Drought effects on vessel plasticity of two endemic Magnolia species in the tropical 1 montane cloud forests of eastern Mexico

The distribution of Mexican Magnolia species´ occur under restricted climatic conditions. As many other tree species from the tropical montane cloud forests (TMCF), Magnolia species appear to be sensitive to drought. Through the use of dendrochronological techniques, this study aims to determine the climate influence on the vessel traits of M. vovidesii and M. schiedeana which are endangered tree species that are endemic to the Sierra Madre Oriental in eastern Mexico. Because most of the tree species in TMCFs are sensitive to climate fluctuations, it is necessary to investigate the differences in the climatic adaptability of the vessel architecture of these trees. This could allow us to further understand the potential peril of climate change on TMCFs. We compared vessel frequency, length and diameter in drought and non–drought years in two Mexican Magnolia species. We used tree–rings width and vessel traits to assess the drought effects on Magnolias’ diffuse–porous wood back to the year 1929. We obtained independent chronologies for M. vovidesii with a span of 75 years (1941–2016), while for M. schiedeana we obtained a span of 319 years (1697–2016). We found that temperature and precipitation are strongly associated with differences in tree–ring width (TRW) between drought and non–drought years. Our results showed anatomical differences in vessel trait response between these two Magnolia species to climatic variation. We suggest that our approach of combining dendroclimatic and anatomical techniques is a powerful tool to analyse anatomic wood plasticity to climatic variation in Magnolia species.


49
Tropical montane cloud forests (TMCFs) represent an example of forest ecosystems 50 where direct anthropogenic disturbance and climate change quickly affect the feedback 51 between forests and the hydrological cycle [1,2]. These ecosystems are influenced by 52 climate change mainly through the carbon cycle [3,4] and play an important role in the 53 regulation of atmospheric flows, humidity and rainfall recycling, which directly influences 54 local, regional and continental climates [5][6][7]. Climate fluctuations show that worldwide 55 TMCFs have experienced a warming trend and an increase in elevation of the 0°C isotherm 56 during the second half of the 20 th century [1,8]. 57 Dendroclimatology is a useful tool that provides past climatic information about the 58 development and ecology of tree species [9][10][11][12]. Several studies have confirmed the 59 relationship between climatic seasonality and seasonal growth rings development as a 60 response from tree species. Such a relationship that can be observed in tropical ecosystems 61 with the use of dendrochronology [11,13,14]. Climate signals such as precipitation (e.g., 62 rain, mist, fog and cloud water), temperature and/or droughts regulate the growth of 63 TMCFs' tree species [15,16]. Dendroecology allows us to identify climatic processes 64 across time and can be used to reconstruct past local and regional climates [9,16,17]. In this 65 context, one of the most important advances in dendrochronological studies has been the 66 additional focus on anatomical features such as vessel traits and/or tree-ring anatomy 67 [13,18,19]. For TMCFs' trees, water conduction through the xylem vessels is relevant to 68 understand the effects of stress caused by climatic changes such as those provided by 69 drought years. In particular, plasticity adjustments of the vessel dimensions could be 70 closely related to temperature and precipitation [20,21]. Most of the tree species inhabiting 71 the TMCFs develop reliable and anatomically characteristic annual growth rings On the other hand, M. schiedeana belongs to sect. Magnolia. It is mostly an evergreen 96 and short-to long-lived woody species (e.g., longevity of M. grandiflora of 318 ± 20 97 years), it commonly grows under the canopy of Quercus, Liquidambar,Meliosma,Fagus 98 and Podocarpus [27][28][29], and is often associated with Ternstroemia, Oreopanax and 99 Styrax. It also occurs in secondary conditions with Alnus and Clethra. This species has a 100 greater morphological variability and a wider distribution, occurring in Hidalgo, Puebla and 101 Veracruz [71,74], ranging from the eastern United States to Mexico 197 1929-30, 1940, 1963, 1970, 1972, 1976, 1983, 1991, 1997, 2012 and 2015-16). We also 198 followed this procedure for two consecutive years before and two years after El Niño only the objects larger than 10, 000 µm 2 , and that had a width smaller than twice their 209 length. Vessel outlines were improved by applying morphological operations (erode-dilate 210 2 × 2 one pass, and calculation of the convex hull).

211
We performed a multiple comparison Tukey test to assess whether the values of vessel 212 traits (frequency, length, and diameter) present a significant difference between drought 213 years (DY) and non-drought years (NDY) for the Magnolia species studied. These analyses 214 were performed in R (version 3.5.1) using the R-package ggplot2 (https://cran.r-215 project.org/web/packages/ggplot2/ggplot2.pdf [57]).

221
The chronologies index also showed a distinct inter-annual variation pattern for the 222 Magnolia species studied, except for those tree-rings developed in strong drought years 223 such as 1929-30, 1940, 1963, 1970, 1972, 1976(Figure 2), 1983, 1991, 1997, 2012  Batda the mean TRW range was 0.456 and 3.02 mm for M. vovidesii (Table 1). Signal-to-227 noise ratio (SNR) and the mean between-trees correlation was high in the Magnolia species 228 studied, suggesting a strong common signal is expressed as climate effect on growth rates 229 ( Correlations with mean maximum temperature (Tmax) were positive in previous Jan (-1) 236 (see Figure 3) (i.e. late Winter and early Spring (January to February)) and during the dry 237 cool season (Oct (-1) to Nov (-1)) before tree-ring development. A negative correlation 238 occurred in previous Sep (-1), Dec (-1) for radial growth chronology, and current late  TRWs' growth rates allowing the individual trees to adapt to these climatic fluctuations.

320
Opposed to this, from June to -September precipitation showed the opposite effect ( Figure   321 3), precipitation had a positive impact on TRW's growth rates in both species having an 322 even greater benefit on growth rates of M. vovidesii.

323
Our analyses revealed that a great variability in vessel's anatomy of diffuse-porous 324 wood from the two Magnolia species studied is related to temperature and/or water 325 availability. These anatomical adaptations result from a strong reduction in vessel 326 frequency, length and diameter during DY compared to NDY, plastic adaptations that play 327 an important role in water transport and safety [67]. We state that wood plasticity is essential to identifying climate adaptability of trees in TMCFs. Our findings support that M. 329 schiedeana is probably more sensitive to drought events than M. vovidesii due to poorer 330 vessel structure when confronted against these types of climatic events (Figure 4). Our 331 wood core digital images results demonstrate that most vessel traits show (i) high plasticity 332 from DY to NDY, and (ii) that there are differences in the diffuse-porous wood anatomy 333 and vessel plasticity of the different Magnolia species inhabiting the TMCFs (Figure 4). 334 We deduce that drought induces increased hydraulic conductance with the consequence of 335 high construction costs but reduced hydric transport efficiency and lower TRW growth