5. Acksen et al.

Lazareto

).6

E
ie
€
2
70.2

0:0 sim
CC NO OO TEN
mNNOAO OO TA
> CCSCCEC
© © OU S53335355)5
5555352357235
Survey date
Säo Pedro

2.015|
5 0.010
£ |
2
= 9.005

9.000 =. WW}
POS EL ONDODOOKK
BRNO OO TFNNAN
>> m A ACC CCCCSC£
ee O0 SO I3IQ735)5
Survev date

I

0.6

Marine Pollution Bulletin 228 (2026) 119525

Salamansa
a

Baia das Gatas

1.151

13.0

in
E
9
=09.05

Os
* S
15 5
B

E04
3
E
92

A

0.00L_ "alas l E10 0.0|
5» MN S NO OO Fk
"NN Orr NQA
ARM CCCCCLC
© US 5S5353Z52F5
Survey date
Palha Carga
0.0020
© 0.0015
£
a 90-0010
3.0005 ''

"FF

1

20 ADMO
NNAN SOOÖr NN
mm CCLCCCLCOCESL
ce © © 335353\353635 35333593
Survey date
Praia Norte

r

252
a’ 7

0.0000 |

NOOOK=-MDOMN
NNBSOOTTENN
> m ACC EL CCECE“LCL
5 OO 55332533T053I55
>53 >>7>>55>
Survev date

|
N 6
TO ND SOME
NN OO OT FNANANN
m CCCCCCCES£
eo © zz 2 33 35 3 SF
= > >23 5375 75772”
Survey date

Fig. 3. Abundance of beach debris items depicted as densities (dark grey) and masses (light grey) that were collected during one initial cleaning (red) and nine
‚ollow-up surveys on six beaches of Säo Vicente. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
chis article.)

0
300
2001

>earson r = 0.9871.8 -
p < 0.001

16.

ı A

1001

j\ 2

£&
R
0

0 Ts

PN L

S BG
Beach

SP PC

ES

Fig. 4. Abundance of mesoplastic particles (dark grey) and of macroplastic
‚tems (light grey) at the six beaches of Säo Vicente (PN: Praia Norte, L: Lazareto,
S: Salamansa, BG: Baia das Gatas, SP: Säo Pedro, PC: Palha Carga). Data were
averaged across the beach sections (mesoplastics) and across the subsequent
surveys (macroplastics), respectivelv.

was the most common (69%), followed by PP (30%) and PS (1%). Only
one PVC particle was detected in Praia Norte (Fig. 5C).
A generalized linear model (GLM) with a negative binomial distriution
 and type III sums of squares was used to assess the effects of
“Beach” and “Beach section” on mesoplastic densities. Both main effects
were highly significant (“Beach”: LR y* = 164.90, df = 5, p < 0.001;
“Section”: LR vn = 55.11, df = 2, p < 0.001), as was the interaction
vetween the two factors (LR X = 83.52, df = 10, p < 0.001), indicating
‘hat the influence of “Section” on mesoplastic densities differs among
beaches (Supplement Table A4).

3.3. Ocean modeling

Backward simulations with the Lagrangian ocean tracer model
PARCELS show that the virtual particles mimicking light and small
olastic items arriving in the release zones around the shores of Säo
Vicente originate from the northwest in the last 100 days before their
arrival (Fig. 6A). 500 days before their arrival most particles are

accumulated close to the West African Coast and in the southeast of the
Sabo Verde archipelago (Fig. 6B). 1000 days before reaching the shores
9f Sao Vicente most particles are spread further along the African and
South European coast, while some passed the Atlantic along the equator
towards the Brazilian coastline (Fig. 6C). 5000 days before their arrival
particles are far spread in the Atlantic Ocean without any clear pattern
visible. It is to note that the simulated particles in PARCELS cannot enter
land cells and are thus recirculated in the ocean. Particle pathways towards
 the release zones around the islands only differed significantly
during the last two weeks before their arrival at the shores of Sao Vicente
Supplement Fig. A8). Before that, they all followed highly similar
yathways.
Particle trajectories over the 500 days preceding arrival at the coast
>f Sao Vicente reveal a distinct pathway from the Cabo Verde archipelago
 towards the West African coast. That is why, the West African
shore was examined closer as a possible important origin of particles
‚Fig. 7A). A minimum of 150 days was needed for the particles to travel
from the West African shore to Säo Vicente. On longer timescales prior to
reaching the Cabo Verde archipelago, more than 20% of all particles
were located on the West African shelf, indicating that most particles
arriving at Säo Vicente travelled along the shelf and remained there for a
zubstantial period (Fig. 7B). Most particles reaching Sao Vicente left the
African shelf at latitudes between 10 and 15° N, with a maximum of 40%
leaving the shelf at 13° N (Fig. 7C). This region comprises the coastlines
9f Senegal, Gambia, Guinea-Bissau and Guinea.
4. Discussion

4.1. Overall pollution level on Sao Vicente

We assessed the abundances of beach debris during initial cleanings
“Supplement Appendix B) as well as beach debris accumulation rates
(‚hereafter on six beaches on Säo Vicente, Cabo Verde. The debris
accumulation rate, when averaged across all surveys and beaches, was
2.05 + 0.06 items/m?/d (0.4 + 0.2 g/m?/d). In the initial cleanings,
which allowed to assess the quantity of debris that accumulated during
an unknown time period, we detected a 16 times higher load