Dendrochronology in Latin America, state of the art and contributions for understanding climate variation

: Dendroclimatology has gained relevance during the XXI century. We analyze the state of-the-art of dendroclimatology in Latin America during the past 28 years (1990 to 2019), identifying the current state and recent advances in the application of dendroclimatology in this region. We carried out a systematic review in ScienceDirect, Web of Science, and Scopus databases using Boolean operators to logically “climatic variability,” and “climatic variability trend”, for each country. Dendroclimatological research conducted in the region focused on climatic reconstruction and the evaluation of new tree species with dendrochronological potential, which increase in 2010, then there was a gap between 1995 and 1996, later increase to present. Dendroclimatological studies in Latin America have been mainly developed in temperate climate zones (82.4 %) followed by tropical or subtropical areas (17.6 %). Dendroclimatological research in Latin America has provided advances in the study of climate variability by defining response functions of tree rings to climate. The generated information allows for a better understanding of the spatial and temporal dynamics of climatic variability and about its effects on ecosystems and society. We also call for increased dendroclimatological research based on subtropical and tropical forests limited studied until today. of

Latin America" and "dendroclimatology and Latin America." The resulting database was filtered by the abovementioned criteria focused on dendroclimatology in order to depurate the list of selected papers.
To systematize the information, the selected articles resulting from queries in the databases were organized according to the assessment of their content (Table 1). During the systematization and analysis stage, a standardized database was created and divided into five main sections: 1) general data, 2) study area, 3) research objectives, 4) methodological developments, and 5) the relationship between climatic variability and dendrochronology. Table 1. General description of the data required for the evaluation of the content of papers regarding the study of dendroclimatology

Concept
Variable Description

General information Who and When?
Author Who wrote the paper? Year In which year was it published?
Title Which is the title of the paper?

Where?
Journal name In which journal was it published? compared each database to identify duplicate articles and articles found in two or more of these datasets. The articles found in two or more datasets were purged leaving a single version of the repeated titles. Finally, a list was made considering the articles that were not duplicated in the databases and a single version of the articles that were repeated in two or more databases.
To observe the annual trend of publications, the nonparametric Mann-Kendall trend test was applied, which provided a qualitative observation parameter to determine a potential trend [18] in the number of annual publications.

Results
The three consulted databases produced similar outputs for searching articles. For instance, in ScienceDirect, Web of Science and Scopus, the query based on the terms "dendrochronology" produced 2,201, 2,189, and 2,289 articles, respectively. The search using the keyword "dendroclimatology" produced 542, 785 and 653 papers in ScienceDirect, Web of Science, and Scopus, respectively. Although using a combing of keywords (Dendrochronology and Trend, Dendroclimatology and Trend,

Dendrochronology and Climatic Variability, Dendroclimatology and Climatic
Variability) always produced a higher number of articles in ScienceDirect in comparison to Web of Science and Scopus databases (Figure 1), while, the last two databases produced a similar number of articles using the same combination of keywords.

Which is the publication trend and where is published the dendroclimatic research in
Latin America?
The number of publications about dendroclimatology in Latin America has increased during the study period. Such a rise displayed two patterns: one with fewer    (12), Holland (10) and Chile (10), rest of the editorial offices by country published less than seven articles ( Figure 3). We considered journals as international when they had offices or editorial representation in different countries.

Which are the study scales in dendroclimatological research?
Of the 220 papers, 102 (46%) studies were developed at the local level and 118 (54%) at the regional level. Of the published research having local scale goals, 57 papers indicated the effects on tree growth of climatic variability, 33 referenced the interaction of dendroclimatology with other sciences, and 12 referred to the reconstruction of past climate. Among the 118 published studies performed with a regional scope, 20 mentioned contributions of dendroclimatology as an auxiliary science for specific objectives, 53 referred to the analysis of climatic variability and tree growth, and 45 had the objective of reconstructing climate. Regional studies are related to the study of atmospheric phenomena or climatic variability while local studies are associated with the glacial thaw, searching for new dendroclimatic species for analysis and dendroclimatic response at different topographic positions.

Where dendroclimatological research is generated?
Among countries that develop dendroclimatological studies, Mexico (84), Argentina (57), and Chile (33) were more active in publishing, followed by Brazil (28)  Other spatial information included in this work was the extension of the studied areas, but only three papers (Costa Rica, Mexico, and Brazil) explicitly provided that information. The most studied biome for dendroclimatological studies is the temperate forest; semidesert biome is in second place, and the rain forest occupied the third place.
The temperate forest is one of the most studied ecosystems given that the species there produce reliable annual rings sensitive to climatic factors.

Sampling and species collected
Dendroclimatic studies are usually carried out by analyzing several types of wood samples. Out of the total published scientific studies, 44.0% used increment-cores, 4.0% increment-cores and cross sections, 14.0% cross sections, 3.0% used available tree-ring series from international databases, and 35.0% did not mention the sampling type used.
The highest sample size in dendrochronological studies consisted of 1477 increment-cores, while others had 1159, 1082 and 916 core samples [19][20][21]. Furthermore, the lowest number of samples was 4 cross-sections [22], with the average being 175 samples per study ( Figure 6). Nevertheless, most of the papers omitted to mention the size of the sampled area, therefore, it became difficult to extrapolate the number of potential for climate sensitivity, mostly in tropical, subtropical, and semiarid environments [23,24]. Conifers are more used for this kind of studies, because they produce well defined annual rings, their species diversity, wide distribution range, and longevity. The common climatic response detected throughout larger areas allows their integration into studies of regional climate variability and the analysis of the influence of ocean-atmospheric circulation modes [25][26][27].
In terms of highly sensitive species for dendroclimatic studies, the genus Pseudotsuga and Austrocedrus are preferred. The latter genus is characterized by producing anatomically well-defined rings and by being sensitive to changes in humidity [20,28,29]. The genus Pseudotsuga is a very climatically sensitive species and mostly responds to the seasonal winter-spring precipitation in northern Mexico and Southwestern United States. Its well-defined annual rings and wide distribution in one of the most extensive arid regions of North America have also been used for their suitability in dendrochronological studies [30][31][32][33][34].
Overall, conifers have been included in 64.0% of the dendrochronological papers growth rates and forest biomass production to establish growth models for forecasting commercial biomass volumes in native forests [37]. Additional dendroclimatological studies deal with the use of dendrochronology in climate warming by analyzing the decrease in precipitation and the rise in temperatures regarding glaciers' stability [20].
The use of dendrochronology to analyze patterns of insect outbreaks in years with droughts and higher temperatures has been an important issue in the last decades [38][39][40].
We identified 16 papers in which dendroclimatic analysis was applied to determine the historical recurrence of forests fires, to assess the evapotranspiration rates from tree species, and to estimate water balances [41][42][43]. The nature of the latter analyses, The reconstruction of past climates is the second most important topic with 57 publications; however, this is difficult to analyze because of the stringency of the tests applied to the data for observing the signal of climatic sensitivity, the low sample depth, and the short-time period covered by dendrochronological series, all of which increase the percentage of error [6]. The climatic signal makes it easier to observe climatic effects and to correlate annual radial increase and forest growth with meteorological variables, as well as to apply techniques like data cross-referencing, which allows determining patterns and their spatial distribution [54]. broadened. In these papers, it is characteristic that several species or even several families were included, mostly using a minimum of 2 species and in one case over 50 species of broadleaf trees from dry climates. Mostly among species from arid and subtropical climates, such as Prosopis ferox, Prosopis flexuosa and a wide variety of other broadleaf tree species, they were frequently studied before 2010, but the trend stabilized thereafter.

How is climatic variability related to dendroclimatology?
In 69.0 % of the papers, the methodological approach was based on establishing a significant association between climatological data and dendrochronological time series.
This technique can be considered a link between dendrochronology and climatology, such that this is the most frequently applied technique in this field of knowledge. The Pearson correlation or a multiple correlation analysis, allowing the production of a reconstruction model that must be calibrated and verified before being used for reconstruction purposes [32,55,56]. One of the most important objectives of dendroclimatological studies is to integrate a dendrochronological network of tree-ring series to analyze the climatic response at the regional or national scale and to determine the significant influence of atmospheric circulatory phenomena. In that regard, the Mexican Drought Atlas (MXDA) provides an estimation of dominant drought conditions based on the Palmer Drought Severity Index for June through August between 1400 and 2012 [57,58].

Discussion
Our search of published scientific information about the topics of dendrochronology and dendroclimatology showed differences between the higher number of papers in ScienceDirect relative to those in the Scopus and Web of Science databases in searches of the key words "dendrochronology and trend," "dendroclimatology and trend," and "dendrochronology and climatic variability." Such differences may have been due to the number of journals subscribed to each database, their geographic areas of influence, and the degree of specialization regarding our topics of interest. The relationship between key words and the application of Boolean operators allowed us to find research spaces within dendroclimatology related to the topic "trends of climatic variability." The estimation of trends in climatic variability has not been undertaken, probably due to the complexity implied in the process of climate reconstruction [35] based on data obtained through dendroclimatological methods, as well as a strong climate signal in a well-reconstructed chronology. Nevertheless, this is a habitual process in climatology that only requires data of temperature and precipitation so that it is possible to calculate trends of climatic variability by means of available specialized software [59].
Dendroclimatology as a proxy tool for past climatic data has gained relevance during the last 12 years, which we observed in the number of published papers. It is probably because of the scrutiny in analyzing the effects of climate change and the need to rely on climatic data of higher and longer temporal resolution [60,61]. We also recognize a lack in Latin America of available dendroclimatology and climate change research until a decade ago [5,62,63]. In that respect, Mexico showed a different trend given that such an increment in published papers did not occur until 2013 [7].
The formation of collaborative research teams developing dendroclimatological research has also favored an increment in publications on this topic. There are mainly three groups of coordinated scholars: one formed by researchers from the United States of America working with Latin American colleagues, a second group including Mexican researchers [7], and a third one including scholars from South America, mainly from Argentina, Chile and Brazil [4,5]. This cohesion has favored the establishment of American continent [17]. These dendrochronological networks enable more robust future analyses, in topics like the expression of climate signal, the influence of general atmospheric circulation phenomena, the effect of global warming and its observation through dendrochronological techniques, as well as the higher involvement of tree species from tropical and subtropical environments.
A high percentage (57.4%) of papers were published by international publishers, perhaps because of the degree of impact and international reach of these journals. The trend, however, is to publish in journals from national research centers, maybe because of the degree of specialization of the research or because of its local and regional relevance. The development of dendroclimatological research in Latin America has been promoted by the training of researchers in recognized tree-ring laboratories and by international collaborations.
The surface of the studied area sampled is infrequently mentioned in papers about dendroclimatology, even though it could delimit the area of influence of climatic variables. We found only three papers providing such information as supplementary data within sections devoted to describing the study area [35,64,65] and giving little importance to it because the potential extension of forest cover is considered the limit of climatic influence [34,66].
Sites suitable for research are required to have certain characteristics for gathering data, such as slope, aspect, and land cover [67][68][69]. The areas fulfilling these requirements are in zones of altitudinal transition caused by the climatic variation present in them [70]. Consequently, dendroclimatic research can be observed at the local and regional scales because of the advantages and disadvantages implied in the extension of the study areas and the spatial distribution of the species to be analyzed. At both scales, however, the objective is to observe the effects of climate variability on forest mass growth [4,7,17]. In Latin America, regional studies are more frequently proposed than local studies, which in some cases could impede defining the study area or the area of influence of the observed parameters, given that local studies mainly focus on analyses belonging to different subdisciplines of dendrochronology.
Another advantage of dendroclimatology is to offer an updated perspective of the degree of anthropogenic impact, which is made through the evaluation of the increment in the cell structures of growth rings. Their cells store chemical components to which terrestrial ecosystems are continuously exposed [5,13,17]. It is still possible to make contributions, above all in the fields of wood anatomy and the identification of the cell structures, as well as the influence of climate warming, particularly in broadleaf woody species [9,26,47,71,72].
Publications about the reconstruction of past climates refer to general atmospheric circulation models like the El Niño-Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO), the Atlantic Multidecadal Oscillation (AMO), the North American Monsoon System (NAMS), as well as to other similar regional phenomena in South America, because these atmospheric phenomena have been shown to correlate with the of the annual radial growth of trees [8,11,58,[73][74][75][76][77][78]. The need for climatological information has grown in the last few years, given that the analysis and modeling of atmospheric dynamics require temporally precise climatological data to propose strategies for facing extreme hydroclimatological events [32,79,80]. The potential of growth rings as an archive of climatic variability has been widely discussed, with several published papers containing references to the capacity for dating climate variations with the added advantage of growth rings responding to annual and, in some cases, seasonal variations [4]. Therefore, one of the main trends of dendroclimatological studies is to reconstruct climatic variability and to compile useful databases to generate climatic and hydrological information in sites lacking instrumental data [5,43,48].  The more used methods for the studies of paleoclimate are based on the physical and chemical properties of the samples to be analyzed, and these could be applied in parallel with dendroclimatology. Lichenometry, ice cores, stratigraphy, and carbon-14 analyses are other methods of chronological dating; however, the temporal resolution of such methods is broad, that may result in dating errors varies between 10 and 100 years [86]. Therefore, dendroclimatology can be used as the most important foundation for the paleoclimatic analysis of the Holocene period because of its relevance being manifest in expressing climatic variability in annual terms for time intervals of several centuries [87][88][89][90][91]. When comparing dendrochronological techniques with other dating methods, it is possible to highlight the resolution offered by growth rings. Consequently, there is a capacity to compare atmospheric climate variability patterns between chronologies obtained in Latin America and other regions of the world [5,92].

Conclusions
Dendroclimatological studies in Latin America have been mainly developed in temperate climate zones and a lower percentage in tropical or subtropical areas, although, some chronologies have been developed using broadleaf tree species. The forest tree genera studied in Latin America are reliable for dating. The tropical genera areas of knowledge to consider in their work the information generated by dendroclimatology, which currently plays a promising role in the analysis of local and regional climatic processes. The generated dendroclimatic information allows for a better understanding of the spatial and temporal dynamics of climate variability and, consequently, about its effects on landscapes and society. This review allowed us to have an updated and broad view about the dendroclimatic studies in Latin America, identifying currently and frequently studied species, as well as geographic less studied areas, that require to increase dendroclimatic research efforts.

Patents
Author