Introduction
Plant diversity and endemism are exceptionally high in the Eastern Arc Mountains of Tanzania and Kenya (Lovett, Reference Lovett1998; Burgess et al., Reference Burgess, Butynski, Cordeiro, Doggart, Fjeldså and Howell2007), which are part of the Eastern Afromontane Biodiversity Hotspot (Mittermeier et al., Reference Mittermeier, Gil, Hoffmann, Pilgrim, Brooks and Mittermeier2004). However, the forests in the Eastern Arc ranges are under severe threat from deforestation, habitat fragmentation and resource extraction, such that, together with other forests in this hotspot, they are considered one of the most threatened sites globally (Brooks et al., Reference Brooks, Mittermeier, Mittermeier, da Fonseca, Rylands and Konstant2002). Conserving these forests is therefore a priority. This is especially true of tropical African plant species, of which over a third are threatened by extinction (Stévart et al., Reference Stévart, Dauby, Lowry, Blach-Overgaard, Droissart and Harris2019).
Of the 13 mountain blocks that make up the Eastern Arc Mountains, the East Usambara Mountains in north-east Tanzania are amongst the best studied in terms of trees and other plants (Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989; Iversen, Reference Iversen1991; Burgess et al., Reference Burgess, Butynski, Cordeiro, Doggart, Fjeldså and Howell2007). The East Usambara Mountains harbour 16 endemic tree species (R. E. Gereau, unpubl. data, Reference Gereau, Beentje, Luke, Malombe, Mashimba and Ndangalasi2024) and > 20 near-endemic tree species (Iversen, Reference Iversen1991; Burgess et al., Reference Burgess, Butynski, Cordeiro, Doggart, Fjeldså and Howell2007). Despite the extensive botanical coverage (reviewed in Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989), new tree species (Cheek, Reference Cheek2002; Dawson & Gereau, Reference Dawson and Gereau2010; Gosline et al., Reference Gosline, Marshall and Larridon2019) and new subpopulations of restricted-range species (Dawson & Gereau, Reference Dawson and Gereau2010) continue to be discovered in the East Usambara Mountains. Here we document the distribution, relative abundance and conservation status of two globally threatened tree species that we have studied since their initial discovery in an unprotected forest remnant of these mountains.
Cola porphyrantha Brenan (family Malvaceae) is categorized as Critically Endangered on the IUCN Red List (Luke et al., Reference Luke, Musili, Barasa, Kalema and Mathenge2018) and Gigasiphon macrosiphon (Harms) Brenan (family Fabaceae) as Endangered (Gereau et al., Reference Gereau, Beentje, Luke, Malombe, Mashimba and Ndangalasi2023). Cola porphyrantha was first described from a type specimen collected in 1978 from the Shimba Hills, Kenya, with additional records coming from one other coastal forest in Kenya (Cheek, Reference Cheek, Beentje and Ghazanfar2007). There was also a collection made in the Shimba Hills in 1968. In Tanzania, the species was found in 2000 at an altitude of 950 m in an unprotected forest fragment in the East Usambara Mountains (Cheek, Reference Cheek, Beentje and Ghazanfar2007). Approximately five mature individuals were reported then, and the habitat was described as heavily affected by human activities. Gigasiphon macrosiphon is known from three coastal forest sites in Kenya and from four disjunct sites in Tanzania, in two Eastern Arc Mountain ranges (one site in the East Usambara Mountains and two in the Udzungwa Mountains) and one on the Rondo Plateau in Lindi Rural District (Luke & Verdcourt, Reference Luke and Verdcourt2004; Ngumbau et al., Reference Ngumbau, Luke, Nyange, Wanga, Watuma and Mbuni2020; Gereau et al., Reference Gereau, Beentje, Luke, Malombe, Mashimba and Ndangalasi2023). The earliest specimens of this species in the East Usambara Mountains were initially reported from Amani in 1906–1912 (Luke & Verdcourt, Reference Luke and Verdcourt2004). Although a cultivated collection was made from Amani in 1937 (Brenan, Reference Brenan, Milne-Redhead and Polhill1967), there were no further wild records of it from there until we found it near Amani in 2003. Currently, G. macrosiphon is known from the same forest fragment in the East Usambara Mountains as C. porphyrantha. Two mature trees of G. macrosiphon were thought to survive in this fragment in 2003.
As the subpopulations of both of these taxa have not been assessed for almost 2 decades in the East Usambara Mountains and given their globally threatened status, conservation assessments are imperative. To locate mature trees and any seedling and sapling recruitment, we surveyed the forest fragment that was known to harbour both species as well as neighbouring forest fragments of varying sizes. We also assessed threats (e.g. tree cutting, encroachment, direct use of tree species) in all sites that we surveyed, to help us identify conservation strategies that could promote the survival of both of these taxa.
Study area
The East Usambara Mountains rise from the coastal peneplain to 1,501 m (Platts et al., Reference Platts, Burgess, Gereau, Lovett, Marshall and McClean2011), with forest starting at c. 250 m and being continuous to 1,100–1,200 m in some places (Fig. 1). Mean rainfall is 2,000 mm per year and is seasonal, with long periods of rain during late March–May and short periods of rain during October–November; proximity to the Indian Ocean induces orographic rainfall in most months (Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989). The mean annual temperature is c. 24 °C.
(a) The East Usambara Mountains, Tanzania, showing the known locations of Cola porphyrantha and Gigasiphon macrosiphon (Table 2), and (b) the area of forest in the central and southern East Usambara Mountains. Amani Nature Reserve is the largest block of forest, extending from Mbomole south to Gereza and east to Mashewa, not including the enclave of Shebomeza. The polygon is the area surveyed for the two tree species. All forest fragments are unprotected and are managed by local tea estates. We have not provided precise localities of the plots or of the two species because of the potential for their exploitation. The Tanzania map was developed using Platts et al. (Reference Platts, Burgess, Gereau, Lovett, Marshall and McClean2011) for the Eastern Arc Mountain blocks, and the background data were from Stamen Design, under a CC BY 3.0 license, and OpenStreetMap, under an Open Data Commons Open Database License.
Loss of original forest cover in the East Usambara Mountains is estimated to be > 50% (Newmark, Reference Newmark1998), driven largely by deforestation and habitat fragmentation as a result of the establishment of sisal and tea estates during the colonial period, dating back to the late 1800s (Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989; Newmark, Reference Newmark2002). On the submontane plateau, at 800–1,100 m, extensive tea plantations dissect the forest in the central and southern regions of the East Usambara Mountains. Numerous forest fragments of varying sizes are surrounded by tea and, in some cases, subsistence cultivation (Fig. 1). Most of these fragments have no protected status, lying outside Amani Nature Reserve (8,380 ha) and Nilo Nature Reserve (6,025 ha) in the south and north, respectively. Comparable to trends in earlier decades (Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989; Newmark, Reference Newmark2002), resource extraction, such as timber harvesting, and encroachment for subsistence agriculture continue to threaten both unprotected and protected forests in the East Usambara Mountains (Tanzania Forest Conservation Group, 2017).
Methods
To evaluate the abundance and distribution of G. macrosiphon and C. porphyrantha in the East Usambara Mountains, we used 15 randomly placed 20 × 20 m vegetation plots to survey the isolated, 17 ha forest fragment that harbours both species (Fig. 1). Additionally, we surveyed 32 randomly placed vegetation plots, of the same size, in seven neighbouring forest fragments (Fig. 1), assuming that similar edaphic factors would increase the likelihood of the presence of the two species. The areas of the seven other adjacent forest fragments we surveyed are 1.3, 4.6, 36.1, 98.3, 114.7, 135.4 and 153.0 ha. To ensure adequate coverage of the forest fragments, we also walked transects through all of the sites, in various directions. Transects were of varying lengths depending on fragment size and shape, and the width of each transect was 5 m either side of the centre line. The vegetation surveys were primarily in May and June 2022 and in January, May and September–October 2023. We made observations of fruiting and flowering of both species and of leaf shedding of the deciduous G. macrosiphon on a monthly basis during January 2022–December 2023. We also made fruiting and flowering observations in our ad hoc visits to the site for other research purposes. These observations are from 2000 onwards.
We identified all trees, shrubs and seedlings in all 47 plots, and we measured the diameter at breast height and height of all individuals of the two target species (i.e. seedlings, saplings and mature trees); we recorded root collar diameter for all individuals < 2 m in height. We also recorded any anthropogenic impacts, such as tree cutting or harvesting, in these plots. We recorded the location of all individuals of the two species > 5 cm diameter at breast height, with a GPS, for future monitoring.
Results
We located C. porphyrantha and G. macrosiphon almost exclusively in Bulwa, the same forest fragment in which they were previously known. We recorded G. macrosiphon and C. porphyrantha in three and six plots, respectively, of the 15 plots in the forest fragment; both species were found together in a localized area on a steep and rocky slope. We found a total of 47 G. macrosiphon, mostly seedlings and small saplings, with several larger saplings and five mature trees (Fig. 2b). We found a total of 153 C. porphyrantha in four of the plots, with 18 mature trees and 38 large saplings, and abundant seedlings and small saplings (Fig. 2a). We recorded a further eight saplings in another two plots: four saplings were at a mean of 120 m to the south-east and four at a mean of 130 m to the north-west from the edges of the clump of mature trees. Although we traversed the entire forest fragment, we did not find G. macrosiphon and C. porphyrantha elsewhere.
Size class distribution of individuals of (a) Cola porphyrantha and (b) Gigasiphon macrosiphon in the East Usambara Mountains, Tanzania (Fig. 1). For C. porphyrantha, data were pooled from four 20 × 20 m plots, and for G. macrosiphon from three plots. Size classes are based on diameter at breast height for individuals > 1.5 m tall and root collar diameter for individuals below this height. Note the y-axis scales are different.
We surveyed a total 46.6 km of transects through the eight fragments (area in ha/distance in km: 1.3/0.6, 4.6/1.2, 17.0/5.3, 36.1/4.8, 98.3/10.6, 114.7/10.5, 135.4/8.1, 153.0/5.5), and surveyed additional transects on the upper parts of the Dodwe River (2.1 km) and along the Vungwe River at Kwezitu (1.6 km). In all of these surveys we found no C. porphyrantha and only a single fruiting tree of G. macrosiphon, in rocky terrain near Kwezitu in October 2023.
In the forest fragment that harboured both species, anthropogenic impacts appeared to be low in the five plots in which mature trees of the two species were found, with only two small saplings of C. porphyrantha having been cut. However, in the other 10 plots, which were on more gently sloping to flatter terrain, we observed cutting of saplings and poles in all of the plots, and larger trees had been felled in six of the plots, including the two plots where we found saplings of C. porphyrantha (all cutting and felling was of species other than G. macrosiphon and C. porphyrantha).
Flowering and fruiting of both species are variable across the various years of observation (Table 1). Leaf-shedding in G. macrosiphon appears to precede the hot, dry season (Table 1). Seed dispersal of C. porphyrantha by one of three resident primate species, the blue monkey Cercopithecus mitis, was observed opportunistically in 2001, 2021 and 2022. The flowers and fruits of C. porphyrantha are shown in Plate 1.
Flowering and fruiting records of Cola porphyrantha and Gigasiphon macrosiphon in the East Usambara Mountains, Tanzania (Fig. 1).
1 Peak months of flowering or fruiting.
2 Also observed fruiting in Kihansi Gorge in October 2005, leaf shedding started in November 2005.
A Cola porphyrantha (family Malvaceae) tree with almost ripe fruit, and the unisexual flowers, in the East Usambara Mountains, Tanzania. The fruits and flowers are cauliflorous. Photos: Fruiting tree and female flower: N. J. Cordeiro; male flower: H. J. Ndangalasi.
Discussion
Although we surveyed a total of 47 vegetation plots and 46.6 km of transects in the eight forest fragments, most of the C. porphyrantha and G. macrosiphon that we located were in the one forest fragment in which they were first recorded in the early 2000s. Only five mature individuals of C. porphyrantha were previously reported from this location (Cheek, Reference Cheek, Beentje and Ghazanfar2007), and two of G. macrosiphon (H.J. Ndangalasi & N.J. Cordeiro, unpubl. data, 2000–2003). However, we recorded 161 and 47 individuals, respectively, of C. porphyrantha and G. macrosiphon, of which 18 and five were of reproductive size (Table 2). In addition, we found a single mature G. macrosiphon near Kwezitu, just north of the intensively surveyed forest fragments.
Locations and status of subpopulations of C. porphyrantha and G. macrosiphon in Kenya and Tanzania.
1 The last two mature trees in Longo Mwagandi were killed by elephants (Luke et al., Reference Luke, Musili, Barasa, Kalema and Mathenge2018).
2 Luke et al. (Reference Luke, Musili, Barasa, Kalema and Mathenge2018) did not provide the number of mature trees but stated that the site is affected by limestone extraction.
3 In six sites in Rondo Nature Reserve, two 1 km transects were surveyed per site and the species was recorded in three sites, but the number of mature trees was not reported.
The forest fragment in which we found both tree species has a south-facing aspect on a steep slope, with an abundance of rocks and boulders. We explored similar habitat in all adjacent forest fragments and at comparable elevations, but we found no additional individuals of either species (Fig. 1). The one mature G. macrosiphon near Kwezitu is beside a river in an area with rocks and boulders. Given the threatened status of both species and that they are known only from an unprotected, small forest fragment outside Amani Nature Reserve and in unprotected forest near Kwezitu, conservation action is required.
Gigasiphon macrosiphon was first known to be present in the East Usambara Mountains from the three syntypes of its basionym, Bauhinia macrosiphon Harms (Reference Harms1915, p. 467). Specimen Braun 1033 was collected in 1906 from the ‘right bank of Dodwe River at Amani’, and Zimmermann 3088 was collected in 1910 ‘at Amani’. These two specimens were subsequently destroyed in the Berlin herbarium during World War II (Hiepko, Reference Hiepko1987). A third syntype, Grote 3763, was collected in 1912 from near Kihuhwi at 500 m; the specimen at Berlin was destroyed but a duplicate was preserved in the East African Herbarium. Despite these early records, G. macrosiphon has not been reported since 1912 from the East Usambara Mountains during the many subsequent wide-scale studies of trees there (Hamilton & Bensted-Smith, Reference Hamilton and Bensted-Smith1989; Frontier Tanzania, 1996a,b, 2001a,b, 2002a,b) or in two targeted searches for it in Amani Nature Reserve (Luke & Verdcourt, Reference Luke and Verdcourt2004) and Kihuhwi (A. Mndolwa, pers. comm., 2023). We also failed to relocate it in the Dodwe River area. Of the eight sites in which G. macrosiphon has been recorded in Kenya (Kaya Muhaka and Gongoni and Mrima Forest Reserves) and Tanzania (Amani Nature Reserve, Bulwa forest fragment, Udzungwa Mountains National Park, Kihansi Gorge and Rondo Nature Forest Reserve), the largest population is in Gongoni Forest Reserve, with 159 mature trees (Fig. 1, Table 2; Malombe et al., Reference Malombe, Kioko and Matano2015). The size class distribution of G. macrosiphon in Gongoni Forest Reserve (Malombe et al., Reference Malombe, Kioko and Matano2015) and the East Usambara Mountains (Fig. 2) is comparable, indicating similar recruitment and the ability to maintain a small subpopulation.
Cola porphyrantha was previously known at two sites in Kenya (Longo Mwagandi in Shimba Hills National Park and Pangani Rocks; Fig. 1, Table 2; Cheek, Reference Cheek, Beentje and Ghazanfar2007). When first discovered in 2000 at 950 m altitude in the East Usambara Mountains, this was the only known record in Tanzania, and was 500 m higher than the records in Kenya (Cheek, Reference Cheek, Beentje and Ghazanfar2007). Only five mature individuals were reported then, and the small forest fragment was described as heavily affected by human activities (Table 2; Cheek, Reference Cheek, Beentje and Ghazanfar2007). The 18 mature trees and high recruitment at different ontogenetic stages (Fig. 2) are suggestive of a subpopulation that is currently stable but small. The eight saplings found at c. 120 m from the main stand of mature trees are probably a result of seed dispersal by blue monkeys, although for the downslope plot, runoff from heavy rains could also have carried seeds from higher up.
Cola porphyrantha and G. macrosiphon are narrowly distributed in the East Usambara Mountains and are currently known to co-occur only in an unprotected forest fragment surrounded by tea plantations and subject to heavy anthropogenic disturbance because of proximity to a large village. Although the part of the fragment where these two species occur was little disturbed, probably because of the steep and rocky terrain, widespread cutting of saplings, poles and trees threatens this forest fragment. These species require protection at this site, by either the government or the village authorities, as does the unprotected forest near Kwezitu where we located a single G. macrosiphon. An additional concern, at least for G. macrosiphon, is that fungal attack, seed predation and seedling herbivory appear to impact early recruitment (Malombe et al., Reference Malombe, Kioko and Matano2015; H.J. Ndangalasi & N.J. Cordeiro, pers. obs., 2022–2024). As we have a basic understanding of the flowering and fruiting periods of these two species (Table 1), and also of seed germination and seedling growth rates (H.J. Ndangalasi & N.J. Cordeiro, unpubl. data, 2006–2024), we recommend in situ planting of seedlings in Amani Nature Reserve as well as in surrounding forest fragments, to safeguard both species in Tanzania. This process is currently being developed in collaboration with the local management authorities.
Author contributions
Both authors contributed equally to study design, fieldwork, data analysis and writing.
Acknowledgements
We thank the following people for their assistance with different aspects of this research: the late S. Baruti, and B. Covarrubias, T. Challange, A. Kajiru, H. Karata, M. Kijazi, B. Matunda, B. Mtui, A. Nyululu and P. Shirk. Logistical support was provided by the Department of Botany, University of Dar es Salaam, Amani Parish, Roosevelt University, The Field Museum, Amani Nature Reserve and East Usambara Tea Company. Permission to conduct research was provided by COSTECH and Amani Nature Reserve, and funding was provided by the National Geographic Society and the Mohammed bin Zayed Species Conservation Fund. This article was supported by Carnegie African Diaspora Fellowship Program awards to HJN and NJC in 2020–2021 and 2021–2022, and was improved by feedback from R. Gereau and an anonymous reviewer.
Conflicts of interest
None.
Ethical standards
This research abided by the Oryx guidelines on ethical standards.
Data availability
The data that support the findings of this study are available on request from the corresponding author (HJN). The data are not publicly available because of the threatened status of the two species, but they have been lodged with the IUCN Species Survival Commission Eastern African Plant Red List Authority.
Open access
