Understanding trends in Zostera research, stressors, and response variables: a global systematic review of the seagrass genus

Abstract

Background: Seagrass meadows are ecologically significant habitats that are globally threatened. Thus, there is increased interest in conservation of seagrasses as they face widespread decline. Biotic and abiotic factors that influence seagrass can be classified as stressors, such as rising temperature and eutrophication. Our study met an imminent need to consolidate data from previous studies to discern knowledge gaps and identify trends in studies, stressors, species, and geographic origination of research for the genus Zostera. For our systematic review, the objectives were to (A) qualitatively assess and summarize the current state of literature focused on seagrass species within the Zostera genus and their stressors; (B) utilize data extracted from full-text articles to identify trends and knowledge gaps for the study of stressors, response variable measurements, species, geography, and study designs; and (C) map the distribution, type, and number of these studies globally.

Methodology: We included articles that focused on stressors associated with Zostera seagrass species, and excluded studies of other seagrasses and non-stressor related articles. We conducted a Web of Science search of all databases, concluding in January of 2021, followed by a standardized review and data extraction protocol using Colandr (colandrapp.com) as our article screening tool. All 15 review participants were trained on the same set of practice articles and decision trees to minimize variation between individuals. After full text extraction, we analyzed our data by frequency and association between species, stressors, and geographic locations studied.

Results: We screened 7,331 titles and abstracts and extracted data from 1,098 full-text articles. We found nutrients, temperature, and light were the most studied stressors. The United States of America produced the most articles in our review, followed by Australia. Zostera marina was most frequently studied, and our review found no stressor studies for five species in the genus. Studies most frequently measured response variables across multiple levels of ecological organization, including the individual plant, biotic community, and environmental conditions. As a part of our review, we made all extracted data publicly available as an interactive map.

Conclusion: Undertaking a review of global studies allowed us to assess more seagrass articles for a single genus than any prior systematic review, summarizing a breadth of stressor studies related to the Zostera genus. A team effort and standardized training minimized bias during screening and data extraction. Evidence limitations may exist due to the single database used in our search protocol, as well as species, geographic, and stressor biases in included studies. Our review creates a centralized knowledge base that serves as a foundational information source for Zostera research, while highlighting existing knowledge gaps in the literature.

Keywords: Coastal ecosystem; Global; Response variables; Seagrass; Stressors; Systematic review; Zostera.

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram.

Diagram for systematic reviews conducted on database searches. Adapted from Page et al. (2021). Diagram contains study-specific exclusion reasons, duplicates removed during the abstract screening process were removed automatically by ColandrApp (http://www.Colandrapp.com), duplicates removed during full-text screening were removed manually by members of the screening team. The PRISMA 2020 flow diagram template, used here, is from Page et al. (2021). The template is distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits distribution, remixing, adaptation and building upon this work, for commercial use, provided the original work is properly cited. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/.

Figure 2. Frequency of study type and study design combinations recorded by full-text articles.

Study type included field, mesocosm, and model studies. Study design included survey and manipulative studies. Numbers may be greater than the total number of articles due to models tagged in conjunction with other study types and study designs.

Figure 3. Publication year and species numbers for full-text articles included in the review.

(A) Number of articles per publication year for included full-text articles, displayed on left y-axis. Cumulative individual publication year as a percent of the total articles displayed on the right y-axis. Many of the 2021 publications were pre-prints at the time of literature search, and article numbers were not yet representative of 2020 and 2021. Data for years 1938–2020 are shown. (B) Frequency of Zostera species in included full-texts. Articles could have multiple species listed, which is why the total overall numbers for all species in this figure exceeds the total article number for our review. Zostera species with 0 articles not shown.

Figure 4. Article numbers per country recorded during full-text extraction.

Number of articles recorded for each country. Colors of bars are associated with continents/regions where purple is Africa, red is Asia, pink is Europe, green is Oceania, blue is North America, and grey is South America. Only countries with greater than five articles are displayed aside from one included article from South America, specifically in Chile.

Figure 5. World map of Zostera species recorded in full-text articles.

World map with dots representing the location(s) of each full text article. The colors of the dots correspond to 13 species of Zostera. Made with Natural Earth. Free vector and raster map data @ naturalearthdata.com.

Figure 6. Number of articles that recorded each of the 23 individual stressors with Zostera species and breakdown of stressors.

(A) Number of full-text articles that recorded each of the 23 individual stressors (left y-axis) and the number of full-text articles that recorded stressors under each of the umbrella categories (right y-axis). Individual stressors are colored corresponding to the umbrella categories shown on the right y-axis (Pollution, Climate Change, Increased Anthropogenic Presence, Intrinsic Factors). Since each article could have stressors under multiple umbrella categories, total full-text article numbers across all umbrella categories exceeds full-text article count. Articles that recorded multiple stressors under the same umbrella category were not double counted. These umbrella categories and their associated stressors are described in-depth in Supplemental Information 5. Abbreviations are as follows: Anthro. Use (Anthropogenic Use), Gen. Diversity (Genetic Diversity), Habitat Frag. (Habitat Fragmentation), Aquacult. Impacts (Aquaculture Impacts), Seagrass Comp. (Seagrass Competition). (B) Top 10 most frequently studied stressors, broken down by the total number of stressors recorded per article. Full-text articles were first grouped based on the total number of stressors recorded (1, 2, 3, 4+), then a count of each individual stressor was calculated within each group to obtain the number of articles. Since each article could investigate more than one stressor, the sum of all bars (e.g., for each stressor) can exceed the total article number (e.g., in the two-stressor group, the bar for nutrients (n = 197), is displaying that nutrients were recorded 197 times by articles that measured two stressors. The 10 most frequent stressors used were determined by the highest overall article counts shown in (A).

Figure 7. Frequency of occurrence for top ten stressors and top eight species in full-text articles.

Heat map showing the normalized frequency of combinations of stressor and species of Zostera. Numbers within the heatmap indicate the percent of articles in our review that studied each stressor for each of these eight Zostera species. Percentages in each column are relative to each species, but the heatmap color is scaled from the lowest possible percentage (0%) to the highest possible percentage of articles (64%) for all species. Normalization was performed by summing all articles that studied each stressor for each species, then dividing by the total number of articles (regardless of stressor) that examined that species. Since each article could investigate more than one stressor, the sum of the percentages within each column (e.g., for each species) can exceed 100%. The total number of articles that studied each of these top eight seagrass species were Z. marina: 722, Z. muelleri: 85, Z. noltii: 194, Z. japonica: 64, Z. capricornii: 59, Z. tasmanica: 12, Z. nigricaulis: 14, and Z. capensis: 20. Note we only included the top ten studied stressors as indicated by counts of articles that recorded each individual stressor, and we only included eight species of Zostera with greater than 14 articles based on counts of articles that recorded each species.

Figure 8. Number of full-text articles that investigated combinations of response variable categories.

(A) The number of full-text articles that investigated combinations of response variable categories. The number of full texts that investigated each individual response variable, which we categorized as (B) Plant Response Variables, (C) Community Response Variables, or (D) Environmental Response Variables. Abbreviations we used are as follows: Plant: Phys. Measures (Physiological Measures), Photosyn. Output (Photosynthetic Output), Reprod. Output (Reproductive Output). Community: Organismal/Microbial Div. (Diversity), Seagrass Comp. (Seagrass Competition). Environment: WC (Water Column), Char. (Characteristics), DO (Dissolved Oxygen).