2.2. Literature review and organization
The first step was to compile all the scientific literature on the coastal wetlands of Ecuador, Peru, and Chile (including original articles and scientific notes). For this purpose, the Google Scholar search engine and the Scopus and SciELO databases were used. The keywords employed for this search were 'wetlands,' 'mangroves,' 'marshes,' 'lakes,' 'lagoons,' 'estuaries,' and 'coastal', chosen in accordance with the definition of coastal wetlands established by the Ramsar Convention (2013). These words were used in conjunction with the target country (e.g., coastal lagoons Peru, coastal lakes Ecuador) in both Spanish and English. The search was conducted in duplicate, in order to maximize the efficiency of data compilation.
2.1. Article categorization
The literature was categorized based on three variables: (1) year of publication, (2) country where the study was conducted, and (3) thematic area, such as the discipline within which the work was carried out. The thematic areas were determined according to the variety of wetland research centers in the region (Salazar-Navarro et al. 2020). The articles were categorized according to a) pathology and public health (articles regarding health, pathogens and their interaction with fauna and humans); b) microscopic organisms (studies focused on microorganisms); c) ecology (documents addressing the ecological interactions of two or more different taxa); d) remote sensing (studies using satellite technologies and image-based prediction or modeling programs); e) management and conservation (articles focused on the correct management of wetlands); f) birds (papers on organisms of this taxon); g) other types of fauna (studies that include taxa other than birds); h) flora and vegetation (research focused on the study of the organisms of these taxa); and i) hydrology and sediments (papers on physical-chemical parameters, sediment sampling, hydrogeology, and hydro-seismic relationships).
2.3. Data analysis
The number of articles produced annually by each country was compared; for each country, 20 values were obtained (one for each year between 2000 and 2019). The Kruskal-Wallis test was used to compare these values per country; this test was chosen because the data did not follow a normal distribution in all cases (p <0.05 for Ecuador in the Shapiro-Wilk test).
Linear regression was performed for the total number of publications (without differentiating the country) and year. The slope of the generated line was used as a reference for growth (where positive, it was considered as a tendency to increase over time). The reliability of the model was calculated using the coefficient of determination (R2).
The diversity of thematic area by country was calculated using the Shannon-Wiener index (Harper, 1999):
H = - ∑(pi). ln(pi)
Where pi is the proportion of studies in a given thematic area, divided by the total number of studies. A diversity T-test (Magurran 1988) was carried out to compare the diversities of thematic area across countries.
A contingency table was prepared to compare whether the proportions of the thematic areas of countries were similar. An X2 test was performed in order to evaluate the association between the proportions of the thematic areas and countries. Cramér's V coefficient was also calculated; this coefficient enables evaluation of the degree of association of qualitative variables (in this case, thematic areas) by category (in this case, countries); this value varies between 0 and 1, where one indicates a strong association (Akoglu 2018).
All the statistical and diversity analyses mentioned in this section were conducted using PAST V 4.03 free software (Hammer et al. 2001).