Search

International audience; Perennial grasses are the main source of fuel during fires in savannas. The demography of these grasses likely varies between species although they have the same general architecture and coexist in savannas. However, very few studies compare their demography. Similarly, their demography likely varies between years because of the variability in weather condition and fire intensity. We described and compared the demography and life-cycle of the four dominant perennial grass species (Andropogon canaliculatus, Andropogon schirensis, Hyparrhenia diplandra and Loudetia simplex) of the Lamto savanna (Ivory Coast) and assessed the influence of their demographic features (stasis, fecundity, growth, fragmentation and retrogression) on this demography. Grass species were monitored over three consecutive years (2015-2016, 2016-2017 and 2017-2018) on three 5 m × 10 m plots. We used a size-classified matrix model with 5 circumference classes (3-10 cm, 10-20 cm, 20-35 cm, 35-50 cm and >50 cm). Results showed differences in the age-based parameters, asymptotic growth rates (λs) and elasticities of the λs of the grass species. The population of A. canaliculatus, A. schirensis and H. diplandra were slowly declining while L. simplex was significantly declining. There were noticeable year-to-year variations in the demography of these four species. The most important demographic parameter influencing λs was the stasis in all species, while retrogression and fragmentation contributed to a relative homogeneity of ages between size-classes 2 to 5. This study provides new insights about the demography of Guinean savanna grasses that could be used to describe the mechanisms of their coexistence, and inform fire management policies.

International audience; Biological nitrification inhibition (BNI) has already led to several studies mainly focused on underlying molecular mechanisms and applications to agriculture. We argue that it is also important to study BNI more systematically from the ecological and evolutionary points of view to understand its implications for plants and soil nitrifiers as well as its consequences for ecosystems. Therefore, we propose here a dedicated research agenda identifying the most critical research questions: (1) How is BNI distributed across plant phylogeny and why has it been selected? (2) What are the costs-to-benefits balance of producing BNI compounds and the relative impacts on BNI evolution? (3) Can we understand the evolutionary pressures leading to BNI and identify the environmental conditions favorable to BNI plants? (4) How has BNI coevolved with plant preference for ammonium vs. nitrate? (5) Diverse BNI compounds and various inhibition mechanisms have been described, but implications of this diversity are not understood. Does it allow inhibition of various groups of nitrifiers? (6) Does this diversity of BNI compounds increase the efficiency, spatial extension, and duration of BNI effect? (7) What are the impacts of BNI compounds on other soil functions? (8) Can field experiments, coupled to scanning of the diversity of BNI capabilities within plant communities, evaluate whether BNI influences plant-plant competition and plant coexistence? (9) Can field quantification of various nitrogen (N) fluxes assess whether BNI lead to more efficient N cycling with lower losses and hence increased primary production? (10) Can the impact of BNI on N budgets and climate (through its impact on N2O emissions and its indirect impact on carbon budget) be evaluated at the regional scale? We discuss why implementing this research program is crucial both for the sake of knowledge and to develop applications of BNI for agriculture.

International audience; Savannas are structured ecosystems characterized by a grass layer interspersed with trees. Trees strongly modify their local environment and favour nutrient accumulation under their canopies. Tree roots can also forage horizontally far beyond the canopy projection to increase nutrient uptake. In the Lamto savanna (Cote d'Ivoire), grasses are able to inhibit nitrification while trees stimulate it.Here, we used a two-patch model simulating nitrogen (N) dynamics in a humid savanna between an open patch (without tree) associated with a low nitrification rate and a patch of tree clump associated with a high nitrification rate. The model also includes horizontal N fluxes between these two patches corresponding to horizontal soil exploration by tree roots. We analysed the impact of spatial heterogeneity in nitrification and soil horizontal exploration on N budget and plant biomass.Despite high N losses under trees due to nitrification stimulation by trees, our results show that the ability of trees to explore horizontally the open allows them to uptake more nutrients in total. This leads to an asymmetric N flux from the open to tree clumps, which contributes to nutrient enrichment under tree clumps and thereby to tree growth.Although trees have the ability to horizontally explore the soil to accumulate nutrients under their canopy, increasing the surface occupied by tree clumps increases N losses per hectare of savanna due to the increased nitrification under trees and the subsequent increase in NO3- leaching.While perennial savanna grasses show a restricted horizontal soil exploration to control nutrient availability, our results predict that the extension of tree roots outside their canopy increases their nutrient acquisition in the Lamto savanna. This study is the first one emphasizing the influence of horizontal exploration of trees and tree cover on savanna N budget and functioning. Overall, the proportion of tree cover and horizontal soil exploration are important factors to consider in savannas characterized by spatial heterogeneity in N cycling created by trees and grasses. These factors appear critical to the functioning of West African humid savannas and should be investigated in other savanna types.A free Plain Language Summary can be found within the Supporting Information of this article.

International audience; Coexistence between trees and grasses in savannas is generally assumed to be due to a combination of partial niche separation for water acquisition and disturbances impacting the demography of trees and grasses. We propose a mechanism of coexistence solely based on the partitioning of the two dominant forms of mineral nitrogen (N), ammonium (NH4+) and nitrate (NO3−). We built a mean-field model taking into account the capacity of grasses and trees to alter nitrification fluxes as well as their relative preferences for NH4+ versus NO3−. Two models were studied and parameterized for the Lamto savanna (Côte d’Ivoire): In the first model, the nitrification only depends on the quantity of available NH4+, and in the second model the nitrification rate is also controlled by tree and grass biomass. Consistent with coexistence theories, our results show that taking these two forms of mineral N into account can allow coexistence when trees and grasses have contrasting preferences for NH4+ and NO3−. Moreover, coexistence is more likely to occur for intermediate nitrification rates. Assuming that grasses are able to inhibit nitrification while trees can stimulate it, as observed in the Lamto savanna, the most likely case of coexistence would be when grasses prefer NH4+ and trees NO3−. We propose that mineral N partitioning is a stabilizing coexistence mechanism that occurs in interaction with already described mechanisms based on disturbances by fire and herbivores. This mechanism is likely relevant in many N-limited African savannas with vegetation composition similar to the one at the Lamto site, but should be thoroughly tested through empirical studies and new models taking into account spatiotemporal heterogeneity in nitrification rates.

Questions Perennial tussock grasses represent the principal fuel source for savanna fires; however, basic information about the impact of fire on their demography remains scarce. Do dominant perennial grass species differ in their demographic parameters? What is the overall impact of tussock circumference and fire regimes on grass demographic parameters? Do grass species differ in the sensitivity of their demographic parameters to fire regimes and tussock circumference? Location The study site is located in the savanna of the Lamto reserve, in the center of Ivory Coast (6 degrees 9 ' to 6 degrees 18 ' N, 5 degrees 15 ' to 4 degrees 57 ' E). Methods Data on the nine dominant species of tussock grasses were sampled over one year under four fire treatments (early fire, mid-season fire, late fire and no fire). We studied the impact of these fire treatments and tussock circumference on five demographic parameters: mortality, fragmentation (the division of a tussock into several smaller ones), growth, retrogression (the decrease in tussock size) and fecundity. Results All demographic parameters varied among species and across the fire regimes. Late fire had the largest negative effect on all parameters except fecundity. Schizachyrium platyphyllum was the most disadvantaged species by fire in terms of mortality, retrogression and growth, whilst Andropogon ascinodis was the most prone to fragmentation. Hyparrhenia smithiana and Sorghastrum bipennatum produced the most seeds and Brachiaria brachylopha produced the most new tussocks. The circumference of perennial grasses impacted all demographic parameters. Small individuals had the highest mortality and growth rates, while large individuals showed the highest fragmentation and retrogression rates, and produced the most seeds. Conclusions This study showed that perennial grass species differed in their sensitivity to fire treatments. Nevertheless, the mid-season fire could be advised to managers for the sustainable management of the Lamto savanna.