Page 99 - AJWEP-22-4

Conservation of the vulnerable Vitellaria Paradoxa

States Department of Agriculture. However, in tropical the north. The mean annual temperature is around 26°C,
24
regions, total annual rainfall and its distribution play a with little or no seasonal variation. 1
more significant role in determining plant distribution
and survival, as temperature remains relatively uniform 2.2. Shea nut tree (V. paradoxa) occurrence data
with little or no seasonal variation. This is true for The occurrence data of V. paradoxa were obtained from
25
the study area in the present project, where the average field surveys and the Global Biodiversity Information
annual temperature rarely exceeds 30°C and shows little Facility (GBIF). Field data were collected from March
or no seasonal variation. to August 2021 and from October to November 2024,
This study seeks to answer the following questions: while GBIF data were accessed on November 26, 2024
(i) Where is the suitable habitat for V. paradoxa? (ii) What (occurrence data and citations in Appendix A1 and
are the essential vegetation phenotype parameters that Table S1). Special attention was paid to the identification
can guide the selection of viable/healthy seeds for of V. paradoxa in the field with the aid of local guides and
V. paradoxa propagation? and (iii) Can the locations experts. The data collected from both sources included
where viable seeds can be sourced and the associated species name, geographic coordinates (in decimal
seed zone in the study area be identified? Accordingly, degrees), sources of information (for data collected from
this study applied remote sensing imaging technologies GBIF), and the name of the closest settlement (only
to extract vegetation-related phenotypic and climatic in the field survey). Additional information collected
data for habitat suitability or SDM, with the goal of included the date and time of each observation.
identifying suitable sites and potential sources of viable For GBIF data, all duplicates and entries with
and healthy seeds for V. paradoxa. questionable geographic coordinates were removed.
The remaining points were spatially thinned at a 1 km
2. Materials and methods resolution using the “spThin” package in R to minimize
26
spatial or geographical bias commonly associated with
2.1. Study area large online databases. This resulted in a total of 100
The study area covered the Guinea savanna region of unique occurrence datasets. Given that V. paradoxa data
Nigeria, including Niger, Kwara, the Federal Capital collected in this study represent presence-only records,
Territory, and parts of Kebbi and Kogi states (Figure 1). pseudo-absence data (locations where the species is not
The vegetation consists primarily of deciduous trees known to be present) were required for most machine
and grasses. Rainfall is seasonal, occurring mostly learning algorithms in spatial modeling. To generate
from June to September, with total annual rainfall of this, between 500 and 2,000 random background
1,200 mm in the southern region and 700 – 900 mm in points were sampled, and 1,000 points were selected

A B C

D

Figure 1. Illustration of the use of Vitellaria paradoxa and the study locations. (A) Benefits of the species.
(B) Charcoal burning sites. (C) Land cover changes. (D) Study area showing local and state administration
boundaries.

Volume 22 Issue 4 (2025) 91 doi: 10.36922/AJWEP025210160