Fig 1.
Map of herbivore GPS points in Etosha National Park in northern Namibia.
GPS points of collected fecal samples from 11 herbivore species are represented by colored points throughout the park, with each color corresponding to a specific species. “BW” represents blue wildebeest (Connochaetes taurinus), “BZ” represents Burchell’s/plains zebras (Equus quagga burchellii), “ED” represents eland (Taurotragus oryx), “EL” represents African elephants (Loxodonta africana), “GB” represents gemsbok/oryx (Oryx gazella), “GI” represents Angolan giraffes (Giraffa camelopardalis angolensis), “HZ” represents Hartmann’s/mountain zebras (Equus zebra hartmannae), “IM” represents black-faced impala (Aepyceros melampus petersi), “KU” represents kudu (Tragelaphus strepsiceros), “RH” represents red hartebeest (Alcelaphus buselaphus), and “SB” represents springbok (Antidorcas marsupialis). Map image is the intellectual property of Esri and is used herein under license. Copyright © 2023 Esri and its licensors. All rights reserved.
Table 1.
Classification of African herbivore community in Etosha National Park, Namibia.
Fig 2.
Core bacterial phyla in Etosha National Park’s herbivore community.
Comparison of 15 major bacterial phyla (present at >1% abundance) to core phyla, defined as phyla present in ≥90%, ≥ 80%, ≥ 75%, and ≥50% of all samples within each herbivore species (n = 11) sampled in Etosha National Park, Namibia. Minor category includes all phyla present at <1%. Herbivore species identification is represented as “EL” for African elephants (Loxodonta africana), “BZ” for Burchell’s/plains zebra (Equus quagga burchellii), “HZ” for Hartmann’s/mountain zebra (Equus zebra hartmannae), “GI” for Angolan giraffes (Giraffa camelopardalis angolensis), “ED” for common eland (Taurotragus oryx), “KU” for kudu (Tragelaphus strepsiceros), “IM” for black-faced impala (Aepyceros melampus petersi), “SB” for springbok (Antidorcas marsupialis), “GB” for gemsbok/oryx (Oryx gazella), “BW” for blue wildebeest (Connochaetes taurinus), and “RH” for red hartebeest (Alcelaphus buselaphus). Herbivore species are listed according to phylogenetic relatedness, with the most evolutionarily separated species on the left and closely-related species on the right. There were two phyla (Bacteroidota and Firmicutes) present at the ≥ 90% core level and three phyla (Bacteroidota, Firmicutes, and Verrucomicrobiota) present at the ≥ 80%, ≥ 75%, and ≥50% core levels that were shared among all herbivore species. The teal-colored phylum present in the major phyla group and 50% core level in eland, giraffes, and impala is Cyanobacteria. The salmon-colored phylum present in the major phyla group and 50% core level in blue wildebeest, eland, kudu, and springbok is Spirochaetota.
Fig 3.
Significant differences in relative abundance of core bacteria based on environmental and physiological factors.
Relative abundance of core bacterial phylum (a) and genera (b-d) found in ≥50% of samples per herbivore species that exhibited significant differences associated with specific environmental or intrinsic factors. (a) Among three putatively divided zones of Etosha National Park (ENP) in northern Namibia, the phylum Verrucomicrobiota was significantly different between Zones 1 and 2 (bootstrapped one-way ANOVA P = 0.007) and Zones 1 and 3 (bootstrapped one-way ANOVA P = 0.04) across all 11 herbivore species. Herbivore species identification is represented as “EL” for African elephants (Loxodonta africana), “BZ” for Burchell’s/plains zebras (Equus quagga burchellii), “HZ” for Hartmann’s/mountain zebras (Equus zebra hartmannae), “GI” for Angolan giraffes (Giraffa camelopardalis angolensis), “ED” for eland (Taurotragus oryx), “KU” for kudu (Tragelaphus strepsiceros), “IM” for black-faced impala (Aepyceros melampus petersi), “SB” for springbok (Antidorcas marsupialis), “GB” for gemsbok/oryx (Oryx gazella), “BW” for blue wildebeest (Connochaetes taurinus), and “RH” for red hartebeest (Alcelaphus buselaphus). Herbivore species are listed according to phylogenetic relatedness, with the most evolutionarily separated species on the left and closely-related species on the right. (b) Monoglobus was significantly more abundant in nonruminants than ruminants (bootstrapped t-test, corrected P = 0.05). (c) Christensenellaceae_R-7 group was significantly more abundant in male versus female herbivores (bootstrap t-test, corrected P = 0.04). (d) P-251-O5 was significantly more abundant in female gemsbok than male gemsbok (permutation t-test, P = 0.05).
Fig 4.
Core bacterial genera in Etosha National Park’s herbivore community.
Comparison of four core genera levels, defined as genera that occurs ≥1% relative abundance and is present in ≥90%, ≥ 80%, ≥ 75%, and ≥50% of all samples within each herbivore species (n = 11) sampled in Etosha National Park, Namibia. Minor category includes all genera present at <1%. Herbivore species identification is ordered according to phylogenetic relatedness (distantly-related species on the far left and more closely-related species on the right) and is represented as “EL” for African elephants (Loxodonta africana), “BZ” for Burchell’s/plains zebras (Equus quagga burchellii), “HZ” for Hartmann’s/mountain zebras (Equus zebra hartmannae), “GI” for Angolan giraffes (Giraffa camelopardalis angolensis), “ED” for eland (Taurotragus oryx), “KU” for kudu (Tragelaphus strepsiceros), “IM” for black-faced impala (Aepyceros melampus petersi), “SB” for springbok (Antidorcas marsupialis), “GB” for gemsbok/oryx (Oryx gazella), “BW” for blue wildebeest (Connochaetes taurinus), and “RH” for red hartebeest (Alcelaphus buselaphus). There were five genera (Christensenellaceae_R-7 group, Coprostanoligenes, Lachnospiraceae_unclassified genera, Rikenellaceae_RC9, and UCG-010) present at the ≥ 90% level, seven genera (the previous five plus Prevotellaceae_UCG-004 and UCG-005) present at the ≥ 80% level, nine genera (the previous seven plus Bacteroides and Uncultured genera) present at the ≥ 75% level, and 16 genera (the previous nine plus Alistipes, Clostridia_UCG-014, Clostridia_VadinBB60, Monoglobus, NK4A214, RF39, and Roseburia) present at the ≥ 50% level that were shared among all herbivore species.
Fig 5.
Relative abundance of core bacterial genera fluctuates by geographic zone in Etosha National Park.
Relative abundance of four bacterial genera (Alistipes, Christensenellaceae_R-7 group, NK4A214, and RF39) among seven African bovid species within three putatively divided zones of Etosha National Park, Namibia. “BW” represents blue wildebeest (Connochaetes taurinus), “ED” represents common eland (Taurotragus oryx), “GB” represents gemsbok/oryx (Oryx gazella), “IM” represents black-faced impala (Aepyceros melampus petersi), “KU” represents kudu (Tragelaphus strepsiceros), “RH” represents red hartebeest (Alcelaphus buselaphus), and “SB” represents springbok (Antidorcas marsupialis). No eland samples were collected in Zone 2. Alistipes abundance was significantly different between Zones 1 and 2 (bootstrapped ANOVA, corrected P = 0.007), genera identified to the family Christensenellaceae_R-7 group were significantly different between Zones 1 and 3 (one-way ANOVA, corrected P = 0.04), NK4A214 abundance was significantly different between Zones 1 and 2 (bootstrapped ANOVA, P = 0.01), and RF39 abundance was significantly different between Zones 1 and 2 (bootstrapped ANOVA, P = 0.02) and Zones 2 and 3 (bootstrapped ANOVA, P = 0.003).
Fig 6.
Gut microbiome alpha diversity is significantly different by sex, host species, and gut morphology.
Boxplots comparing two gut microbiome alpha diversity metrics for 11 African herbivore species. Sampled species included African elephants (Loxodonta africana), Burchell’s/plains zebras (Equus quagga burchellii), Hartmann’s/mountain zebras (Equus zebra hartmannae), Angolan giraffes (Giraffa camelopardalis angolensis), common eland (Taurotragus oryx), kudu (Tragelaphus strepsiceros), black-faced impala (Aepyceros melampus petersi), springbok (Antidorcas marsupialis), gemsbok/oryx (Oryx gazella), blue wildebeest (Connochaetes taurinus), and red hartebeest (Alcelaphus buselaphus). a) Microbial species richness (SR) was significantly more abundant in female eland than males (permutation t-test, 2000 reps, corrected P = 0.03). b) Mean Bulla evenness was significantly greater in impala (0.552) than giraffes (0.509; one-way ANOVA, Tukey’s post-hoc corrected P = 0.02). c) Mean Bulla evenness was weakly but significantly greater in ruminants (0.531) than nonruminants (0.517; bootstrapped t-test, corrected P = 0.03).
Fig 7.
Unweighted UniFrac distance is positively correlated with geographic distance between elephant samples.
Unweighted UniFrac dissimilarity correlates weakly (Mantel statistic r = 0.07) but significantly (p = 0.05) to geographic distance (in km) between pairwise elephant (Loxodonta africana) samples (n = 33) in Etosha National Park, Namibia. Each point represents the pairwise difference between two elephant samples, thus the total number of points is 33x33.