Native palms and trees mediate drought impacts on dry neotropical pastures

Dry forests are the most threatened biome across global tropics (Blackie et al., 2014; Miles 2006). Massive dry forest conversion into mostly treeless pastures and crop fields has decreased biodiversity, altered vegetation-climate feedbacks biogeochemical cycles (Harrison & Bruna, 1999; MEA, 2005) diminished positive contribution trees can have on farm productivity livestock well-being (Harvey 2011). Nevertheless, in many tropical subtropical rural areas, local farmers usually retain some scattered within their grazing (Bernardi 2016; Manning Farmers actively select for provisioning shade, fodder, timber, firewood, construction material (wood thatch roofs), food also aesthehic reasons recognizing economic ecological value During droughts, facilitate grass biomass, especially (Hernández-Salmerón Holmgren, 2022), may be crucial pasture climate change adaptation (Altieri 2015). Trees ameliorate abiotic stress through several mechanisms that growth, improve quality (Jose Dollinger, 2019; Solorio 2017) enhance cattle welfare by providing forage allowing behavioural thermoregulation (Broom 2013). Indeed, considered keystone species systems (Manning Despite recognition of effects productivity, much debate remains levels tree cover desirable (Augustine 2011; Veldman 2015), underlying (Dohn 2013) specific 2011) environmental conditions (Rivest under which positive, negative or neutral rangelands. Our current understanding interactions between plants natural ecosystems guide our search generalizations agropastoral (e.g. Galindo 2017; Gargaglione Holmgren Scheffer, 2010). Plant facilitation results as net outcome an interplay both above- below-ground. Facilitative more common increases, but this general pattern differs depending functional traits interacting species. Several theoretical models meta-analysis indicate plant peaks at intermediate is outweighed competition extreme stressful very low resources (Guignabert 2020; 2012; Valladares 2016). based what been learnt facilitative interactions, we suspect understory grasses growth rangelands during seasons, responses vary among lost drought. There a strong need how neotropics make decisions to manage on-farm these impact productive functions pastures. A better existing modifies critical designing sustainable incorporate palms climate-resilient biodiversity conservation purposes. Here, assess dispersed native pasturelands southeast Mexico. We aim (a) composition pastures, (b) affect different precipitation (c) influence droughts recovery rates rainy seasons. assessed abundance conducted correlational study tree–grass types groups farms, municipality Tizimín, Yucatán, Mexico (Figure 1). The warm sub-humid (Köppen Geiger, 1930) with marked (November–May) wet seasons (June–October). Mean annual 1263 mm (dry season: 392 average, 214 2018 150 871 530 431 2019) mean temperature 26°C (min. 19°C, max. 42°C), (period 1969–2016; SMN, 2020). Interannual variation partly explained El Niño Southern Oscillation (Philander, 1983). region flat elevations ranging 10 40 m a.s.l. Soils classified leptosols (Estrada-Medina Thin layers unfertile topsoil bare rock sites, soil depth generally not exceeding 30 cm. about 75% state Yucatan (INEGI, 2017). However, they being rapidly converted rangelands, planted C4 exotic (Bonilla-Moheno Aide, Cattle often surrounded secondary sparse trees. selected 23 used 2.5 5 ha each, representing gradient total density (see Table S1 Supporting Information). These belonged six farms no irrigation, similar stocking (i.e. 0.5 unit ha−1) rotation system was present same three consecutive days every 2 weeks). All were established 15–20 years ago logging burning forests. They remained fire free since then. breeds mixed race (Bos indicus x Boss taurus) raised breeding stock. This type production input dependent seasonal rainfall (Harrington Tow, dominated Brachiaria brizantha, Cynodon plectostachyus Megathyrsus maximus (cover ≥70%). Permission obtained from landowners official licence permit numbers required. Although data collected animals involved; therefore, ethical approval Between February March 2018, characterized counting identifying all woody stems DBH ≥5 cm found along five parallel transects 4 × 50 m, separated least each other evenly distributed pasture. measured canopy points transect using Forestry Suppliers, Inc Spherical Convex Densiometer. Canopy breast height 15 3 transects) averaged estimate per Each individual palm identified level knowledge botanical collections (Universidad Autónoma de México). groups: Palms, N2-fixing (NF), non-N2-fixing (NNF) because strategies. Variation root 2013), organic matter carbon accumulation (Binkley, 2005), architecture could lead productivity. Palms superficial smaller canopies than monitored changes subset 12 farms. In pasture, abundant Under canopy, 1-m2 permanent plot paired open grassland one distance (varying size plot), plots facing north stem maximize desired shade effect S1). aimed solitary individuals > density, diameter, DBH. Individual periodically, Suplier® convex crown densitometer, sampling 2019. herbaceous layer 120 adjacent once month (30–35 days) (March, April, May) (July, August, September) plot, quadrant intersections 20 percent green tones, flexible), yellow brown crunchy), shrubs seedlings) soil. greenness response variable less influenced biomass is. Grass commonly characterize physiological conditionof grasslands Gartzia Yan 2019). center plot. At end season, weight) clipping, drying weighing (20 cm). did use exclusion cages comparing active present. registered microsite calculated difference peak season. Samples taken weeks, always plots, first below min later grassland. moisture top IMKO HD2 hand-held meter sensor integrated TDR electronics Trime®-PICO 32; recorded light irradiance air minute HOBO® pendant Data Logger UA-002-64. Throughout whole duration study, hourly relative humidity two dataloggers U23-001 permanently placed pole San Jose Xhoppel Finally, R package climatrends (de Sousa analysed carbon, phosphorus, calcium, potassium magnesium content subsample 60 plots. overstory group Palm: Sabal sp., tree: Acacia pennatula, Coccoloba spicata) sampled season when possible sample 100 g upper core small shovel too rocky. Soil samples stored sealed plastic bags, dried 48 h 55°C sieved (mesh-width mm). laboratory analyses Universidad Yucatán (Table S2). performance index (OCE = Performanceundertree – Performanceopen) ranges −100% 100%, 0 there its outside canopy. greenness, height, weight, To overall estimator monthly monitoring linear repeated measures, palms, trees) dry, wet) fixed factors, random factors; nested farm. included DBH) moisture, temperature, irradiance) covariates backward selection AIC criteria. community cover. calculate effect, multiplied OCE corresponding (COCE * abundancefunc group). (palms, (dry, factors had significant (different variable, S3) covariates. standardized absolute proxies resilience drought (Ingrisch Bahn, 2018). pre-disturbance time impact, divided expressed percentage 2019, state, I ((x̄greenness∙wet∙2018 x̄greenness∙dry∙2019)/x̄greenness∙wet∙2018) 100%). quantified condition after day RR=x̄green∙wet − x̄green∙dry/days recovery). number period, experienced transition transitions season; rates. impacted afterwards, factor. (R Core Team, 2020) lme4 (Bates 2014) perform models. Overstory ranged 0% 30% total, 167 belonging 35 45% (15% species), followed 33% (46% species) 22% (39% 2a). yapa S. mexicana (Guano), A. pennatula C. spicata. families Arecaceae (dominated genus Sabal), Fabaceae (well represented genera Acacia, Lysiloma Piscidia). grasses, varied direction intensity function type, S4). higher weight (Figures 3, Figure (p 0.34). Grasses greener 0.02) taller 2b). maintained consider compared < 0.01, 2c, S3). depended denser 40% 3a). 0.02, 2) S2) Only severe rainfall, observed As 0.001). clearly visible 0.01), Non-N2-fixing except driest equally 0.02). 2c). 0.34, regardless Woody rather 0.05) slightly tended increase N2 non-N2 fixing became 3c). With canopies, understory. N2- 3d). year, constant, dense 29% 42% 44% trees). retained whereas lose leaves ∆ −2% −15% −12% season). onset leaved out reflected stronger amelioration (Irradiance −54% −50% grassland). Irradiance dropped roughly 50% 3°C comparable reduced 0.01) reduce strongly S3 S5). Contrastingly, 0.01). reducing diverse nutrients trees, soils richer C P 0.01 p 0.001) grasslands, Ca, K Mg content. None differed grasslands. ameliorated 0.001, S6). growing 4A). 2019 (characterized drop 242 average), reduction lower 4B). subsequent recovered faster those 4C, Overall, slowed down severity After slowly palms. detailed functioning shows enhanced increased, determine relationships tree–pasture pastures; groups, depend maintains water scarce prolonged droughts. facilitated recruitment result combination below-ground while maintaining sufficiently high grasses. constant uniformly distribute radicular exudates maintain (Leite addition, diazotrophic bacteria associated potentially contribute (Reis 2000). general, lowest supply nitrogen strongest reductions leafing It plausible think trade-off transpiration demands nutrient levels) particularly show importance depends characteristics benefactor conditions, probably trade-offs resource conditions. Meta-analysis recent literature field experimental (Holmgren Stahlheber D'Antonio, (Caldeira Cooper tend non-linear irradiance. plays key role competitive effects. densities increasing seedling denser. Below amount increase. only optimal increases canopies. vegetation surveys semi-arid regions demonstrated directly habitat adapted and, indirectly, affecting outcomes (Soliveres Remaining do services farmers, including discussed manuscript allow keeping seed sources shady microsites where recruit. never abandoned regenerate, existence remaining recolonization (Maza-Villalobos McClanahan Wolfe, 1993). complex seasonally dependent. systems, deciduous semi-deciduous synchronized strategy escape reproductive success months (Eamus, 1999). Therefore, ameliorating it most. implications will plants. More drought-tolerant light-demanding benefit deeper tolerant able keep photosynthetic capacity temperatures availability (Chaves 2003). That why promoted (Edwards behaviour play important patterns interactions. hot periods, increased provided alleviate heat Both attract concentration grazing, ruminating resting explain proportion becomes extremely valuable periods cooler On contrary, slow greening responds progresses limiting. Further research needed uncover might recovery. Interestingly, although largest failed recover contrast, despite earlier. seemingly puzzling microclimate shallower understory, evaporation demands. therefore point towards opposite contributions resilience. While offer pockets Future experiments combined continual long-term direct indirect microclimate, levels, magnitude help influences (Hernández-Salmerón, 2021). preliminary observations suggest noted frequently higher, Animals looked (N2 idling activities, grazed seeking cool negatively trampling likely Scattered provide opportunities own. wide diversity conserved pool species; Hernandez-Ramirez Garcia-Mendez, carefully documented motivated conserve over widespread occurrence highly diversified uses (Araújo Lopes, 2012), landscapes widely overlooked. Attention focused benefits worldwide (Barfod anecdotally references (Calle contributes understand neotropical explaining demonstrating tropics. Whether perceive clear them threat obvious though, remain given counterbalance considering provision ecosystem such shelter animal 2013; Murgueitio 2011), even profit directly. Farmer's choices retaining maximizing yields reflect holistic resilience, adaptability transformability social-ecological livelihoods (Walker 2004). findings demonstrate designed tolerate fundamental Removing vulnerable previously buffered, compromising (Folke introduction dominance studies insects, example, turn bird mammal communities (Pompermaier integrating mitigate loss facilitating surrounding drought, variability weather extremes (Matocha management restoration known (Griscom Ashton, Considering rapid expansion farming throughout Wassenaar 2007), replacement lands continue advance future. fostering maintenance periods. become increasingly warming progresses. larger appreciation remnant develop contributing conservation. Transforming visions take provides evidence transformation build agricultural challenges Iván Raniero Hernández-Salmerón Milena conceived ideas, interpreted wrote manuscript. Hernández-Salmerón, Jannes van Hoeflaken Valentina Delconte data. Francisco Javier Solorio-Solórzano contributed collection interpretation. authors critically drafts gave final publication. thank insights work R. funded PhD grant CONACYT (Mexico). declare conflict interest. available via Zenodo Digital Repository https://doi.org/10.5281/zenodo.6824373 2023). S1. Schematic view design. Different geometric figures represent groups. effect. S2. Error bars standard error. S3. Difference variables Negative values S4. Positive Characteristics coordinates studied Yucatan, Palm characteristics, (n 117 due misidentification surveys. Results level. variables. S5. S6. factor Please note: publisher responsible functionality any supporting information supplied authors. Any queries (other missing content) should directed author article.