See discussions, s a s, and au ho p o iles o his publica ion a : h ps://www. esea chga e.ne /publica ion/384687258
Public Pe cep ion o Biodi e si y Landscape Elemen s and Au onomous
Technologies in Small-Scale P oduc ion Sys ems
Con e ence Pape · Sep embe 2024
CITATIONS
0
READS
33
3 au ho s, including:
And eas Gab iel
Baye ische Landesans al ü Landwi scha
56 PUBLICATIONS465 CITATIONS
SEE PROFILE
All con en ollowing his page was uploaded by And eas Gab iel on 07 Oc obe 2024.
The use has eques ed enhancemen o he downloaded ile.
28 29 h Sep embe 2024, Reading, Uni ed Kingdom
Ha pe Adams Business School
Ha pe Adams Uni e si y
Uni e si y o Reading
h ps://www.ha pe -adams.ac.uk/ esea ch/gia e/
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 12
Public Pe cep ion o Biodi e si y Landscape Elemen s and
Au onomous Technologies in Small-Scale P oduc ion
Sys ems
And eas Gab iel, Johanna Ga ni z and Oli ia Spykman
Ba a ian S a e Resea ch Cen e o Ag icul u e, Ge many
Abs ac
The pe cep ion and e alua ion o u al landscapes esul ing om human in e ac ion wi h
na u e is highly subjec i e. Howe e , unde s anding how he non-ag icul u al popula ion
iews he impac o an al e ed landscape image is c ucial. This pape explo es he Ge man
popula ion's pe cep ions o changes in ag icul u al landscapes b ough abou by mul i-c op,
small-scale ield s uc u es (s ip in e c opping) combined wi h he in oduc ion o
biodi e si y landscape elemen s and ield obo ics. An online su ey was conduc ed wi h
Ge man esiden s aged 18 and olde (n = 2,022). P e e ences and he impo ance o indi idual
image componen s we e analysed based on ou images depic ing a ield wi h s ip
in e c opping, ea u ing a ious combina ions o ac o s, obo s, and lowe ing s ips.
indings e eal ha nea ly wo- hi ds o esponden s p e e ed he image ea u ing a lowe
s ip and a ac o , associa ing i wi h concep s such as g een, na u e, and en i onmen
( lowe ing s ip), as well as he adi ional image o ag icul u e ( ac o ). Among he wo
images wi hou lowe s ips, he ac o was p e e ed o e he obo by mo e han a six old
ma gin. Con e sely, he image wi h a obo and lowe s ips was chosen abou as equen ly
as he image wi h a ac o bu wi hou lowe s ips. Addi ionally, he s udy highligh s how
socio-demog aphic cha ac e is ics may in luence he e alua ion o ag icul u al landscape
changes. Two logis ic eg ession models indica e ha ac o s such as age, gende , di ec
impac p e e ences o speci ic landscape componen s. O e all, he esul s sugges a
p e e ence o landscapes ha a e bo h amilia and en i onmen ally o ien ed. Ne e heless,
he use o au onomous echnologies and he shi owa ds small-scale di e si ied p oduc ion
sys ems a e no b oadly ejec ed.
Keywo ds
Au onomous a ming echnologies; biodi e si y; public accep ance; u al landscape; s ip
in e c opping.
P esen e P o ile
And eas Gab iel is a membe o he 'Digi al Fa ming' wo king g oup a he Ba a ian S a e
Resea ch Cen e o Ag icul u e. Wi h ex ensi e expe ience in empi ical social esea ch, his
wo k ocuses on in es iga ing he social accep ance and adop ion o digi al echnologies in
ag icul u al p ac ice.
* Co esponding Au ho : And eas Gab iel, Ba a ian S a e Resea ch Cen e o Ag icul u e,
Ins i u e o Ag icul u al Enginee ing and Animal Husband y, 94099 Ruhs o , Ge many;
email: and e[email p o ec ed].de
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 13
In oduc ion
The isual pe cep ion o a u al landscape ("landscape image") is an impo an ac o o
accep ance o indi idual ea u es in ag icul u al s uc u es among bo h ag icul u al
s akeholde s and he gene al public. This pe cep ion is in luenced, e.g. by associa ed a ming
p ocesses and en i onmen al e ec s and is s ongly shaped by he subjec i e pe spec i e o
he indi idual (c . Ro h e al., 2011). The e o e, in e ac ions such as he in oduc ion o new
p oduc ion sys ems and s uc u al elemen s (e.g., ag o o es y sys ems, lowe s ips, e c.) o
he use o new echnologies (e.g., ield obo s) o p omo e ecological sus ainabili y mus also
be discussed and e alua ed in e ms o i s impac on he landscape.
In ligh o cu en e o s o p omo e biodi e si y-enhancing p oduc ion sys ems (FAO, 2023;
Rugge i Lade chi e al., 2024), small-scale di e si ied c op p oduc ion sys ems such as s ip
in e c opping a e gaining impo ance (c . Ala cón-Segu a e al., 2022; Spykman e al., 2023).
S ip in e c opping e e s o he simul aneous cul i a ion o di e en c ops on he same ield
in pa allel s ips (Vande mee , 1989). I es ablished on a la ge scale, his p oduc ion sys em
has a - eaching impac s on he landscape image compa ed o con en ional a ming. I is
assumed ha he managemen o such small-scale di e si ied p oduc ion sys ems can be
made labou -e icien h ough au oma ion (e.g., au oma ic s ee ing sys ems and sec ion
con ol), o by using au onomous echnologies such as ield obo s o d ones (c . Lowenbe g-
DeBoe , 2021; Gacks e e e al., 2023). Pa icula ly, he in oduc ion o au onomous
echnologies would u he change bo h he aes he ic appea ance o he landscape and
ag icul u al p ac ices.
P e ious esea ch has demons a ed he in luence o use expe ience and knowledge abou a
Den zmann and Goldbe ge (2020) examined images o a biodeg adable al e na i e o
con en ional polye hylene mulching oil in ocus g oup discussions wi h a me s. I was ound
ha he e alua ion o his al e na i e was s ongly dependen on he expe iences o he
esponden s, wi h unc ional knowledge in luencing he isual assessmen (Den zmann and
Goldbe ge , 2020). The isual assessmen o he landscape image wi hin he p o essional
g oup is hus also based on knowledge abou a ming me hods, hei easibili y, and economic
p ospec s.
Howe e , i is no easy o de e mine how g oups ha a e no amilia wi h he ope a ional
unc ions o landscape-shaping a ming measu es will eac o changes in he landscape.
Posi i e ecological e ec s o en occu as pa o conse a ion measu es associa ed wi h
"diso de ", bu hese measu es do no necessa ily diminish a ce ain p e e ence o " idy"
landscapes and amilia landscape images. In his ega d, a me s di e om he non-
ag icul u al socie y in hei pe cep ion and e alua ion o he landscape (Bu on, 2012). In
con as o a me s, he non-ag icul u al socie y pa ly e alua es linea i y in landscape images
as nega i e and "unna u al" (La oche e al., 2018). The aes he ic pe cep ion weighs hea ie
han o he e alua ion c i e ia such as ag icul u al p oduc ion o conse a ion. I is pos ula ed
ha plan ing na u al elemen s (e.g., bushes) in linea , s uc u ed cul i a ion o ms (e.g.,
s aigh ows) can e oke eelings o "cul u al dissonance" (La oche e al., 2018). Howe e , he
ype o landscape image cul u ally es ablished is ele an in his con ex . Fo example, an
ag o o es y sys em wi hin adi ional o cha ds gene a es highe accep ance (e.g., measu ed
in highe willingness o pay) i mo e han one c op is g own be ween he ee ows (Alcon e
al., 2020), i.e., i mo e s uc u es a e p esen . Howe e , no only he isual quali y o he
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 14
landscape was e alua ed, bu also he associa ed ecosys em se ices and cul u al he i age,
ep esen ed by manual managemen as opposed o a ac o (Alcon e al., 2020). This
app oach also poin s o he complex in e play o isual pe cep ion and associa ed p ocesses
o he non-ag icul u al popula ion. Wa en-K e zschma and Von Haa en (2014) emphasize
he ele ance o posi i e isual e alua ion by socie y as an impo an aspec besides he
ecological bene i s o ag icul u al p ac ice. This likely also gene a es accep ance o a change
in he cul u al landscape.
In addi ion o changes in he landscape image h ough new ag icul u al sys ems o s uc u al
elemen s, an impac om he use o echnologies in he ields is expec ed. Al hough
au onomous echnologies such as ield obo s a e associa ed wi h a ious bene i s, including
educed labou cos s (Lowenbe g-DeBoe e al., 2021), a su ey o a me s showed ha
conce ns abou a nega i e image o "aliena ed ag icul u e" in he popula ion can in luence
he planned acquisi ion o ield obo s (Spykman e al., 2021). P e ious esea ch on he
popula ion sugges s ha ield obo s end o be a ed neu al o posi i e (P ei e e al., 2020).
Howe e , Willmes e al. (2022) desc ibe a nega i e impac on he willingness o pay o ood
p oduced wi h he help o digi al echnologies. The au ho s add ha his nega i e impac can
be educed by addi ional ecological bene i s o he echnologies. These indings a e e lec ed
in a choice expe imen on au onomous echnologies in weed con ol, whe e he me hod o
weed con ol (mechanical s. he bicide b oadcas and spo -sp aying) in luenced he decision
mo e han he deg ee o au onomy o he echnologies used (Spykman e al., 2022). Howe e ,
a join conside a ion o au onomous echnologies and al e ed p oduc ion sys ems has no ye
been unde aken.
The aim o his pape is o analyse he pe cep ion o he Ge man popula ion ega ding new
small-scale di e si ied p oduc ion sys ems, he in eg a ion o s uc u al (biodi e si y)
elemen s such as lowe s ips, and he use o au onomous echnologies such as ield obo s
using an online su ey. A special ocus is on iden i ying and e alua ing he igge s o po en ial
p e e ences and he connec ions o indi idual isual componen s. This is done by ca ego izing
sho associa ions p o ided by su ey pa icipan s in connec ion wi h hei p e e ence
decisions. Fu he mo e, his pape also includes a segmen a ion analysis and illus a es how
a ious sociodemog aphic cha ac e is ics o he popula ion in luence he e alua ion o
ag icul u al elemen s such as lowe s ips o he use o au oma ed echnologies.
Me hods
Online su ey among he Ge man popula ion
A na ionwide online su ey o he Ge man popula ion aged 18 and olde was conduc ed om
mid-Sep embe o mid-Oc obe 2023. Access o his consume panel was acili a ed h ough
he engagemen o a ield se ice p o ide . The use o a consume panel allows he sepa a ion
o pe sonal da a and con en da a, so ha esea ch e hics can be assu ed. The panel enables
a p e-s a i ica ion o he sample o ensu e ha pa icipan s we e ep esen a i e o he
Ge man popula ion in e ms o age, gende , size o esiden ial a ea, and ede al s a e. In
addi ion o a ious sociodemog aphic da a, in o ma ion on leisu e ac i i ies in u al a eas,
pe sonal connec ions o ag icul u e, a i udes owa ds echnology, local ood p oduc ion and
sus ainable consump ion, and knowledge o ag icul u e, was ga he ed using es ablished
ma ke esea ch me hods. A e he inal da a alida ion, he su ey sample comp ised 2,022
usable and comple ed da a se s.
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 15
Analysis o p e e ences, mo i es, and sho associa ions
In a ques ion se ega ding isual e alua ion, pa icipan s we e asked o assess a ious aspec s
o a landscape image wi h s ip in e c opping using ou pho omon ages. All ou image
a ian s we e based on an iden ical s ip in e c opping image which shows a machine passage.
The di e ences included he use o a ield obo ins ead o a ac o and he p esence o lowe
s ip. All ou images we e pho omon ages ha we e delibe a ely no ealis ic (sligh ly
di e gen size o he machines) bu we e designed o inc ease he ecognizabili y o he
a ious componen s o pa icipan s (Table 1).
Table 1: Choice o image a ian s o esponden s
No e: Image sou ces: Pho omon ages, Ba a ian S a e Resea ch Cen e o Ag icul u e, 2023.
The ou images we e p esen ed simul aneously o he pa icipan s and wi hou andomised
a angemen s. A e selec ing hei p e e ed image a ian , pa icipan s we e asked o
choose h ee ou o six p ede e mined image componen s ha in luenced hei decision,
he la e componen e e s o he o de
and s aigh ness o he pa allel ield s ips as a s uc u ed o m o cul i a ion wi hou any
beau i ul ow o ees in he b
esponse. The selec ed image componen s we e coun ed and weigh ed acco ding o hei
speci ied ank ank 1 ecei ed a iple weigh , ank 2 double weigh s, and ank 3 single
weigh s. This app oach allows o he conside a ion o all h ee men ioned image componen s
and a composi e anking.
In a ollow-up ques ion, su ey pa icipan s we e asked o p o ide up o h ee sho
associa ions in he o m o keywo ds ela ed o he decisi e image componen ( i s ank).
These a he spon aneous associa ions o he pic u e componen s shown o e addi ional
insigh s in o he decisi e image componen and he choice o image a ian . While he anking
o p ede e mined image componen s se ed he cogni i e e alua ion by he pa icipan s, he
a ec i e and hus emo ion-based app oach o sho associa ions p o ides ano he dimension
o de e mining accep ance (Busch e al., 2019; P ei e e al., 2020; Lange e al., 2022). A e
da a cleaning, a o al o 4,872 usable keywo ds as spon aneous associa ions o he
componen s o he ou images we e a ailable. Mos o hese we e ela ed o image 4
( ac o /wi h lowe s ip), o which a o al o 3,092 keywo ds we e analysed and ca ego ised,
manually and in se e al i e a ions, in o 33 ca ego ies. F om hese, he 16 mos equen ly
men ioned ca ego ies (co e ing 2,995 keywo ds) we e iden i ied and p epa ed o his
con ibu ion.
Va ian s
Robo /
no lowe s ip
T ac o /
no lowe s ip
Robo /
wi h lowe s ip
T ac o /
wi h lowe s ip
Visualiza ion
o he image
a ian s o
selec ion
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 16
Modelling he ac o s in luencing p e e ences
Ano he goal o his con ibu ion is o iden i y possible sociodemog aphic in luences on he
p e e ence o one o he ou image a ian s. Based on simila s udies, i was assumed ha
pe sonal ac o s such as age, gende , size o esiden ial a ea, o li ing in a speci ic egion (e.g.,
Eas Ge many wi h la ge-s uc u ed landscapes) play a ole, as may he esponden s' di ec
connec ion o an ag icul u al en i onmen (De lin, 2005; Booga d e al., 2008; P ei e e al.,
2020). Addi ionally, he G een Consump ion Value (GCV), which e lec s he esponden s'
endency owa ds en i onmen ally iendly shopping beha iou , was used as a alue- and
a i ude-based ac o . This was measu ed using six i ems (Haws e al., 2014). These six i ems,
p esen ed in a Like - ype scale o ma , we e condensed in o an indi idual s anda dized ac o
sco e h ough ac o analysis and conside ed as a me ic p edic o o he selec ion o he
image a ian . Since esponden s could also choose be ween he use o a ac o and a ield
obo in he images shown, he a i ude owa ds echnology (ATT) was assessed using nine
i ems in a Like scale o ma and condensed in o a s anda dized ac o sco e (Edison and
Geissle , 2003). Fo bo h scales, nega i e ac o alues indica e a s onge mani es a ion o
his cha ac e is ic, while posi i e alues indica e a lowe mani es a ion. While he ypology o
su ey pa icipan s ega ding GCV is igh -skewed, indica ing ha pa icipan s' pu chasing
beha iou is p edominan ly en i onmen al-conscious acco ding o hei s a emen s, a i ude
owa ds echnology is mo e e enly dis ibu ed, showing a balanced a io be ween echnology-
o ien ed and ech-a e se esponden s (Figu e 1).
Figu e 1: Dis ibu ion o he ac o alues o g een consump ion alue (GCV) and he
a i ude owa ds echnology (ATT) o he esponden s (n = 2,022); GVC: median: -0.11;
skewness: 0.774; ku osis: 0.525; ATT: median: -0.09, skewness: 0.339; ku osis: -0.127
Mul i a ia e eg ession models de e mine he ela ionships be ween mul iple p edic o
a iables and a dependen a iable. Fo binomial and ca ego ical dependen a iables, logis ic
p ocedu es a e used o de e mine he p obabili y o he occu ence o non-occu ence o an
e en (e.g., selec ion o an image) based on he alues o he included p edic o a iables
(Backhaus e al., 2018). Logis ic eg ession p o ides in o ma ion abou he ans o ma ion o
he dependen a iable logi (p):
(1)
whe e p is he p obabili y ha he selec ion o a pa icula image is in luenced by he
-selec ion. The odds a io
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 17
ep esen s he a io o hese wo p obabili ies. When inco po a ing k p edic o a iables, he
model akes he ollowing o m:
(2)
The eg ession equa ion p o ides in o ma ion abou he impo ance o each p edic o based
), allowing o he c ea ion o a hie a chy o he measu ed
a iables' e ec s on g oup assignmen (Backhaus e al., 2018).
To cap u e he o e all e ec s and explana o y con ibu ion o he selec ed in luencing ac o s
ed. Fo his pu pose,
he image p e e ence was dummy coded as he dependen a iable (Model A: 1 = one o he
wo images wi h a obo was chosen; Model B: 1 = one o he wo images wi h a lowe s ip
was chosen). Sociodemog aphic cha ac e is ics included gende (1 = emale), age (1 = < 40
yea s), size o esiden ial a ea (1 = < 20,000 inhabi an s), geog aphical loca ion (1 = wes e n
Ge man s a es), and educa ional le el (1 = no gene al highe educa ion en ance
quali ica ion). Responden s' s a emen s ega ding pe sonal connec ion o ag icul u e was
included in he modelling ei he as pe sonal employmen in he sec o (1 = yes) o h ough
pe sonal con ac wi h ag icul u e in he ci cle o iends o acquain ances (1 = yes) (c . P ei e
e al., 2020). The me ic ac o sco es o GCV and ATT we e also in eg a ed in o he wo models
as addi ional independen cha ac e is ics.
Resul s
Dis ibu ion o p e e ences and selec ion mo i es
In a cen al ques ion, pa icipan s we e asked o e alua e changes in he landscape based on
single images, conside ing bo h he use o obo s ins ead o ac o s and he addi ional use o
lowe s ip. The o e all dis ibu ion o he s a ed p e e ences indica es ha he a ian wi h
lowe s ip in conjunc ion wi h ieldwo k pe o med by ac o s is p e e ed (Image 4 in Table
15.3% o he 2,022 esponden s, while 14.6% chose he obo in combina ion wi h lowe s ip
(Image 3 in Table 1). Only 2.5% o he su ey pa icipan s a ou ed image 1 (see Table 1), in
which he obo was depic ed on he ield wi hou lowe s ip.
Figu e 2 shows he esul s o he anking o he image componen s ha we e decisi e o he
pa icipan s' p e e ence choices. Fo images 3 and 4, which depic he obo and he ac o
espec i ely, he lowe s ip shown in bo h images is he p ima y componen (41% and 43%,
espec i ely).
This is ollowed by he echnical aspec obo o ac o wi h 27% and 28%, espec i ely.
also equen ly men ioned o bo h images 3 and 4, wi h 13% each. Fo images 1 and 2, which
depic he obo o ac o wi hou he lowe s ip, he ocus is p ima ily on he echnological
aspec , ci ed as he eason by 39% o he obo and 40% o he ac o . The second place in
he amilia image o adi ional ag icul u e and ha he lowe s ip is pe cei ed as a he
dis up i e o co e ieldwo k. Rega ding image 4 ( ac o wi h lowe s ip), some esponden s
ema ked ha he lowe s ip speci ically symbolizes na u e and animal conse a ion o
P oceedings o he 7 h Symposium on Ag i-Tech Economics o Sus ainable Fu u es 18
hem. Fo images 1 and 3, which show ieldwo k done by a obo , inno a ion, po en ial
e iciency gains, and no el y we e men ioned as dis inc mo i es o selec ion.
Figu e 2: Dis ibu ion o p e e ences and selec ion mo i es
Fac o s in luencing p e e ences
The wo binominal logis ic eg ession models examining he in luence o sociodemog aphic
cha ac e is ics on he selec ion o images wi h obo s and images wi h lowe s ip
demons a e dis inc e ec s. Gende in luences he selec ion o images wi h he obo (Table
2) as emale pa icipan s a e signi ican ly less likely o choose images wi h a obo compa ed
o men (Odds Ra io = 0.434). I esponden s ha e pe sonal con ac s wi h acquain ances in he
ag icul u al sec o , he likelihood o selec ing he obo image is signi ican ly lowe .
In e es ingly, esponden s wi h pe sonal ag icul u al expe ience exhibi an opposi e, hough
no s a is ically signi ican e ec . No addi ional in luence ac o s, such as he a i ude owa ds
echnology o o igin om wes e n o eas e n Ge man s a es, a ec he p e e ence o a ield
obo compa ed o a ac o .
P edic o s Model A Field obo
B SE Wald p
Odds
Ra io
95% CI
LL UL
Gende (1= emale)* -0.835 0.337
6.137 0.013
0.434
0.224 0.840
Age (1=younge han 40 yea s) 0.067 0.392
0.029 0.864
1.069
0.496 2.304
Educa ion (1=no A-le els and below) 0.100 0.332
0.090 0.764
1.105
0.576 2.118
Size o place o esidence (1=less han 20k
inhabi an s)
-0.288 0.342
0.706 0.401
0.750
0.384 1.467
Region (1=Wes e n Ge many s a es) -0.314 0.367
0.730 0.393
0.731
0.356 1.501
Own ag icul u al expe ience (1=yes) 0.405 0.670
0.365 0.545
1.499
0.403 5.572
Pe sonal con ac wi h a me s (1=yes)* -2.088 1.043
4.008 0.045
0.124
0.016 0.957
A i ude owa ds echnology (ATT)
(nega i e ac o alue = highe deg ee)
-0.034 0.141 0.059 0.808
0.966
0.732 1.275
G een Consump ion Value (GCV)
(nega i e ac o alue = highe deg ee)
0.087 0.178
0.239 0.625
1.091
0.770 1.546
Cons an *** -1.398 0.406
11.892
0.000
0.247
No e:
assignmen classi ica ion (con ibu ion o p edic o a iables) = 90.3% | Sou ce: own su ey
