4. Discussion
The temporal variations in the species richness of propagules deposited by hydrochory show similarities with what has been observed in other rivers and what we propose in our hypotheses. These temporary changes are determined by the increase in the level and volume of water in the rivers during the rainy season (or higher water flow) which in turn increase the number of species that are dispersed towards the riparian zones (Boedtelje et al., 2004; Fraaije et al., 2017; Moggridge et al., 2009). In contrast, the abundance of propagules dispersed by hydrochory towards the riparian zone did not increase during the rainy season. Previous work has found that there is a greater number of propagules during the season of higher water flow in rivers due to the greater capacity of propagule dispersal through water currents (Boedtelje et al., 2004; Esper-Reyes et al., 2018; Fraaije et al., 2017; Moggridge & Gurnell, 2010; Naiman et al., 2010). However, there is no temporal trend in the deposition of propagules by hydrochory and it has been suggested that this may be due to the great production of propagules of one species (e.g. Veronica anagallis-aquatica ) throughout the year (Gurnell et al., 2008), a similar phenomenon could be occurring in the study area.
In our case, we observed that the dominant tree species in the riparian zone (Alnus acuminata ) (Flores-Galicia et al., 2021) had a peak in propagule release that was detected in the traps for other types of dispersal at the end of the rainy season (November 2017). Subsequently, the increase was reflected in the hydrochory traps at the beginning of the dry season (January 2018) (Annex 1). The effect of A. acuminata propagule production on temporal variations in the number of propagules deposited by hydrochory is confirmed if this species is eliminated from abundance analyses. By doing this, the number of propagules deposited by hydrochory in the rainy season is greater than in the dry season (rainy: \(\overset{\overline{}}{x}\) = 21.8, S.D. = 44.3, dry: \(\overset{\overline{}}{\text{x\ }}\)= 3.1, S.D.= 6.3,W = 502, d.f .= 45, p <0.01). The foregoing indicates that, in this system, the temporal patterns of abundance of propagules dispersed by hydrochory are not only determined by the temporal variations in the flow of water from the rivers, but also that the local production of propagules.
The differences in the composition of the propagules dispersed by hydrochory between the dry and rainy seasons were consistent with our hypothesis. These temporary changes in composition are related to differences throughout the year in the production of propagules of the species that make up the riparian communities or of the adjacent forests (Esper-Reyes et al., 2018), the accumulation of these in the riparian zones and their subsequent remobilization during river floods (Boedtelje et al., 2004; Fraaije et al., 2017; Gurnell et al., 2008). In our system we observed that some species with peaks of propagule production during the rainy season were only observed during the same season in propagules deposited by hydrochory, such as A. hickelli , Pinus sp.,Solanum sp and some Asteraceae (Asteraceae 2 y Asteraceae 6).
For the spatial dimension, there are contrasts between what our hypotheses propose and what we observed in this study, specifically for the richness patterns, the number of propagules and the composition of the communities of propagules dispersed by hydrochory along the river. Contrary to what was expected, a greater number of species dispersed by hydrochory was not observed with increasing distance from the origin site of the river. The absence of a positive relationship between the number of species and the distance to the origin of the river may be the result of the fact that downstream there was no accumulation of species from upstream, as has been observed in other rivers (Andersson et al., 2000; Andersson & Nilsson, 2002). The nesting analysis supports the above since the composition of the communities of the sites close to the origin of the river (sites one and two) is not a subset of the communities further away from the origin of the river. The low nesting indicates that the species near the origin of the river were not transported downstream (sites beyond 7 km) or if they did, the propagules were not deposited in the downstream riparian zones and continued their course in the column. of water beyond the study area.
Regarding the number of propagules deposited by hydrochory along the rivers, our results contrast with what we propose in our hypotheses and what has been observed in other studies (Andersson & Nilsson, 2002; Esper-Reyes et al., 2018). As occurs for the temporal variations in the number of propagules, the absence of an increase in the propagules deposited along the river could be related to the local production ofA. acuminata . However, when A. acuminata is eliminated from the analyses, an increase in the number of propagules is observed along the rivers in the hydrochory traps in the dry season (estimated= 0.15, z = 2.65, p = <0.01) but not in the rainy season, where instead a decrease in the deposited propagules is observed (estimated= -0.14, z = -2.46, p = 0.01). This same analysis (exclusion of A.acuminata ) in the traps for other types of dispersal yields similar results (rainy: estimated= -0.20, z = -3.04, p = <0.01; dry: estimated= 0.17, z = 2.42, p = 0.01). This could indicate that the production of propagules in the sites near the origin of the river is greater in the rainy season, which increases the deposition of propagules by hydrochory in the area, while in the dry season there is a greater production of propagules in the sites furthest from the origin of the river.
The results that we observed with respect to the spatial dimension (absence of relationship between the distance to the origin of the river with the estimated species richness or with the number of propagules, absence of nesting of the upstream communities in those of downstream), has important implications for theoretical frameworks such as RHC that hypothesize regarding longitudinal changes in diversity of riparian plant communities. The RHC postulate that the number of plant species increases to a maximum point in the intermediate zone of the rivers from which it decreases towards its mouth, these changes are associated with: (i) a linear increase in the disturbance in the riparian zones at along the river and (ii) a constant increase in the number of species and propagules dispersed by hydrochory (Nilsson et al., 1994, 2010). As there is no greater deposition of species and propagules deposited downstream, the patterns of diversity of the riparian plant communities that the RHC supposes could not be observed in rivers of mountainous systems (see for example Flores-Galicia et al., 2021; Pielech, 2021).
Some methodological limitations of this work did not allow us to conclude whether the contrasts between what we propose in our hypotheses with respect to the spatial distribution patterns of the propagules along the river (that is, a greater number of propagules and species deposited by hydrochory along the river, as well as nesting of the upstream communities in the downstream communities) and what we observed (absence of changes in the number of propagules, in the richness and decrease in the nesting of the communities). This is due to the fact that in this system: (i) the propagules transported in the water column travel a shorter distance than has been observed in other studies (Andersson et al., 2000; Boedeltje et al., 2003) or (ii) the propagules are effectively transported over long distances but due to the characteristic erosive capacity of mountain rivers (Wohl, 2010) the propagules deposited in the riparian zone are quickly remobilized to downstream and were not detected in our traps. Future studies could determine which of these hypotheses is correct through controlled propagule release experiments and the measurement of the distances they travel (Boland, 2017; da Cunha et al., 2017) or through sampling of propagules in the water column (Boedeltje et al., 2003).