Biodiversity declines in the Anthropocene continue to increase compared to the presumed prehuman background rate, with profound effects on ecosystem functioning and services (Loss et al., 2015; Rosenberg et al., 2019). Tropical ecosystems are important biodiversity reservoirs compared to other biomes, but their integrity continues to be threatened by existential anthropogenic threats such as habitat loss, climate change, unregulated harvest, and other forms of human-caused mortality (Barlow et al., 2018a; Ceballos et al., 2015; Gardner et al., 2009; Pollock et al., 2022; Rosenberg et al., 2019). Increasing temperatures coupled with changing rainfall patterns in the tropics are expected to impact species’ distribution patterns and population dynamics (Freeman et al., 2018; Freeman and Class Freeman, 2014; Magurran et al., 2010; Toms et al., 2012), in addition to driving upslope range shifts of lowland tropical species across taxa (Freeman et al., 2018; Freeman and Class Freeman, 2014). There is strong theoretical and empirical evidence indicating that tropical biotas are more strongly affected by anthropogenic ecosystem changes than their temperate counterparts (Colwell et al., 2008; Sekercioglu et al., 2008). Due to the global importance of tropical forests as carbon sinks and biodiversity, mitigating anthropogenic impacts on ecosystems and conserving tropical biodiversity has become an increasingly urgent research priority (Barlow et al., 2018b; Pollock et al., 2022). Despite general consensus that loss of montane forest habitat can lead to a decline of these small and isolated populations that are already elevationally constrained (Guo et al., 2013; Lomolino, 2001; McCain, 2009), few studies exist along elevational gradients especially in tropical Africa (Kittelberger et al., 2021). Furthermore, few elevational studies in Africa have focused on species trends despite the continent having steep elevation gradients many of which are unprotected (Elsen et al., 2018; La Sorte et al., 2014; Sheldon, 2019).

Baseline data are urgently needed to document how tropical species along elevational gradients have responded and will respond to rising temperatures, deforestation, and other anthropogenic threats (Freeman and Beehler, 2018b). However, there is a scarcity of baseline data in the tropics on species’ distributions (Collen et al., 2008; van der Hoek et al., 2020). There are few published elevational baselines of raw data that include the number of individuals of each species detected (Freeman et al., 2018a; Freeman and Beehler, 2018b; Pagaduan and Afuang, 2012). Here, we used a 13-year (2011 to 2023) population study of understory birds along an elevation gradient of a tropical montane forest – the longest study of its kind in the Eastern Arc Mountains – to evaluate long-term population trends from an isolated but intact forest reserve in southern Kenya. We used the number of unique individuals captured as an index of abundance (Pollock et al., 2022) and modelled the populations of 18 out of the 56 resident bird species, with the goal of determining how their abundances had changed over the 13-year sampling period.

The trends in the number of species captured at each site over time in Mt. Kasigau elevational transect are shown in Figure 2 . We registered a total of 2149 bird captures constituting 1755 unique individuals from 61 species in 30 families over the 19 sampling sessions that took place in 9 years, i.e. 2011–2015, 2019 and 2021–2023 ( Figure 2 ; Supplementary Table S1 ). The total sampling time amounted to approximately 912 net hours within the 13-year period. Out of the 61 species registered in the study period, 18 species had at least 20 individual captures and were used in our models ( Figure 3 ).

Overall, the models indicated that 15 of the 18 species analyzed (83.33%) had stable local abundances between 2011 and 2023 ( Figure 4 ; Supplementary Table S1 ). The exceptions were the Green twinspot (Mandingoa nitidula, APC= −0.16, APC_95%CI = [−0.23, −0.09]; TPC= −0.90, TPC_95%CI= [−0.97, −0.69]), and the Taita White-eye (Zosterops silvanus, APC=−0.11, APC_95%CI=[−0.18, −0.04]; TPC=−0.78, TPC_95%CI= [−0.92,−0.40]), whose declines were statistically significant, and the Eastern Nicator (Nicator gularis, APC= 0.07, APC_95%CI= [0.02, 0.14]; TPC=1.54, TPC_95%CI=[0.23, 4.23]) whose increase was statistically significant.

When we compared the estimated increments in local abundance obtained through parametric bootstrapping (Δn) between 2011 and 2023 ( Figure 5 ; Supplementary Table S2 ), our results differed slightly. In this case, besides the Green twinspot (Δn= −12.31; Δn_95%CI= [−19.81, −7.65]) and the Taita White-eye (Δn=−2.89; Δn_95%CI= [−7.53, 0.4]), the Red-throated Twinspot (Hypargos niveoguttatus) also showed a decreasing trend in this analysis (Δn= −9.61; Δn_95%CI= [−2268.93, −1.74]. According to theparametric bootstrap on the local abundance increment, four species showed a local abundance increase. As in the previous analysis, Eastern Nicator (Nicator gularis, Δn=4.77; Δn_95%CI= [0.92. 8.1]) was one of them. The other three were White-starred Robin (Pogonocichla stellata, Δn= 3.26; Δn_95%CI= [1.31, 5.04]), Bleating Camaroptera (Camaroptera brachyura, Δn= 3.26; Δn_95%CI= [1.19, 5.13]), and Bearded Scrub-robin (Cercotrichas quadrivirgata, Δn= 2.54; Δn_95%CI= [0.27, 4.79]).

Our results show a stable population trend for most studied bird species in Mt. Kasigau, despite inter-seasonal and inter-year fluctuations ( Figure 5 ). These results are in line with those in Manu National Park (Peru), where bird populations and rainfall were found to remain reasonably stable in a 40-year period (Martínez et al., 2023). However, our models are in contrast with those of other recent studies on montane bird population trends in tropical and Paleartic region that report generalized declines over the last decades (Blake and Loiselle, 2024; Brown et al., 2019; Lehikoinen et al., 2014; Neate-Clegg et al., 2021; Pollock et al., 2022; Riegert et al., 2021; Zamora and Barea-Azcón, 2015). Many tropical mountains have experienced changes in their climatic conditions and vegetation structure, implying changes in species niches that are likely to be driving the population fluctuations (de la Fuente et al., 2023; Dulle et al., 2016; Riegert et al., 2021). In contrast, the vegetation cover in Mt. Kasigau has remained largely stable over the study period, as have the temperatures at the sampling sites and the rainfall at the county level (Nyambariga et al., 2023). This general intactness of the montane forest habitat in Mt. Kasigau, which is largely devoid of human activities besides small scale firewood collection in the lower altitudes, livestock grazing in dry conditions and ecotourism, may explain the largely stable bird populations. Further, tropical climates are assumed to have constant environmental conditions which lead to constancy of resources and, hence, more stable populations especially in areas unaffected by significant human activities (Blake and Loiselle, 2015a; Sigel et al., 2006).

Despite this apparent stability, the declining trend for a globally threatened endemic species Taita white-eye, Zosterops silvanus, alongside Green twinspot Mandingoa nitidula ( Figures 4 , 5 ) may be indicative of changing habitat parameters or species-specific life history factors that do not favour these species. These declines are not likely to be caused by habitat changes as there has been minimal human disturbance in our study area during the course of the study. Besides natural tree-falls and isolated cases of selective removal of high-quality timber trees, small scale firewood collection and ecotourism are unlikely to have caused these declines. An increase in ecotourism may lead to negative impacts on birds due to behavioural changes caused by noise (Canaday, 1996), but there has not been a substantial increase in this activity during the study to affect bird populations. This suggests that other factors besides habitat changes may be driving these declines and may include but are not limited to disease and/or climate change. There have been earlier sentiments regarding the declining Taita White-eye when the species was suspected to have experienced a population crash in Mt. Kasigau (BirdLife International, 2022). The decrease in the Taita White-eye’s abundance is particularly concerning since Mt. Kasigau was identified as the species’ stronghold in the late 1990s, estimated to harbour 78% of the world’s population of this threatened species (Mulwa et al., 2007). A potential cause contributing to the Taita White-eye’s abundance declines could be inbreeding, which can cost lifetime fitness in birds (Harrisson et al., 2019). Supporting this hypothesis is that the gene flow between Mt. Kasigau and Taita Hills populations is scarce owing to isolation of the former, and the percentage of heterozygosity is relatively low (Habel et al., 2014). Inbreeding has been shown to have a lifetime fitness cost in birds (Harrisson et al., 2019). Restricted dispersal and frequent inbreeding within “sky island” systems, such as Mt. Kasigau for the Taita White-eye, can occur even in highly mobile bird species (Ceresa et al., 2024). In contrast, the Eastern Nicator showed remarkable increases in its abundance in the same period. Further studies are needed to understand the underlying causes of these population changes despite relatively stable rainfall and temperature conditions across the study period. Moreover, other species in the same diet guilds similar to these species have not experienced similar population changes.

Despite overall stability in bird populations, our analyses show short term dips and peaks in bird population trends. These dips may be attributable to the number of sampling sessions undertaken: there was only one sampling session undertaken in 2011, 2015 and 2019 while there were 2–4 sessions in all other years. However, periodic dips and peaks in bird abundances have been observed in other studies and have been hypothesised to be the effect of large-scale climatic cycles, such as El Niño-Southern Oscillation (ENSO) which can affect bird populations in different ways across geographical regions (Ballard et al., 2003; Blake and Loiselle, 2015a; LaManna et al., 2012). These ENSO events may not only influence the breeding success on birds both positively and negatively but could also affect their foraging behaviour and diet. In our study area, periodic changes in weather parameters were suggested as driving these periodic declines in an earlier study (see Wambugu et al., 2024). Data from the Kenya Meteorological Department (Kenya Meteorological Department, n.d) reveals that both 2011 and 2015 corresponded to La Niña and El Niño extreme weather periods, respectively, which may partly explain these dips in bird population trends in our study. It is however worth noting that many of the species in our study were only captured in one or two years at each site. Thus, our dataset only provides an indication of the general trend in bird community over time but it’s unclear regarding the turnover between species.

Several species were consistently present in the first half of the study (2011–2015) but absent in the second half (notably Common bulbul and Lesser Honeyguide in Site I; Spotted Flycatcher in Site II; and African Pygmy Kingfisher and African Goshawk in Site III). Likewise, other species appeared in the second half such as the Red-throated Twinspot and the Variable Sunbird in Site I. Other studies have termed similar observations as winner–loser species replacements, which may be triggered by changes in habitat parameters (Lees and Peres, 2006; Tabarelli et al., 2012). Winner–loser patterns are clearer in human–modified landscapes due to widespread habitat changes (e.g. Filgueiras et al., 2021) but less so in more intact habitats. In our study, these changes appear to have occurred immediately after population collapse, thought to be due to changes in resource availability as a result of unconfirmed events ( Figure 2 ).

Our study shows that most bird species populations in Mt. Kasigau remained stable despite reported declines in similar locations elsewhere in the tropics. This stability indicates the role of montane areas as refugia for birds in the face of a warming climate, along with other anthropogenic pressures. Our study further emphasizes that relatively small mountain regions can play an important role in maintaining stable bird populations as global environmental transformation continues to escalate. However, future studies should focus on species-specific life history aspects to better understand population trends of Afromontane birds, especially those that appear to be undergoing declines.