Functional Changes and Threats to Hyperseasonal Neotropical Savannas After Australian Acacia Invasion

The hyperseasonal savanna experiences regular ooding and drought stresses and is a neotropical vegetation type threatened by global change including Acacia spp. invasion. To deepen the understanding of hyperseasonal savannas after Acacia invasion in a climate change scenario, we aimed to answer if: i) the plants of the studied hyperseasonal savanna are separated into C3, C4 or CAM species; ii) Acacia invasion can change the hyperseasonal savanna functioning for C3, C4 and CAM plants; iii) how invasive Acacia uptake water compared to native species in this hyperseasonal savanna. We detected both C3 and C4 metabolic groups of plants but two C3 species are possibly CAM facultative. The functioning of C3 plants as a group was not affected by the Acacia invasion, but this result does not exclude a species turnover between C3 herbs and C3 trees. The C4 plants of invaded Mussununga lost their response of increasing water use eciency to the increasing Leaf N%. Plants of hyperseasonal savannas depend on the same water source as the soil water from recent rains. There are differences in d 18 O among species because some grow mostly during the rainy season with the 18 O-enriched water meanwhile the invader Acacia mangium grows throughout the year whenever it rains. According to our results, the threat to C4 plants is high and they can be excluded from Mussunungas and from hyperseasonal savannas. However, hyperseasonal savannas are threatened as a vegetation. Therefore, hyperseasonal savannas should be considered critically endangered because of global change, especially bacause Acacia invasions. Initiatives for conservation of hyperseasonal savannas could save these remarkable ecosystems. variables; F, biplot of Leaf C% responding to δ15N. G-I: G, coecients of Leaf N% responding to responding to explaining variables; H, explaining percentages of Leaf N% responses by predictor variables; I, biplot of Leaf N% responding to C/N. coecients C/N responding to explaining variables; K, explaining percentages C/N responses by predictor variables; C/N responding to Leaf N%.


Introduction
Along with land use change and climate change, alien plant invasion are considered a major threat globally to biodiversity and ecosystem functioning (Strayer 2012;Ens et al. 2015). In most terrestrial ecosystems the climate is predominantly shifting to one with more heavy precipitation, more hot extremes and more consecutive dry days (IPCC 2021a, b). Consequently, the climate is changing towards more extreme seasonal climate in most of the tropical regions. In addition, biological invasion as one of the main factors of loss of biodiversity and ecosystem services, also contributes to global change  (Sarmiento 1984) which are expected to deepen with the changing climate. Besides climate change, a better understanding of species coexistence and functioning of hyperseasonal savannas in the context of Australian Acacia spp. invasion is crucial for its conservation.
The tropical hyperseasonal savanna is associated with at plains of poorly structured soil with layers of water-carried deposits that slow down drainage at the point that the vegetation is stressed by months of ooding during rainy season. Few months after the rainy season the soil becomes completely dry and the vegetation is stressed by months of drought. This extreme seasonality creates an open vegetation dominated by grasses and xeromorphic sedges with few woody plants or palms (Sarmiento and Monasterio 1975;Sarmiento 1984). Mussununga is a hyperseasonal sandy savanna which occurs in small areas under the Brazilian Atlantic Rainforest distribution with Spodosol soils evolved from Tertiary sandstones with a cementation layer of complexed water-carried deposits at a variable depth ( Alien plant invasion can determine irreversible biodiversity decline due to the disappearance of native species, and subsequent ecosystem instability through drastic changes in vegetation composition and function (Hooper et al. 2004(Hooper et al. , 2005. Successful alien plants are considered to possess a high resource-use e ciency (e.g. water, nutrients) with improved performance facilitated by high resource availability resulting from disturbance or low resource uptake by the native plant community (Funk and Vitousek 2007;Funk 2013 Considering the threat to hyperseasonal savannas in the context of global change and to deepen the understanding of the functioning of hyperseasonal savannas before and during Acacia invasion, we aimed to answer the following questions: i. Are the plants of the invaded hyperseasonal savanna separated into C3, C4 and CAM species?
ii. Can Acacia invasion change the hyperseasonal savanna functioning for C3, C4 and CAM plants?
iii. How does invasive Acacia uptake water compared to native species in this hyperseasonal savanna?

Vegetation sampling
The vegetation plots were set in three different Mussunungas divided into two series of plots arrangements: Marcetia centered plots and Acacia centered plots. Each circular plot with 6m of diameter was centered in different shrubs or treelets of Marcetia taxifolia or Acacia mangium. These two focal species were chosen because the former was abundant, non N-xer and easily found and the latter because is the N-xing invader. All plants of all life forms were sampled and most plots had one or few small termite mounds. The Acacia plants were less abundant despite usually larger than Marcetia plants. There were settled ve pairs of plots in each Mussununga ( Figure S1). At the sampling time, Acacia mangium was the only species of Acacia invaders, but Acacia dealbata and Acacia auriculiformis have been also reported in Mussunungas of the region.

Stable isotopes
Five leaves from each plant were collected during the rainy season (March 2013) and dried in an oven at 70-80 °C, excluding the petiole and evident veins. The dry material was ground in a ball mill to reduce it to particles of no more than 40 µm. For an elemental analysis of C and N, 5 µg of samples were weighed on a precision balance (XM 1000 P, Sartorius) and deposited in 5x9 tin capsules (EuroVector, Milano).
The C/N and the stable isotope ratio 13 C/ 12 C, 15

Results
In the studied hyperseasonal savannas, we detected the C3 and C4 metabolic groups of plants among the species according to the d 13 (Tables 2-4 and S1-S3, see Figure S2).
Leaf C% increases as C/N increases and as d 15 (Table S1).
The monocots Lagenocarpus rigidus and Actinocephalus ramosus as well as Blechnum serrulatum, a pteridophyte species, are 15 N-enriched with low N-nutritional status and with high d 13 C (Tables 2-4 and S1-S3, see Figure S2). Syngonanthus nitens and Panicum trinii are also C3 monocots that have high d 15 N, high d 13 C and low C/N (Tables 2-4 and S1-S3).
Only two species are C4, Urochloa sp and Cyperaceae sp1 with d 13 C values varying between -12.9 and -11.7 (Table S4). All other species have d 13 C between -42.7 and -24 and are C3 or could be CAM facultatives as is likely the case of Marcetia taxifolia and Actinocephalus ramosus (Table S5). The association between termite mounds and certain species with high d 13 C and high WUE in hyperseasonal savannas deserves attention because can shed light in an old controversy about the process in hyperseasonal savannas that originates Murundus, mounds attributed to termites or differential erosion, possibly with a facilitative in uence of termite mounds according to our results (see Marimon et al. 2015 Some C3 monocots as Lagenocarpus rigidus and Actinocephalus ramosus are 15 N-enriched possibly because they are commonly associated to termite mounds, dungs and urine but they have low Nnutritional status, especially Lagenocarpus rigidus that is amongst the species with the highest C/N.

Discussion
Possibly, their high d 13 C means high WUE that depends on 15 N-enriched N that is better mineralized and mostly available during pulses of N availability during rainy season for small C3 monocots. Blechnum serrulatum, a pteridophyte, is another species that bene ts from the 15 N-enriched in the same way of the small monocots Lagenocarpus rigidus and Actinocephalus ramosus.

Conclusions
Are the plants of the studied hyperseasonal savanna separated into C3, C4 or CAM species?
Most individuals and plant species in the studied Mussunungas are C3. Actinocephalus ramosus and Marcetia taxifolia, besides C3 could also be CAM facultative and deserve further attention as they can be indicative of hyperseasonal savannas. Two plant species are probably C4: some individuals of an unidenti ed species of Cyperaceae, a native sedge, and few individuals of Urochloa decumbens, an african grass species commonly used in pastures.
Can Acacia invasion change the hyperseasonal savanna functioning for C3, C4 and CAM plants?
Among C3 plants, the more N-limited the plants (higher C/N), the more dependent on 15 N-enriched N (possibly from termite mounds, dungs or urine) and the more dependent on nitrogen pulses. Thus, the higher d 15 N, the higher the d 13 C (the higher WUE), indistinctly if the Mussunungas are invaded or not.
Therefore, the responses of the C3 plants were not affected by the Acacia invasion, but this result does not exclude a composition shift caused by a turnover among C3 species from small monocots/herbs to C3 trees that compete for light as Acacia invasion out ows N into the ecosystem and shifts the hyperseasonal savanna ecosystem towards a dense woodland. to the C/N variation. If we consider that this effect can be associated with shading promoted by the Acacia invasion, a rapid exclusion of C4 plants in Mussunungas is conceivable. This effect could be general for hyperseasonal savannas, at least for the hyperseasonal savannas invaded by Acacia.
How do invasive Acacia uptake water compared to native species in this hyperseasonal savanna?
Mussununga plants depend on the same water source as the soil water from recent rains. This is because during the dry season the soil water is completely depleted. However, there are differences in d 18 O among species because some are annuals and geophytes that grow mostly during the rainy season with the 18 O-enriched water of heavy rains, sometimes with high d 18 O possibly boosted by CAM facultative metabolism, meanwhile other species grow throughout the year whenever it rains, such as the invader Acacia mangium. The perennial Marcetia taxifolia plants can also grow throughout the year, but its high d 18 O can also indicate that Marcetia taxifolia is CAM facultative. The high d 18 O of Marcetia taxifolia plants means that they use proportionally more water from heavy rains and/or they grow during shorter periods than Acacia mangium and for longer periods than annuals like Actinocephalus ramosus or geophytes like Blechnum serrulatum. This is a clear leverage for the biomass production of Acacia invaders, as they grow throughout the year every time it rains.

Concluding remarks
According to our results, the threat to C4 plants is high and this metabolic group of plants can be excluded from Mussunungas very soon and possibly can be excluded from hyperseasonal savannas in general as Acacia invasion advances. However, not only C4 plants are threatened by Acacia invasions with the formation of dense woodlands, but also the hyperseasonal savannas as a whole, since most of their plant species are shading intolerant. Therefore, hyperseasonal savannas should be considered critically endangered because of global change including Acacia invasions. An IUCN initiative (see IUCN 2015) could be an eye-opener for researchers to deepen the understanding of this remarkable type of ecosystem that potentially occurs worldwide in tropics but it is not even mentioned outside Neotropics.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarymaterial.docx