn om he applica ion o he p o ocol.
717
Conduc ing la ge su eys on seed sys ems equi es o comp omise on he le el o de ails and accu acy
718
o da a. Gene ally speaking, la ge su eys makes i much mo e di icul o o e come he local no ms
719
and aboos as compa ed o case s udies ha allow o build p oximi y and us wi h su ey pa icipan s
720
and makes possible o ex end in e iew du a ion and de ails and access o in o ma ion on sensi i e
721
aspec s. This especially a ec ed ou abili y o ec ui women, as explained in he me hods sec ion. I
722
also limi ed ou capaci y o ge de ails abou he iden i y o seed o in o ma ion p o ide , a ec ing in
723
u n he abili y o conduc comple e ne wo ks da a sampling as ini ially planned.
724
Combining his quan i a i e su ey da a wi h ine-g ain quali a i e and comp ehensi e su eys is hus
725
c ucial o gain a good unde s anding o he p ocesses a s ake. The p esen ed p o ocol pu he ocus on
726
quan i a i e me hods, bu we insis on he impo ance o combining i wi h quali a i e da a, as well as
727
ex ensi e li e a u e e iew. This is especially needed o ge a comple e unde s anding o he go e nance
728
o seed sys ems and ma ke dynamics ha play a key ole in smallholde a me s’ access o seed. I
729
allows also o ge in o he local ules and no ms ha d i e seed and in o ma ion ci cula ion in u al
730
socie ies, and documen mo e di use o hidden p ocesses ha could no be documen ed h ough
731
ques ionnai es. In ARISER p ojec , case s udies a e conduc ed by s uden s in each s udy si es, which
732
p o ides an in-dep h unde s anding o he local c op di e si y managemen p ac ices. A se ious game
733
was designed o ob ain complemen a y da a abou a me s’ decision-making ega ding seed and
734
in o ma ion sou cing. We a e conduc ing a li e a u e e iew co e ing scien i ic publica ions and epo s
735
as well as o icial documen s conce ning public policies and legal amewo ks ela ed o he seed sec o .
736
Las , wo kshops a e o ganized wi h a me s and local de elopmen o ganiza ions o discussing ou
737
esul s, which is decisi e o building obus and accu a e conclusions.
738
The p esen ed p o ocol is expec ed o make signi ican con ibu ions o esea ch on smallholde seed
739
sys ems by o e ing a uni ied amewo k o conduc ing u u e compa able esea ch. Such a amewo k
740
will allow o implemen a la ge-scale assessmen o he e ec o di e en c op seed and in o ma ion
741
sou cing modali ies on smallholde s’ access o c op di e si y ac oss di e en con ex s. This will add ess
742
an impo an challenge in his esea ch ield, as schola s claim ha local social ne wo ks in ol ed in
743
c op seed and in o ma ion di usion play a key ole in smallholde s’ access o c op di e si y (Coomes
744
e al. 2015, Pau asso e al. 2013), bu conc e e e idence o which cha ac e is ics o hese ne wo ks is
745
lacking. This p o ocol is also expec ed o ad ance ou unde s anding o he p ocesses in ol ed in
746
smallholde esilience in a con ex o high clima e a iabili y. This is pa icula ly impo an as esea ch
747
Scoones I, Thompson J. The poli ics o seed in A ica’s g een e olu ion: Al e na i e na a i es and
936
compe ing pa hways. Ids Bull. 2011;42(4):1-23.
937
Snijde s, T. A. & Boske , R. J. Mul ile el heo ies, mul is age sampling, and mul ile el models (Chap e
938
2). in Mul ile el analysis: An in oduc ion o basic and ad anced mul ile el modeling (Sage, 2012).
939
Thomas, M., & Caillon, S. E ec s o a me social s a us and plan biocul u al alue on seed ci cula ion
940
ne wo ks in Vanua u. Ecol and Soc. 2016; 21(2).
941
Schlingmann A, G aham S, Benyei P, Co be a E, Ma inez Sanes eban I, Ma elle A, e al. Global
942
pa e ns o adap a ion o clima e change by Indigenous Peoples and local communi ies. A sys ema ic
943
e iew. Cu Opin En i on Sus ain. 2021;51:55-64.
944
Tengö M, B ondizio ES, Elmq is T, Malme P, Spie enbu g M. Connec ing Di e se Knowledge
945
Sys ems o Enhanced Ecosys em Go e nance: The Mul iple E idence Base App oach. AMBIO.
946
2014;43(5):579-91.
947
Tengö M, Hill R, Malme P, Raymond CM, Spie enbu g M, Danielsen F, e al. Wea ing knowledge
948
sys ems in IPBES, CBD and beyond—lessons lea ned o sus ainabili y. Cu Opin En i on Sus ain.
949
2017;26-27:17-25.
950
To kelsson, Å. Resou ces, no capi al: a case s udy o he gende ed dis ibu ion and p oduc i i y o
951
social ne wo k ies in u al E hiopia. Ru al Sociol. 2007;72, 583–607.
952
UNEP-WCMC. A spa ial analysis app oach o he global delinea ion o d yland a eas o ele ance o
953
he CBD P og amme o Wo k on D y and Subhumid Lands. Da ase based on spa ial analysis be ween
954
WWF e es ial eco egions (WWF-US, 2004) and a idi y zones (CRU/UEA; UNEPGRID, 1991).
955
2007. A ailable om: h ps://da a-gis.unep-
956
wcmc.o g/po al/home/i em.h ml?id=789 cac8959943ab9ed7a225e5316 08
957
Vanloque en G, Ba e PV. How ag icul u al esea ch sys ems shape a echnological egime ha
958
de elops gene ic enginee ing bu locks ou ag oecological inno a ions In: Food So e eign y,
959
Ag oecology and Biocul u al Di e si y. Rou ledge; 2017.
960
Ve nooy R. Does c op di e si ica ion lead o clima e- ela ed esilience? Imp o ing he heo y h ough
961
insigh s on p ac ice. Ag oecol Sus ain Food Sys . 2022;46(6):877-901.
962
Villa, T. C. C., Max ed, N., Schol en, M., & Fo d-Lloyd, B. De ining and iden i ying c op land aces.
963
Plan gene es. 2005; 3(3), 373-384.
964
Wasse man, S, and Faus , K. 1994. Social Ne wo k Analysis: Me hods and Applica ions. Camb idge
965
Uni e si y P ess.
966
Wencélius J, Thomas M, Ba billon P, Ga ine E. In e household a iabili y and i s e ec s on seed
967
ci cula ion ne wo ks: a case s udy om no he n Came oon. Ecol Soc. 2016;21(1).
968

Wes engen OT, Dalle SP, Mulesa TH. Na iga ing owa d esilien and inclusi e seed sys ems. P oc
969
Na l Acad Sci. 2023;120(14):e2218777120.
970
971
Sec ion numbe o ype o answe
Ques ions asked
each yea
Ti le o ques ion asked Choice o answe s
A. A. INTRODUCTION
A.1 A.1 PRESENTATION OF THE ARISER STUDY
Yes
Summa y, P ocess, and Pu poses o he Su eys :
The ARISER s udy aims o be e unde s and how di e en c ops enable a me s in [commune name] o cope wi h dis up ions, pa icula ly
clima ic ones. This s udy also seeks o unde s and whe e a me s ob ain hei seeds o hese a ious c ops and how hese di e en seed supply
s a egies in luence households' abili y o cope wi h dis up ions.
The esul s om his su ey will lead o scien i ic publica ions, p esen a ions a con e ences, o in o ma ional documen s o de elopmen
s akeholde s. We will y o dissemina e he esul s and discuss hem wi h all he illages in he a ea so ha all esiden s o he commune can
bene i . All w i en epo s om his s udy will be deli e ed o he illage chie and will be accessible o anyone who wishes o see hem.
I you wish o ob ain in o ma ion abou he ollow-up o he p ojec , you can con ac he lead esea che a any ime.
Yes
Pa icipan s' Righ s :
You a e olun a ily pa icipa ing in his su ey, which means ha you a e ee o pa icipa e o no , and you can choose no o answe one o
mo e ques ions asked o you.
You pa icipa ion is ee o cha ge, wi h no inancial compensa ion o cos o you.
You can s op he su ey and wi hd aw om his s udy a any ime, wi hou needing o jus i y you sel . To do so, simply con ac he lead
esea che .
The indi iduals esponsible o his su ey ensu e ha you pa icipa ion o wi hd awal has no nega i e consequences o you. Thus, you
esponses and iden i ying da a will only be known o he su ey eam. The da a om his su ey will only be accessible o hose in cha ge o he
su ey and he esea che s associa ed wi h he p ojec o analysis pu poses, and hey will be secu ed h ough passwo d-p o ec ed access. The
esul s ha will be dissemina ed a he end o his p ojec will no allow you o be iden i ied, as only agg ega ed da a, ha is, global da a, will be
sha ed.
I any changes o he p ojec 's objec i es o he use o you esponses a e conside ed, we will in o m you o hese changes and eques you
consen again.
You ha e he op ion o indica e now whe he you would ag ee o ha e you da a po en ially eused o o he esea ch p ojec s (o he han he
cu en p ojec ). You e usal will no ha e any nega i e consequences o you and does no call in o ques ion you pa icipa ion in his su ey. I
you ag ee, we will ask o you consen again o his euse.
Yes
Da a P o ec ion :
The in o ma ion collec ed by Ci ad as he da a con olle is ga he ed o he pu pose o s udying he ole o seed supply me hods in he esilience
and ood secu i y o a ming households, o esea ch pu poses.
This p ocessing is based on you explici consen , exp essed e bally and signed by he wi ness be o e he in e iews begin.
You da a will be kep o he du a ion o he s udy (5 yea s, om 2022 o 2027) and up o 30 yea s a e he end o he s udy. This ex ended
e en ion is ini ially necessa y o allow he exploi a ion o he da a o publica ions o 10 yea s a e he end o he s udy, hen o an addi ional
15 yea s o allow, i necessa y, compa a i e analyses by o he p ojec s. A he end o his e en ion pe iod, he da a will be des oyed.
The in o ma ion collec ed is in ended o he ARISER esea ch eam and subcon ac o s in ol ed in he s udy and may be communica ed o
pa ne s unde con iden iali y condi ions and wi hin he sole amewo k o ARISER. In acco dance wi h he applicable egula ions, conce ning
he Gene al Da a P o ec ion Regula ion (GDPR), you ha e he igh s o access, ec i ica ion, e asu e, and po abili y (when applicable)
conce ning you da a, as well as he igh o es ic and objec o hei p ocessing o legi ima e easons. You can exe cise hese igh s by
con ac ing he Da a P o ec ion O ice o he da a con olle a dpo@ci ad. . You also ha e he igh o lodge a complain a any ime wi h he
compe en au ho i y (in F ance, he CNIL).
Yes
P o ec ion o T adi ional Knowledge and Know-How Associa ed wi h Gene ic Resou ces :
In he con ex o access and bene i -sha ing (ABS), adi ional knowledge and know-how associa ed wi h gene ic esou ces e e o he
knowledge, inno a ions, and p ac ices o indigenous and local communi ies ela ed o gene ic esou ces. They esul om he expe ience gained
by people o e cen u ies, a e adap ed o local needs, cul u es, and en i onmen s, and a e passed down h ough gene a ions.
The collec ion o his adi ional knowledge and know-how equi es compliance wi h he p ocedu es se by he Nagoya P o ocol o 2010,
pa icula ly he ai and equi able sha ing o bene i s a ising om esea ch.
In he con ex o his su ey, you a e sha ing wi h us in o ma ion abou adi ional knowledge and know-how associa ed wi h local gene ic
esou ces. The p ojec has aken he necessa y s eps o comply wi h he Nagoya P o ocol.
A.2 A.2 CONSENT REQUEST
1. Ha e you ecei ed and unde s ood he in o ma ion ha was communica ed o you?
2. Ha e you been able o ask ques ions?
3. Do you unde s and ha you can decide o s op pa icipa ing in he su ey a any ime?
4. Do you ag ee o pa icipa e in he s udy?
5. Do you unde s and you igh s ega ding you pe sonal da a?
6. Do you ag ee ha he da a collec ed in his s udy can be used acco ding o he pu poses, e en ion pe iod, and ecipien s indica ed in he
in o ma ion le e ?
7. Do you ag ee ha you da a can be eused o ano he p ojec ? (i applicable)
8. Do you ag ee o be con ac ed again o ano he su ey? (i applicable)
Yes/no Yes
Does he esponden gi e hei consen e bally?
Unique choice Yes
Name o he in e iewe Lis o he local in e iewe s ec ui ed
B.
B. SURVEY IDENTIFICATION
da e Yes
Su ey da e
Unique choice Yes
Name o he su eyed illage Lis o he illages selec ed o he s udy
Unique choice Yes
Name o he su eyed hamle o neighboo hood Lis o hamle s o neighbo hoods s udied in he illage when applicable
geopoin No
GPS coo dina es o he house GPS coo dona es gi en by Kobo applica ion, o by GPS Ga min
Unique choice Yes
Fa m iden i ie (ID) Lis o he a ms selec ed o he s udy and iden i ied by an unique numbe
C.
C. FARM
C.1
C.1 FIELD
C.1.a
C.1.a Land s a us and ield su ace a ea
In ege o decimal No
Wha is he o al a ea o ag icul u al land ha you household owns?
Yes/no No
Does you household ha e a home ga den?
Mul iple-choice No
Wha a e he main c ops you g ow each yea in you home ga den? Lis o he mos commonly c ops species and ca ego ies o c ops species encoun e ed in he s udy a ea (e.g. Wa e melon, melon, peppe s, oma oes, and o he ege ables)
Yes
O e he las ag icul u al season :
In ege o decimal
How many ields in o al ha e you cul i a ed ?
( he o al numbe o ields includes owned ields, en ed ields, bo owed ields, and sha ec opped ields)
In ege o decimal
I you le some land uncul i a ed, wha o al a ea i ep esen s?
(e.g. land le allow) Di e en uni s o a ea measu emen a e p o ided (e.g. km², hec a es, m²)
C.1.b C.1.b Desc ip ion o each ield cul i a ed o e he las ag icul u al season
No
Ques ions asked o each ield cul i a ed o e he las ag icul u al season :
Mul iple-choice
Wha is he ype o land enu e ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea
In ege o decimal
Wha is he a ea o he ield ? Di e en uni s o a ea measu emen a e p o ided (e.g. km², hec a es, m²)
Mul iple-choice
Wha a e he di e en ypes o soil p esen in he ield ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea
Mul iple-choice
Wha a e he di e en species o ce eals, legumes, and ube s ha you cul i a ed in his ield du ing he p e ious ag icul u al season? Exhaus i e lis o c op species o ce eals, legumes and ube s in en o ied in he s udy a ea (e.g. maize, so ghum, mille , peanu , peas, beans, cassa a)
Yes/no
Did you allow he li es ock o g aze on he esidues om he p e ious c ops ?
Mul iple-choice
Wha a e he di e en species o ce eals, legumes, and ube s ha you sown in his ield o e he las ag icul u al season ? Exhaus i e lis o c op species o ce eals, legumes and ube s in en o ied in he s udy a ea (e.g. maize, so ghum, mille , peanu , peas, beans, cassa a)
Mul iple-choice
Wha a e he c op species, o he han ce eals, legumes, and ube s, ha you sown in his ield o e he las ag icul u al season ? Exhaus i e lis o he o he c op species in en o ied in he s udy a ea (e.g. seconda y c ops, bissap, ok a)
No
I se e al c op species ha e been sown in his ield o e he las ag icul u al season :
Unique choice
How a e hese c ops species associa ed in he ield ? Lis o he di e en c op species associa ions obse ed in he ields in he s udy a ea (e.g, in e c opping, mixed, in sepa e ed a eas)
C.1.c C.1.c Fa ming p ac ices
Yes
Ques ions asked o all he ield cul i a ed o e he las ag icul u al season :
Mul iple-choice
Wha ype o illage ha e you pe o med on you ields ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea (e.g. manual, mecanized)
Yes/no
Did you use chemical e ilize s ?
Mul iple-choice
Did you use manu e ? And how ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea (e.g. applica ion in all ields, applica ion in ields wi h in e ile soil)
Mul iple-choice
Did you use p oduc s o elimina e insec s, diseases, o ungi ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea (e.g. chemical p oduc , biologic p oduc )
No
I he e a e pe ennial species in you ields o e he las ag icul u al season :
Mul iple-choice
Wha a e he pe ennial species ? Lis o pe ennial species obse ed in he ields in he s udy a ea (e.g. mango ee, neem ee, jujube ee)
In ege o decimal
How many do you ha e app oxima ely ?
In ege o decimal
On how many ields a e hey p esen ?
(I p esen on all ields, please indica e he o al numbe o cul i a ed ields)
C.2
C.2 CROP SPECIES
Yes
Lis o c op species cul i a ed o e he las ag icul u al season
Mul iple-choice
Lis o all he species o ce eals, legumes, and ube s ha you cul i a ed ac oss all you ields :
(Do no include homega den c ops and do no o ge seconda y c ops occupying small a eas in he ields)
Exhaus i e lis o c op species o ce eals, legumes and ube s in en o ied in he s udy a ea (e.g. maize, so ghum, mille , peanu , peas, beans, cassa a)
Yes/no
Do you con i m ha you ha e no cul i a ed he c ops ha we e no checked in he p e ious lis ?
( he in e iewe should lis all unchecked c ops be o e answe ing and ensu e ha no c op species we e o go en o added)
Yes
Ques ion asked o each species o ce eal, legume, and ube checked in he p e ious lis :
In ege o decimal How many a ie ies o his specie did you sown ac oss all you ields o e he las ag icul u al season ?
C.2.a/b/c C.2.a/b/c CROP VARIETY
Yes
Sec ions C.2.a/b/c : ques ions asked o each a ie y coun ed o each species o ce eal, legume, and ube checked in he p e ious lis
C.2.a C.2.a Named a ie y, sowing and ha es ing
Yes
O e he las ag icul u al season :
Unique choice
Wha is he name o he a ie y you sown ? Exhaus i e lis o he 28 named a ie ies o ce eals, legumes and ube s in en o ied in he s udy a ea
Unique choice
Wha is he na u e o he a ie y you sown ? Lis o he di e en cases and possibili ies encoun e ed in he s udy a ea (e.g. local a me a ie y, in oduced a ie y)
Unique choice
Would you spon aneously use one o hese e ms o desc ibe his a ie y ? (lis adap ed o each s udy si e, illus a ion in Senegal s udy si e) possible choices : local ( hiossane, cosaan) in odui e ( écemmen in odui e) ce i iée, ne sai pas
Yes/no
Did you pa en s al eady cul i a ed his a ie y ?
Yes/no
Is i he i s ime you g ow his a ie y ?
Yes/no
Did you cul i a ed i las yea ?
Unique choice
E en i you don' g ow i each yea , since app oxima ely how long a e g owing i ? possible choices : ...< 5 yea s; 5<...<10; 10<...<20; 20<...<40; 40<...
In ege o decimal
Wha quan i y did you sown ? Di e en uni s o weigh measu emen a e p o ided (e.g. kilog am, onne, g am, 50 kg sack)
Unique choice
When did you sown : beginning, middle o end o which mon h o he yea ?
Yes/no
Did you do a e-sow ?
In ege o decimal
Wha quan i y did you ha es ed ? (en e 0 i he e was no ha es ) Di e en uni s o weigh measu emen a e p o ided (e.g. kilog am, onne, g am, 50 kg sack)
Unique choice
Do you coun you ha es in g ains, shells, ea s, o pods ?
Unique choice
When did you ha es ed : beginning, middle o end o which mon h o he yea ?
Mul iple-choice
I no ha es : how do you explain ha ? Lis o 22 main easons desc ibed by he a me s o he s udy a eas (e.g. soil e ili y issues, excessi e ain all, no ge mina ion due o poo -quali y seeds)
Yes
I ha es ed :
Unique choice
A e you sa is ied wi h he ha es ?
Mul iple-choice
I he ha es was poo o unsa is ac o y : in you opinion, wha a e he easons o his? Lis o 22 main easons desc ibed by he a me s o he s udy a eas (e.g. soil e ili y issues, excessi e ain all, no ge mina ion due o poo -quali y seeds)
Mul iple-choice
I he ha es was sa is ac o y o e y sa is ac o y : In you opinion, wha a e he easons o his? Lis o 8 main easons desc ibed by he a me s o he s udy a eas (e.g. a o able wea he condi ions, good-quali y seeds)
Unique choice
Using he e ms "no a all, a li le, a lo , all," es ima e he sha e o he o al ha es ha was in ended o :
-sale
-human consump ion
-animal eed
- o be used as seeds o he ollowing yea
Mul iple-choice
Why do you p e e his a ie y o e ano he ?
O wha a e he ad an ages o cul i a ing his a ie y ? Lis o 23 main easons desc ibed by he a me s o he s udy a ea (e.g. High g ain yield, high selling p ice, esis an o pes s)
C.2.b C.2.b In o ma ion sou ces
Unique choice Yes
Whe e did you knowledge abou he a ie y come om ha made you decide o cul i a e i ?
(selec he main sou ce o in o ma ion)
-F om pe sonal expe ience
-F om obse ing a neighbo 's ield
-Th ough amily adi ion/gene a ional knowledge ans e
-F om common p ac ices among a me s in he illage, commune, o egion
-F om ad ice gi en by a a me
-F om ad ice gi en by a selle (ma ke , local shop, specialized s o e)
-F om ad ice gi en by an ag icul u al p ojec (go e nmen , NGO, coope a i e, associa ion)
-F om in o ma ion sha ed ia media ( adio, newspape , ele ision)
-F om in o ma ion sha ed on social media
-Can' emembe
ex Yes I he in o ma ion came om a p ojec , coope a i e, o a me s' o ganiza ion, wha is i s name?
Yes
Whe e was his sou ce o in o ma ion loca ed ?
(i.e., place o obse a ion, loca ion o he ma ke , s o e, a me who p o ided he ad ice, o p ojec )
(O icial lis s a ailable on he go e nmen websi e)
Unique choice
Name o he egion in he s udy coun y Lis o he egions o Senegal
Unique choice
Name o he depa emen in he s udy egion Lis o he depa men s in he s udy egion (Ka ine)
Unique choice
Name o he illage in he s udy depa men Lis o he illages/ci ies in he s udy depa m en (Ka ine)
Wha e e he answe gi en o he ype o in o ma ion sou ce:
Unique choice
Wha is he na u e o you ela ionship wi h he pe son who ga e you he ad ice, o wi h whom you made he obse a ion ?
-I don' know hem.
-I don' know hem pe sonally, bu I ecognize who hey a e (i 's an acquain ance).
-I 's a iend.
-I 's a amily membe
Unique choice
I he sou ce o in o ma ion was loca ed in he s udy depa men : is he pe son a neighbo ing a me ? someone om he illage ?
- ield neighbo
- illage neighbo
-bo h
nei he
Yes
I he pe son associa ed wi h he sou ce o in o ma ion is someone known, a iend, o a amily membe :
Unique choice
Gende
ex
Las name
ex
Fi s name
Unique choice
Age ca ego y Lis o 6 age ca ego ies (< 20 yea s, 20-30, 30-40 yea , 40-50, 50-60, and > 60 yea s)
Yes
I he pe son is a amily membe :
Unique choice
Does he pe son belong o you own amily o o you spouse's amily ?
Unique choice
Wha is he na u e o he kinship ie ? Lis o kinship ies ac oss h ee gene a ions acco ding o he gende o he esponden (e.g mo he , a he , b o he , sis e , g andmo he (ma e nal/pa e nal), g and a he (ma e nal/pa e nal))
C.2.c C.2.c Seed sou ces
Yes
O e he las ag icul u al season :
In ege o decimal
How many di e en sou ces (o seed lo s) do he seeds o he a ie y you cul i a ed come om ?
(seed lo : all he seed o he same a ie y ob ained om he same sou ce on he same da e by he household)
Yes
Ques ions asked o each seed lo coun ed :

Unique choice
Was his seed lo sel -p oduced (seeds kep om a p e ious ha es ), o sou ced ou side he a m las yea ? -Sel p oduced
-Sou ced ou o he a m
Yes
I he seeds we e sel -p oduced on he a m:
Unique choice
How did you selec you seeds?
-Selec ion in he ield
-Selec ion a he g ana y/s o age (s o ed ha es bags)
-O he
Unique choice
Whe e did you s o e you seeds be o e using hem o he las ag icul u al season ?
-T adi ional g ana y (sakk)
-Keu gui s o age
-Deukêbi s o age
-Household s o age
-In a ee
-O he
Unique choice
When was he las ime you ob ained hese seeds?
(i.e. las acquisi ion be o e sel -p oduc ion)
-Can' emembe
-Be ween 2 and 5 yea s ago
-Be ween 5 and 10 yea s ago
-Mo e han 10 yea s ago
-Inhe i ed
Yes
I he seeds ha e been inhe i ed (answe "inhe i ed" abo e l.108) :
Unique choice
Does he pe son belong o you own amily o o you spouse's amily ?
Unique choice
Wha is he na u e o he kinship ie ? Lis o kinship ies ac oss h ee gene a ions acco ding o he gende o he esponden (e.g mo he , a he , b o he , sis e , g andmo he (ma e nal/pa e nal), g and a he (ma e nal/pa e nal))
ex
Las name
ex
Fi s name
Unique choice
Gende
Unique choice
Age ca ego y Lis o 6 age ca ego ies (< 20 yea s, 20-30, 30-40 yea , 40-50, 50-60, and > 60 yea s)
Yes
I he seeds we e no sel -p oduced on he a m (answe "ou o he a m" l.104) :
Mul iple-choice
Why did you es ock you seeds ? Lis o 8 main easons desc ibed by he a me s o he s udy a ea (e.g. los o he a ie y, es o a new a ie y, oppo uni y)
Unique choice
Whe e do he seeds o he a ie y you cul i a ed o e he las e ag icul u al season come om ?
-Louma (weekly ma ke )
-Daily ma ke in he illage
-Specialized s o e (seed and ag icul u al supplies)
-Village shop
-Fa me
-Fa me s' o ganiza ion (associa ion, coope a i e)
-Ag icul u al p ojec (go e nmen , NGO)
-Seed mul iplie (ce i ied seed p oduce )
-Can' emembe
Unique choice
How we e hese seeds ob ained: pu chase, dona ion, ba e , o in exchange o o he seeds?
Unique choice
I he seed sou ce is a ma ke , wha ype o selle was i ?
-local a me s selling punc ually he own ha es
-local seed eselle s (buying and eselling seeds in hei communi y)
-ex e nal seed selle s
-o he
ex
I he seed sou ce is a specialized s o e, illage shop, p ojec (go e nmen , NGO), coope a i e, o a me s' associa ion: wha is i s name ?
Yes
Wha e e he answe gi en o he ype o seed sou ce: whe e was i loca ed ? (O icial lis s a ailable on he go e nmen websi e)
Unique choice
Name o he egion in s udy coun y Lis o he egions o Senegal
Unique choice
Name o he depa emen in s udy egion Lis o he depa men s in he s udy egion (Ka ine)
Unique choice
Name o he illage in s udy depa men Lis o he illages/ci ies in he s udy depa m en (Ka ine)
Yes
Wha e e he answe gi en o he ype o seed sou ce:
Unique choice
Wha is you ela ionship wi h he pe son om whom you ob ained he seeds ?
-I don' know hem.
-I don' know hem pe sonally, bu I ecognize who hey a e (i 's an acquain ance).
-I 's a iend.
-I 's a amily membe
Unique choice
I he seed sou ce was loca ed in he s udy depa men : is he pe son a neighbo ing a me ? someone om he illage ?
- ield neighbo
- illage neighbo
-bo h
nei he
Yes
I he pe son associa ed wi h he seed sou ce is someone known, a iend, o a amily membe :
Unique choice
Gende
ex
Las name
ex
Fi s name
Unique choice
Age ca ego y Lis o 6 age ca ego ies (< 20 yea s, 20-30, 30-40 yea , 40-50, 50-60, and > 60 yea s)
Yes
I he pe son is a amily membe :
Unique choice
Does he pe son belongs o you own amily o o you spouse's amily ?
Unique choice
Wha is he na u e o he kinship ie ? Lis o kinship ies ac oss h ee gene a ions acco ding o he gende o he esponden (e.g mo he , a he , b o he , sis e , g andmo he (ma e nal/pa e nal), g and a he (ma e nal/pa e nal))
Yes
Excep in cases whe e he seeds we e gi en as a dona ion o inhe i ed
Mul iple-choice
Why did you choose his sou ce o seeds?
(I he esponse ela es o us , inqui e u he abou he basis o ha us be o e eco ding he answe )
Lis o 14 main easons desc ibed by he a me s o he s udy a ea (e.g. unique sou ce ha p o ides hose seeds, ecommended seed sou ce, lowe seed selling p ice)
D. D. HOUSEHOLD
D.1 D.1 HOUSEHOLD MEMBERS
D.1.a No D.1.a Household head cha ac e is ics
ex
Fi s name
ex
Las name
In ege o decimal
Phone numbe
In ege o decimal
Age
Unique choice
Gende
Unique choice
Na i e language g oup (lis o g oups adap ed o each s udy si e, illus a ion in Senegal s udy si e) Exhaus i e lis o he di e en e hic g oups iden i ied in he s udy a ea (e.g. Se ee , Peulh)
Unique choice
Region, depa men and illage o bi h O icial lis s a ailable on he go e nmen websi e
Unique choice
Educa ion le el
In ege o decimal
How many yea s ha e you been a ming as he head o he ag icul u al ope a ion?
Mul iple-choice Do you ha e a peci ic social s a us wi hin he illage ? (o se e al, adap ed o each s udy si e, illus a ion in Senegal s udy si e) Lis o he di e en cases encoun e ed in he s udy a ea (e.g. illage chie , neighbo hood chie , lineage chie , no able, p esiden o a illage associa ion)
Yes/no
A e you a bene icia y o one o mo e ag icul u al p ojec s?
(e.g.NGO p ojec , go e nmen p ojec )
Mul iple-choice
Name o he p ojec s Lis o ag icul u al p ojec s iden i ied in he s udy a ea
Yes/no
A e you a membe o one o mo e a me s' o ganiza ions?
(e.g.coope a i e, associa ion)
Mul iple-choice
Name o he a me s' o ganiza ion Lis o a me s' o ganiza ions iden i ied in he s udy a ea
D.1.b No D.1.b Household composi ion
In ege o decimal
How many women a e you ma ied wi h and li e in he household ?
In ege o decimal
How many people, li e in you household h oughou he yea o o a signi ican pa o he yea ?
(O he amily membe s o no , welcomed by he head o he household, p esen a all meals o e y egula ly)
Fo each spouse coun ed :
Unique choice
Na i e language g oup Exhaus i e lis o he di en e hic g oups iden i ied in he s udy a ea (e.g. Se ee , Peulh)
Unique choice
Region, depa men and illage o bi h O icial lis s a ailable on he go e nmen websi e
In ege o decimal
Age
D.2 D.2 SOCIOECONOMIC DATA
D.2.a No D.2.a Household income-gene a ing ac i i ies
Mul iple-choice
Wha a e he di e en ac i i ies o he men and sons in you household ha con ibu e solely o he household income ? Lis o he di e en ac i i ies desc ibed by he a me s in he s udy a ea (e.g. ailo , mechanic, ca pen e , ma ke ga dene )
Mul iple-choice
Wha a e he di e en ac i i ies o he women and daugh e s in you household ha con ibu e solely o he household income ? Lis o he di e en ac i i ies desc ibed by he a me s in he s udy a ea (e.g. ailo , mechanic, ca pen e , ma ke ga dene )
In ege o decimal
Can you ank he di e en ac i i ies men ioned in descending o de o impo ance in e ms o income gene a ed?
(Use he same numbe i mul iple ac i i ies ha e he same impo ance)
D.2.b No D.2.b Household esou ces
Mul iple-choice Does you household ha e hese asse s ? (lis o asse s adap ed o each s udy si e, illus a ion in Senegal s udy si e)
-House made o mud o b icks
-Me al o du able ence
-Sola panel
-Tele ision
-Jaka a scoo e
-Well
-Access o elec ici y ia he collec i e g id
-None
Yes/no
Has you household e e aken ou a loan o bo owed money?
Yes/no
A e he e people who don' li e in he house, bu who ecei e money and/o p oduce om you a m ?
Yes/no
A e he e people who don' li e in he house bu send you money and/o ood?
D.2.c No D.2.c Domes ic and li es ock animals (adap ed o each s udy si e)
In ege o decimal
How many d a animals does you household own ? (e.g.donkeys, ho ses, oxen - adap ed o he s udy si es)
In ege o decimal
App oxima ely, wha is he size o you sheep lock (in numbe o heads) ?
In ege o decimal
App oxima ely, wha is he size o you ca le he d (in numbe o heads) ?
D.2.d D.2.d Labou o ce
Yes
O e he las ag icul u al season :
Yes/no
Did you use hi ed wo k o ce o cul i a e he ields ?
In ege o decimal
I yes : wha is he o al amoun o money you es ima e o his expense ela ed o hi ed wo k o ce (in na ional cu ency) ?
In ege o decimal
Including he people in you household and hose li ing in he amily compound, how many people in o al (including you sel ) helped wi h he
wo k in he a m ?
D.3 D.3 PERCEPTIONS OF LIFE CONDITIONS AND FOOD SECURITY
Yes
O e he las ag icul u al season :
Yes/no
Did you ha e any di icul ies accessing he seeds you wan ed o cul i a e ?
Mul iple-choice
I yes : o which c op species ? Exhaus i e lis o c op species o ce eals, legumes and ube s in en o ied in he s udy a ea (e.g. maize, so ghum, mille , peanu , peas, beans, cassa a)
Yes/no
Ha e you hea d abou o he a ie ies ha migh in e es you ?
Yes/no
I yes: would i be easy o you o ob ain in o ma ion abou hese o he a ie ies ?
Unique choice
In ligh o he di icul ies ha amily a ming ace in you a ea, do you hink i is be e o mee you needs sa is ac o ily by:
-Does no exp ess an opinion
-Cul i a e mul iple species
-Focus on a educed numbe o species
Unique choice
In you opinion, he income o you household and you p oduc ions:
-Did no mee he amily's needs ( ood, heal h/illness...)
-Only ba ely me he amily's needs ( ood, heal h/illness...)
-Me he amily's needs ( ood, heal h/illness...)
-Ve y well me he amily's needs ( ood, heal h/illness...)
Yes/no
Ha e you been o ced o change you ea ing habi s due o inancial di icul ies?
Unique choice
I yes: how do you assess he signi icance o hese changes?
-Low
-Medium
-High
Yes/no
Ha e you had di icul ies managing ce ain ag icul u al expenses ?
Mul iple-choice
I yes: wha a e hese ag icul u al expenses? Lis o he di e en expenses desc ibed by he a me s o he s udy a ea (e.g.paying o labo , ac o , inpu s)
Unique choice
I yes: how do you assess he signi icance o hese di icul ies ?
-Low
-Medium
-High
Yes/no
Ha e you had di icul ies managing o he expenses ela ed o he amily's li ing condi ions?
(e.g. school ees, heal h expenses, expenses o a ce emony)
Unique choice
I yes: how do you assess he signi icance o hese di icul ies ?
-Low
-Medium
-High