Master’s Thesis (60 ECTS)
Social-ecological Resilience for Sustainable Development Master’s programme 2018/20 (120 ECTS)
Looking for the present in the past: Social-Ecological Memory and Palaeoecology to explore changes in Ciénaga
Grande de Santa Marta-Colombia
Lina Gutierrez-Cala
Stockholm Resilience Centre
Sustainability Science for Biosphere Stewardship
“El retraso en los caños nos permitió ver a pleno día la barra de arenas luminosas que separa apenas el mar y la ciénaga,
donde había aldeas de pescadores con las redes puestas en la playa, y niños percudidos y escuálidos que jugaban al futbol con pelotas de trapo”
Gabriel García Márquez – Vivir para contarla-
Acknowledgments
The work presented here is the result of many hearts genuinely beating and working together to keep it (and me) going forward for more than a year.
Thanks to my supervisor, Fernando for providing me both freedom to explore ideas and sense groundedness, whenever each was needed. Thanks to the co-supervising team, Constanza and Sam, for their support in planning, field work and feedback during the project.
Gracias a mi familia, que a pesar de estar lejos, siempre son el soporte que me permite dar el siguiente paso, cualquiera que ese paso sea. A Elena y Catalina, que son la familia que construyo cada día.
To my Masters Class, we have yet to realize this connection that started 2 years ago. So, looking forward to a life of friendship to continue enjoying it.
To Catalina, Andrés and Jorge, for your constant, encouragement, inspiration, and dedication.
Thank you for showing me the real meaning of a support system.
To Lina S. for jumping into this project to guide me and have my back every step of the way.
A Jesus Suárez, Javier de la Cruz y los demás pescadores de Buenavista y Nueva Venecia, por abrirme las puertas de ese mundo mágico que son los palafitos.
Table of contents
Abstract ... 6
Introduction ... 7
General Objective ... 9
Specific objectives ... 9
Theoretical Framework ... 10
Palaeoecology as a tool to inform resilience ... 10
Social-Ecological Memory (SEM) ... 10
Methods ... 12
Case Study ... 12
Ontology and Epistemology ... 14
Data Collection and Analysis ... 14
Paleoenvironmental reconstruction ... 15
Radiocarbon dating and sediment analyses ... 15
Demographic reconstruction based on archaeological 14C dates ... 16
Historic reconstruction with local communities ... 17
Qualitative data collection and analysis ... 17
Critical reflections on methods and data sources ... 18
Results ... 20
Palaeocological approach ... 20
Chore Chronology ... 20
Geochemical composition ... 20
Demographic reconstruction based on archaeological 14C dates ... 21
Participatory Historic Reconstruction ... 23
Agriculture and Extractive activities (Q 1-2) ... 23
Crops and cattle ranching ... 23
Mangrove use and exploitation ... 24
Massive Biodiversity Mortalities (Q 3-4) ... 24
Mangrove mortalities ... 24
Fish Mortalities ... 25
Impacts of degradation on local communities (Q 5-7) ... 26
Reduction of fisheries diversity and productivity ... 26
Social and ecological impacts of sedimentation ... 27
Reference (baseline) descriptions of CGSM... 27
Water flow modifications ... 28
Analysis of results and discussion ... 34
Period 1 - Freshwater – terrestrial influence (56cm-100cm -~5500 yr BP) ... 34
Period 2 High Variability - (56-11cm – 4036 yr BP) ... 35
Period 3 – Marine Incursion (11 cm – 2869 yr. BP- present) ... 36
Conclusion ... 38
Literature Cited ... 40
Interim SRC Research Ethics Review Form ... 50
Abstract
Mangrove forests are unique coastal ecosystems, formed through a complex network of terrestrial, estuarine, and marine processes that have provided a diverse assortment of societal benefits across time. Compounding anthropogenic pressures are driving critical mangrove degradation worldwide, threatening the wellbeing of coastal populations historically associated with these systems. The Ciénaga Grande de Santa Marta (CGSM) in northern Colombia is the largest coastal lagoon-delta in the Caribbean. It is inhabited by stilt-house communities who have developed an intricate livelihood and cultural relationship with the mangroves. The CGSM has experienced sustained social and ecological degradation over the last 6 decades, triggered by land-use change and disruption of hydrological connections. This study integrates Social- Ecological Memory and Palaeoecology to develop a historical contextualization of the biophysical and social dimensions of environmental change in CGSM. Integration of geochemical sediment analysis, C14 radiocarbon dating, and demographic inferences from archaeological evidence revealed three distinct periods over the last 5000 years. During this time sea level rise and hydroclimatic variability shaped the transition from freshwater to prevailing marine conditions, and modulated human occupation patterns in the area around 2000 years ago. In addition, participatory reconstructions with local communities offered nuanced descriptions about the spatial, temporal and contextual aspects of the degradation process, with profound social-ecological consequences. The interdisciplinary approach of this study indicates that CGSM is a highly dynamic social-ecological system that has been changing and reconfiguring across different time scales in response to both natural and human-induced processes, and contributes to the preservation of collective memory in this unique stilt-house community. Finally, it reveals the relative effects of biophysical and social drivers on driving social-ecological change under both millennial and decadal scales.
Resumen
Los bosques de manglar son ecosistemas únicos, inmersos en una compleja red de procesos terrestres, estuarinos y marinos que han contribuido innumerables beneficios sociales a través del tiempo. El efecto acumulado de diversas presiones antrópicas está acelerando la degradación de los manglares a nivel mundial, amenazando el bienestar de las comunidades costeras que históricamente han habitado estos sistemas. La Ciénaga Grande de Santa Marta (CGSM) al norte de Colombia, es el delta costero más grande del Caribe, y es hogar de pueblos palafíticos, cuyos habitantes han desarrollado relaciones profundas culturales y de subsistencia con los bosques de manglar. La CGSM ha experimentado una degradación social y ecológica sostenida durante las últimas 6 décadas, desencadenada por cambios en el uso de la tierra y alteración de sus conexiones hidrológicas. Este estudio integra perspectivas de Memoria Socio- Ecológica y Paleoecología para desarrollar una contextualización histórica acerca de las dimensiones biofísicas y sociales del cambio ambiental en la CGSM. La integración de análisis geoquímicos de sedimentos, datación por radiocarbono C14 e inferencias demográficas basadas
en evidencia arqueológica, sugiere tres períodos diferenciados a lo largo de los últimos 5000 años. Durante este tiempo, el aumento del nivel del mar y la variabilidad hidroclimática promovieron, hace alrededor de 2000 años, una transición de condiciones dulceacuícolas hacia el establecimiento de condiciones marinas que perduran hasta hoy; modulando a su vez los patrones de ocupación humana de la época. Adicionalmente, reconstrucciones participativas con comunidades locales ofrecieron descripciones detalladas sobre los aspectos espaciales, temporales y contextuales del proceso de degradación, y las profundas consecuencias socio- ecológicas que ha conllevado. Esta aproximación interdisciplinaria indica que la CGSM es un sistema socio-ecológico altamente dinámico que se ha ido reconfigurando a través del tiempo, como respuesta a procesos tanto naturales, como humanos. Asimismo, este estudio contribuye a la preservación de la memoria colectiva de estos pueblos palafíticos, únicos en Colombia.
Finalmente, el estudio revela los efectos relativos de factores sociales y biofísicos sobre el cambio socio-ecológico en escalas temporales de milenios a décadas.
Introduction
Mangrove forests are coastal ecosystems distributed over tropical and subtropical coastal regions, representing a unique forest assemblage that thrives under the influence of terrestrial, estuarine, and marine processes (Urrego et al. 2018). This complex environmental setting displays high variability in salinity, sea-level rise, precipitation, and hydrological connectivity;
which together shape a myriad of unique ecological adaptations over different time scales (Kathiresan and Bingham 2001). Mangrove ecosystems sustain human coastal populations as they provide several ecosystem services including fisheries (Nagelkerken et al. 2008), timber fuel (Castellanos-Galindo et al. 2020), coastal protection from erosion (Kamal et al. 2017), extreme weather events (Hochard et al. 2019), and can act as major global C sinks reducing atmospheric CO2, thus they are fundamental in a climate change context (Atwood et al. 2017).
Notwithstanding their social-ecological importance, mangroves are facing degradation rates of
~2% every year, and an overall loss between 35-85% in the last 25 years worldwide, mainly as a result of land conversion for agriculture, aquaculture, and coastal development (Duke et al.
2007, Alongi 2015). Mangrove forests in Colombia have withstanded high anthropogenic pressures from severe landscape transformations since European colonization times (Sánchez Páez et al. 2004, López-Angarita et al. 2016), thus when studying these systems it is crucial to consider the compound effects of long-term environmental and human drivers.
In the context of long-term degradation, paleoenvironmental reconstructions offer a unique standpoint to explore mangrove ecosystem dynamics at different temporal scales (Dahdouh- Guebas and Koedam 2008, Ellison 2008, González et al. 2010, Velez et al. 2014). These reconstructions provide historical patterns of disturbance and recovery, informing ecological resilience and highlighting the implications of these processes for human wellbeing (Saulnier- Talbot 2016).
Traditional ecological knowledge (TEK) (Berkes et al. 2000) is an attribute of societies historically attached to resource use, it accumulates over time, and is constantly updated through experience (Ohmagari and Berkes 1997). TEK is anchored in local observation of environmental processes, empirical understanding through practices of resource use, and belief systems that create meaning around people and their environment (Berkes et al. 1994).
Combining paleoenvironmental and traditional knowledge sources may, thus, facilitate the contextualization of long-term ecosystem processes and social dynamics. However, retrospective research in mangroves has remained largely within the disciplinary boundaries of natural sciences (i.e ecology, geology, chemistry), and approaches that combine social perspectives are scarce (Dahdouh-Guebas and Koedam 2008).
The Ciénaga Grande de Santa Marta (CGSM) is a unique social-ecological system (SES). It is the largest coastal deltaic-estuarine complex system in the Caribbean and it is among the most productive wetlands globally (Hernández-J. and Gocke 1990, Cloern and Jassby 2008). It is home to stilt-house communities, unique cultural heritage societies that develop livelihoods around multidimensional connections with the surrounding water and mangrove forests (Vilardy Quiroga 2009). The CGSM has undergone critical environmental degradation in the last 60 years, triggered by land-use changes and highway constructions that altered key hydrological connections between the lagoon, the Caribbean Sea, and freshwater sources (Vilardy et al. 2011). This has resulted in soil hypersalinization, massive biodiversity mortalities (Botero and Salzwedel 1999, Jaramillo et al. 2018), and loss of key ecosystem services (Vilardy et al. 2011).
Management interventions have focused solely on restoring hydrological connectivity through dredging of silted channels. However, the wetland has not completely recovered due to the interplay of complex processes such as ENSO (Blanco et al. 2006, Rivera-Monroy et al. 2006) and emerging ecological interactions such as species reconfigurations and invasive species (Perdomo et al. 1999, Röderstein et al. 2014) . Instead, it has triggered dependency dynamics in which lack thereof institutional support for sustained dredging has led to further degradation (Jaramillo et al. 2018).
Previous paleoenvironmental studies in the CGSM have focused on millennial scales showing the influence of Holocene sea-level rise and subsidence processes on the establishment of marine conditions, around 2000 years before present, hereafter (yr. BP) (Wiedemann 1973, Velez et al. 2014). Likewise, historical insights from bibliographic revisions describe a transition from subsistence agriculture and fisheries in pre-colonial times, towards increased land-use intensification since the XIX century (Vilardy et al. 2011). In the context of environmental degradation, participatory methods with local communities have informed the loss of ecosystem services during more recent times (Vilardy et al. 2011). Environmental change is monitored through of quantification of mangrove cover, salinity, sedimentation rates, hydrological flow. However, the timespan of these monitoring is limited to the onset of the degradation process around the 1960’s (Botero and Salzwedel 1999, Blanco et al. 2006,
Jaramillo et al. 2018, INVEMAR et al. 2019), and thus little is known about the century-scale ecosystem processes that may influence the overall response of the CGSM system.
By integrating the longer time-scale paleoenvironmental insights and the more recent historical perspectives of local communities, this study seeks to answer the following research question:
¿How can social-ecological memory and palaeocological reconstructions explain social- ecological change in the CGSM?. It pioneers an integrative approximation to understand both the environmental and social implications of the loss of this important system in Colombia while preserving collective territorial memory of the stilt-house communities.
General Objective
To develop a historical perspective on social-ecological change in the CGSM, by integrating palaeocological analyses and participatory approaches with local communities.
Specific objectives
1) Identify signals of biophysical change across time and early patterns of human occupation through palaeoecological record analyses, using C14 radiocarbon dating,. geochemical composition analysis of sediments form the CGSM and archaeological evidence.
2) Retrieve local perspectives, through participatory methods, on the temporal, spatial and contextual aspects of the degradation process in the CGSM.
3) Integrate evidence of environmental and social change across millennial and decadal scales.
Theoretical Framework
Palaeoecology as a tool to inform resilience
Paleoecology offers mutli-proxy approaches to identify signals of change that can inform past ecosystem features (i.e vegetation, hydrology, climate) and processes, including anthropogenic influence (Quinlan et al. 2008, Crumley 2018). Walther’s Law of Uniformitarianism underpins palaeoecological reconstructions as it states that sediments from different facies juxtapose, and accumulate in cross-sections that lie on top of each other (Middleton 1973). Hence, as in undisturbed records the age of sedimentary units increases with depth (Ellison 2008), studying one sediment core can inform processes at the catchment basin scale and provide temporal perspectives of ecological and sedimentological processes.
Paleoenvironmental reconstructions play a key role in ecosystem management and restoration, by informing reference conditions, targets, and ecosystem services trends that may be difficult to identify based solely on current ecosystem configurations (Dearing et al. 2012, Smol 2014).
For instance, paleoenvironmental reconstructions of volcanic activity integrated with local indigenous perspectives on landscape change, facilitated the co-creation of volcanic hazard management plans in New Zealand, thereby reducing risk exposure and enhancing adaptation responses to volcanic eruptions in local communities (Cronin et al. 2004). Similarly, integration of palaeocological analyses and ecosystem services assessment in the CGSM (Velez et al. 2018) underpins the proposed Ecosystem-based risk management plan for the system, by informing biophysical change over time and defining management outcomes based on ecosystem state prior to generalized degradation.
In the context of resilience, Palaeoecology can provide insights into ecosystem resilience by describing reference conditions, timing, magnitudes of change and potential feedback mechanisms behind it (Saulnier-Talbot 2016). However, narrow definitions of the concept halt the potential to inform the social implications of the long-term perspectives offered by the discipline (Davies 2018). By understanding social-ecological resilience as the capacity of a SES to absorb disturbance and reorganize in the face of change, as to retain its identity and feedbacks (Folke. 2016), this study explores the complex links between social and ecosystem dynamics across timescales from an interdisciplinary approach to SES change.
Social-Ecological Memory (SEM)
Communities engaging in ecosystem management practices or natural resource use, develop (SEM), understood as the historic pool of knowledge, experiences and practices related to ecosystem management, collectively held by a community in a Social-Ecological System (SES) (Barthel et al. 2010). Prior to the concept of SEM, the idea of “social sources of resilience”
(Berkes et al. 2002) already included social memory within the arrangement of social capital features that enhance resilience in a SES, by enabling diversity, renewal and reorganization
capacity in the face of change . Therefore, by explicitly incorporating ecological perspectives, SEM can enhance adaptive capacity in SES, increasing response diversity to deal with disturbance or uncertainty (Davidson-Hunt and Berkes 2003, Norberg and Cumming 2008).
However, given the dynamic nature of memory, maladaptive resource management practices can also be incorporated within SEM, potentially amplifying social-ecological degradation dynamics and reinforcing lock-ins of undesirable resilience (Dornelles et al. 2020).
SEM can be understood as a dynamic feedback constituting the Person - Practice - Place complex, with ecoliteracy, identity of place, and attachment to a given place, emerging as features from these interactions (Fig.1) (Kim et al. 2017). In this framework a
“person” refers to the smallest unit of memory accumulation and extends to the individuals with whom the person shares practices, social-ecological interactions, and approaches to ecosystem
management. The
“Practice/Experiences” in a SES
represent collective, past and current habitual activities regarding ecosystem management or use, including cultural, historical, and spiritual aspects. “Place” is the physical space in which an individual develops and accumulates experiences, practices and knowledge about ecosystem management.
In this framework, interactions between the constituents of SEM, generate emergent features relevant for enhancing SEM. Ecoliteracy emerges from the interaction of “person” and
“practice/experience” and it is defined as “the ability to identify names, uses, and related stories of living organisms and natural phenomena within their social-ecological systems, perpetuated by oral transfer of traditional ecological knowledge (TEK)” (Kim et al. 2017). As it builds upon TEK (Berkes et al. 2000), ecoliteracy is thus manifested and fostered by practical engagement with the ecosystem, making TEK holders key ecoliterate members in the community who will facilitate intergenerational knowledge transfer (Berkes 2009). “Identity of a place” is defined by the accumulation of “practices/experiences” of communities in the physical place over long periods of time. This feature is especially relevant when identifying changing perceptions of wellbeing and sustainability of the space. “Place attachment” broadly relates to the person-place bond, built over large temporal scales through a person’s interaction with a physical space (Kim et al. 2017). This complex connection has been thoroughly explored from different anthropological angles, and termed “topohilia” (Buttimer and Seamon 2015). Nevertheless, these approaches have emphasized subjective and cultural perspectives related to a place, rather than knowledge, understanding about environmental aspects.
Ecology and Society 22(2): 27 https://www.ecologyandsociety.org/vol22/iss2/art27/
practices (Barthel et al. 2010). Studies of SEM may include the attributes of practices (or experiences) of ecosystem management, and the place where the practices occur. In our study, we adhere to this understanding of SEM and suggest it as a person-practice-place complex. The dynamic feedback among these three pillars manifests as ecoliteracy, place attachment, and the related identity of a place (Fig. 1).
Fig. 1. A conceptual framework for social-ecological memory as a person-practice-place complex with ecoliteracy, place attachment, and identity as emergent outcomes of their complex linkages.
Here “person” refers to the individual as the main agent of memory, and to individuals who influence or share practices with the main agent. In an SES, individuals are regarded as potential stewards of their communities (Chapin et al. 2009). An individual being is also an actor in social-ecological processes, and is regarded as the smallest scale or unit in the analysis of SES resilience (Cumming et al. 2015). An individual’s knowledge, worldview, social-ecological interactions, and approaches to ecosystem management result from the individual’s experiences;
experiences inform perceptions that define the individual’s attitudes and behaviors; and depending on the nature of social feedback, the individual’s actions can be either reinforced or corrected. In turn, individuals, through their ecological knowledge, cognitive skills, and behaviors, which are collectively termed as ecoliteracy (Capra 1996), can impact ecosystem patterns and processes (McBride et al. 2013).
To refer to a person’s ecological competence, we purposely choose ecoliteracy over environmental literacy and ecological literacy, because the concept of ecoliteracy includes spiritual and holistic components, thereby suggesting a character with well-rounded abilities of head, heart, hands, and spirit (McBride et al. 2013).
We also draw from Lam (2014) and Pilgrim et al. (2007) in defining ecoliteracy as the ability to identify names, uses, and related stories of living organisms and natural phenomena within their social- ecological systems, perpetuated by oral transfer of traditional ecological knowledge (TEK). Given that TEK is defined as a knowledge-practice-belief complex (Berkes 2009), a person possessing TEK may also be considered an ecoliterate person.
Studies on TEK have discussed how the knowledge systems of
local and indigenous groups contribute to building resilience and adaptive capacity in an SES, particularly at the community level (e.g., Berkes et al. 2000, Berkes and Seixas 2005, Berkes and Turner 2006, Berkes 2008, 2009).
Many traditional communities, particularly resource-dependent communities, have sustained themselves through their ecoliterate members (Pilgrim et al. 2007). In the modern context of industrialization and globalization, ecoliterate leaders guide their communities to premeditate and withstand undesirable social- ecological impacts, with their TEK and holistic beliefs related to their SES. A well-known example is of Sarah James, a Native Alaskan who leads her community’s efforts on protecting the Porcupine caribou habitat and the sustenance of indigenous communities dependent on the caribou, by resisting the moves of the oil industry of the United States in Alaska (Goleman et al.
2012).
A person’s ecoliteracy is developed or manifested though
“practice” in a real SES, which refers to the past or present habitual activities and experiences in relation to ecosystem management. Barthel et al. (2010), for example, discuss gardening activities in allotment gardens in the Stockholm urban area, Sweden as an example of such an ecosystem management practice. Practice is the result of long-term interaction between person and place, thus incorporating cultural, historical, and spiritual aspects. Although some ecologists are skeptical of indigenous and local knowledge and practice, much of the current scientific knowledge on biology has also undergone verification in apprenticeship with local knowledge holders (Berkes 2008). It is in this sense that various SES studies acknowledge the importance of SEM without explicitly referring to “social- ecological memory” and instead focus on the alternative knowledge systems related to local ecosystem stewardship practices.
“Place” is the physical site in which a person has experienced and learned through practice about ecosystem management, complex systems thinking, and the link between nature and humans.
Barthel et al. (2010) label these physical sites in urban areas as
“pockets of social-ecological memory.” Physical sites may vary in spatial-temporal scale, but all have a “spatial identity” that is influenced and shaped by endogenous and exogenous variables within the SES; similarly “place identity” is directly established by people’s experiences and interaction within a place. The spatial characteristics of an SES are critical to the resilience of the system (Cumming 2011a), thus it is important to integrate the concepts of place and space within the scope of SES research.
A comprehensive modern discussion of the concepts of space and place was initiated by Tuan (1974, 1977). He described the scope of individual experiences in differentiating space and place, and further explained that an individual’s scope of experience is the arena in which the individual learns from her/his encounters and is influenced by her/his senses. The concept of place is related to the individual’s perceptions and interactions within her/his environment, the resulting feeling of well-being, and the cognition or comprehension of sustainability of the space. Experience is thus the starting point for recognizing sustainability-related issues and attempting to manage the resilience of the individual’s self and place or connected environment. Our study draws attention to the role of autobiographical memory with such experiences in
Figure 1. Social-Ecological Memory framework proposed by Kim et a. (2017)
Methods Case Study
The Ciénaga Grande de Santa Marta (CGSM) is one of the largest coastal-lagoon-delta ecosystems in Colombia, extending over 1280 km2 (Jaramillo et al. 2018) and comprising a complex mosaic of ecological units including lagoons, alluvial plains, creeks and inner mangrove swamps (Botero and Salzwedel 1999). The main water bodies are the CGSM lagoon (450 km2), and the Pajarales Complex (800 km2), which includes the stilt-house communities of Nueva Venecia and Buenavista, and a wider network of streams and inner lagoons (i.e: La Ahuyama, Luna, Conchal, Alfandoque (750 km2) (Fig. mapa). It is separated from the Caribbean Sea by the Isla de Salamanca sandbar on the north, on the east it is bounded by the Sierra Nevada de Santa Marta mountain (SNSM), and 5 rivers draining from the Sierra Nevada mountains, and by the Magdalena River to the west. The zone has an annual water deficit of 1031 mm/yr (Botero and Salzwedel 1999). It is located in the Intertropical Convergence Zone (ITCZ) which determines a bimodal pattern of precipitation (Haug et al. 2001, Urrego et al.
2018), with a dry season between December-April, July-August, and a rainy season from May- June and September-November (Botero and Salzwedel 1999, INVEMAR et al. 2019) (Botero y Salzwedel, 1999, INVEMAR 2016). The region experiences El Niño Southern Oscillation (ENSO) events, resulting in prolonged droughts, severe reductions in freshwater streamflow, and sea surface temperature and salinity increases (Giannini et al. 2000, Blanco et al. 2006).
The system has been severely intervened by the construction of two highways: one “coastal highway” connecting the cities of Barranquilla and Ciénaga on opposite ends of the lagoon, which interrupted key hydrological flow between the Caribbean Sea and the CGSM (1955- 1960); and another parallel to the Magdalena River (“river highway”) interrupted hydrologic connections between the River and the system along the alluvial plain during the early 1970’s (Vilardy et al. 2011). Both projects failed to consider drainage systems to allow hydrological connection of the wetlands on both sides of the road and, coupled with climatic regime alterations (i.e ENSO) and agriculture pressure (i.e freshwater diversions), led to increased pore salinity, triggering mangrove mortality and overall acute ecosystem degradation (Cardona and Botero 1998, Perdomo et al. 1999, Blanco et al. 2006, Jaramillo et al. 2018). Mangrove cover has reduced to nearly half between 1956 - 1996, slightly bouncing around 2005 but displaying overall poor recovery (Rivera-Monroy et al. 2006, INVEMAR et al. 2019).
Similarly, ecosystem services associated to fisheries have significantly decreased (Vilardy et al. 2011) affecting the most important income source for local communities. Increased runoff from Magdalena river and smaller tributaries has caused siltation in most of the channels of the alluvial plain, further limiting freshwater exchange with the PC and CGSM (Jaramillo et al.
2018). Management interventions in late 90’s focused on re-opening the clogged channels:
Aguas Negras, Clarín, Renegado to recover freshwater flow and decrease salinity. Helped by an unusually intense precipitation in 1999 (Röderstein et al. 2014), the strategy seemed
effective, however under regular precipitation regimes, dredging is insufficient to promote mangrove recovery (Jaramillo et al. 2018).
Stilt-house communities are inhabited by around 3000 people, (Aguilera-Díaz 2011, Suárez J 2019, pers.comm). Education coverage is significantly low, as 57,7% of the population over 15 years old in Nueva Venecia and 31,4% in Buenavista are illiterate (Aguilera-Díaz 2011). Access to public services is low, with just partial access to electricity, and absence of sewage management system and drinking water (Pineda Vivas 2018). Main livelihood fisheries include crustaceans and mollusk resources. Environmental change, destructive practices, and overfishing below reproductive size, have led to sharp reductions in the resource (Aguilera- Díaz 2011, Vilardy et al. 2011). In sum, historic low access to basic services and infrastructure, reinforce negative dynamics between poverty and environmental degradation.
Figure 2. Geographical location of the CGSM. Map describes mangrove cover, location of stilt-house communities and sampling location of the sediment core.
Ontology and Epistemology
This study builds on Critical Realism as ontological perspective by assuming that one reality exists, in this case: a degraded CGSM, but its inherent intractability of social and natural phenomena demands multiple methods, as all offer limited explanations. Therefore, in order to achieve the aim of this thesis, this reality must be studied from the broadest perspective possible and any claim about it, should be critically examined (Moon and Blackman 2014). Critical Realism realizes a “compatibility of ontological realism, epistemological relativism and judgmental rationality” (Bhaskar and Hartwig 2010) under which to integrate diverse evidence sources with the aim of understanding complex phenomena.
Epistemologically, this study builds on two fundamentally different approximations to knowledge acquisition. The palaeocological approach is built on an Objectivist epistemology which assumes that signals of ecosystem change are identifiable through the sediment record, and by doing so we can understand the biophysical dimensions of ecosystem degradation (Patton 2002, Moon and Blackman 2014). Contrastingly, the exploration of social-ecological memory with local communities builds on a Constructivist epistemology as this approach assumes that local communities, as subjects, develop their own understanding of the changes in the CGSM, the object, through their own engagement with their realities as inhabitants of the mangrove. As such, no single “truth” is to be determined and any claim is the result of cultural, historical and social perspectives in relation to the space they inhabit (Crotty 1998, Creswell and Creswell 2017).
Therefore, the philosophical perspectives of this research are rooted in Eco-Phenomenology (Ingold 2000, Brown and Toadvine 2003) as it aims to understand how people have experienced change in a place and how collective knowledge about this can inform degradation processes;
and a Post-Positivist perspective that multiple methods and perspectives must be integrated to understand and predict social-environmental changes (Moon and Blackman 2014).
Data Collection and Analysis
The mixed methods approach developed in this thesis aim to provide quantitative and qualitative evidence on the degradation processes taking place in CGSM over time (Fig. 3).
Palaeocological and qualitative data in this study were retrieved empirically through fieldwork in the area conducted from October 2019 to January 2020.
Paleoenvironmental reconstruction
A 100-cm sediment core was collected at a water depth of 0.7 m, on the western flank of the main lagoon of the CGSM (CGSM-P1, 10°55'0.71"N, 74°30'37.97"W) using a piston corer (Uwitec-trademark). The coring site was chosen as it is simultaneously close to the stilt-house communities, allowing for shared social and environmental inferences, and far from the confluence of rivers into the CGSM and sites of dredging operation that could alter the sediment layer deposition. The core was vertically extruded and sliced into 102, 1-cm thick, sections which were individually labeled and stored at 4ºC. Each section was divided in halves, one kept as reference material and the other further divided for subsequent analyses of: radiocarbon dating, organic and carbonate content and geochemical composition. All working halves were dried in a BINDERTM Drying Chamber (Thermo Fisher Scientific) at 60ºC overnight and grinded with mortar and pestle to homogenize.
Radiocarbon dating and sediment analyses
Age chronologies for the sediment record were retrieved from 4 samples along the core using Accelerator Mass Spectrometry (AMS) for 14C radioisotope (Direct AMS, WA-USA).
Geochemical composition was measured through X-Ray Fluoresence (XRF) using a handheld spectrometer analyzer (XMET 7500 Oxford Instruments). For each 1cm slide, 2-3 gr of sample were mounted in plastic rings and covered with Chemplex film to standardize transmittance.
Mean value from duplicate measurements were recorded. Data were log +1 transformed to control for scale differences.
Figure 3. Methodological integration of Social-Ecological Memory and Palaeocological record.
Arrows indicate a temporal scale spanning from millennia – decades. Methods are distributed according to the temporal scale they inform.
Percentage dry weight, organic matter and carbonate content (%) were determined by sequential loss on ignition (Dean 1974, Heiri et al. 2001). A set of porcelain crucibles were weighted to 3 decimal digits in a FisherbrandTM Analytical Balance (Thermo Fisher Scientific). Subsequently, a fraction of 1-2 gr of wet sample was placed in each crucible and re-weighted. Crucibles were placed overnight in a drying chamber set to 105ºC, re-weighted, and the percentage weight remaining was calculated by subtraction. The same sample was used for organic content determination through loss on ignition using a Muffle Furnace (Thermo Fisher Scientific) set to 550ºC for 2 hours, then re-weighted and the percentage weight remaining was calculated by subtraction. Carbonate content was determined putting the same sample in the furnace set to 950ºC for 4 hours, re-weighted and the percentage weight remaining was calculated by subtraction (Heiri et al. 2001).
Demographic reconstruction based on archaeological 14C dates
To make inferences about fluctuations in human populations across time, I used statistical modelling based on frequencies of calibrated C14 radiocarbon dates of archaeological remains (Riede 2009) . This analysis assumes that the frequency and distribution of archaeological records in a place is proportional to the population size in that place at a point in time (Rick 1987, French and Collins 2015). For this I collected 104 calibrated 14C dates retrieved from sites with dated archaeological remains (i.e pottery, animal fragments) in the Caribbean coast of Colombia. Data was obtained from the database of the Colombian Institute of Anthropology and History (ICANH). Each calibrated radiocarbon date is associated with a probability distribution of the likelihood of the estimated age, for that given remain (Ramsey 2017). Using the “rcarbon” v. 1.4.1 package in R (Crema and Bevan 2020, Bevan 2020) I applied Bayesian statistical inference to estimate the summed probability distribution of the 104 radiocarbon dates, which serves as a proxy to inform the demographic variability across time in the Caribbean coast of Colombia.
This analysis is based on frequencies of available dates, thus it is susceptible to sampling bias (Ramsey 2017). To control for over representation of evidence in a given place, C14 dates were grouped by sampling site and an association limit of 100 years was set, such that only samples from the same place and < 100 years apart were grouped together. Taphonomic conditions and other environmental factors are also acknowledged to influence the preservation of evidence and, thus, the probability distribution (Bocquet-Appel and Demars 2000). However, it is a powerful tool for estimating demographic fluctuations over large temporal scales (Rick 1987).
Historic reconstruction with local communities
Through focus group discussions (FGD) and participatory mapping with local fishermen I retrieved local interpretations and detailed descriptions of the temporal, spatial and contextual aspects of the main landscape and livelihoods transformation processes of the inhabitants of the CGSM. Discussions focused on three main topics: Agriculture and Extractive Activities, Massive Biodiversity Mortalities, and Socioeconomic impacts of degradation. The specific questions (Q) within each topic were:
Agriculture and extractive activities
1. When, where and how have main agriculture and extractive activities such as cattle ranching, extensive crops, and mangrove logging, taken place?
2. When, where and how have major landscape and hydrological modifications associated with such activities, taken place?
Massive Biodiversity Mortalities
3. When, where and how was the onset of massive mangrove mortalities?
4. When, where and how was the onset of massive fisheries resource mortalities?
Socioeconomic impacts of degradation
5. When, where and how did significant reductions in fisheries captures were first perceived?
6. When, where and how was the onset of navigability issues associated with increased siltation?
7. Other impacts not considered in the questions?
Qualitative data collection and analysis
During the field season between November 2019 and January 2020, one focus group discussion (FGD) was conducted in each of the stilt-house communities: Nueva Venecia and Buenavista.
I visited the communities twice to establish initial communication with the local contact person and to describe the purpose of my visit, the goal of the activity and the selection criteria for participants. In the case of the FGD in Nueva Venecia, the local contact person was a community leader who works as a fisheries observer, whereas for the FGD in Buenavista was the president of the Fishers Association of Buenavista. Following the conversation with the local contacts, a date was set for developing the activity.
Selection criteria for participant recruitment included people over 60-years old, whose livelihoods were associated with natural resource use - assuming these activities are the most affected by ecosystem degradation- and who had lived over 30 years in the CGSM to increase the possibility of retrieving older records of events. Recruitment followed a “snowball approach” (Tenzek 2018), whereby local contacts recruited other community members to
participate, groups ranged between 4 to 9 participants. This strategy facilitated contacting fishermen whom would have been difficult to invite otherwise, while still attaining specific criteria for participation (Krueger 1989). The temporal reconstruction was bounded by the birth date of the oldest participant in each group plus 10 years, to cover events occurring within the participants’ lifespan.
The FGD was divided in two sections: Introduction and socialization of the project, where I emphasized voluntary participation, confidentiality and consent forms were signed. Space was given for extensive discussion with participants about the project before launching the activity.
The other section addressed the main discussion topics, through the specific questions abovementioned. Answers were collated in a time line diagram and a map of the area.
Discussions were video and audio recorded, and audio files for both FGDs were transcribed verbatim. Transcripts were coded deductively (Morgan 1997, Saldana 2015), following a structural coding method (Namey 2008, Saldana 2015) guided by the 3 overarching themes addressed in the FGD. Subsequently, I applied subcoding, also known as “nested coding”
(Gibbs 2007, Saldana 2015) to identify social-ecological memory components following (Kim et al. 2017). Finally, a round of inductive coding allowed for the identification of emerging topics that were not addressed through pre-defined questions but were relevant from participant’s perspectives.
Critical reflections on methods and data sources
This study encountered the inherent methodological limitations of any other historical ecology approach including use of proxy variables, disparate evidence sources and partial explanation of processes with no causal links (Crumley 2019). Paleoecological evidence for this study was retrieved from one single sediment core, which is a standard practice in this discipline given resource intensive analyses, but observed patterns could change if more replicates were included. Similarly, evidence only considers geochemical indicators as proxies of environmental change but does not include bio-indicators directly informing ecological dynamics of mangrove communities.
Qualitative data through focus groups is valuable as a first attempt to use participatory methods in CGSM with the explicit aim of retrieving detailed historical records and local testimonies about degradation. This method was chosen given its potential to gather diverse perspectives in limited time; however some voices dominate these settings concealing other perspectives.
Representation in the FGD was biased as only one women participated in the activity as they don’t usually engage in fishing activities. Also not all the participants who were available to participate met the age criteria, but their input was equally valuable. Overall, this is the first effort to integrate palaeoecological and social evidence sources to understand the complexities of the CGSM case, so it provides a novel approach and a useful base upon which to conduct further methodological improvements in the future.
Photographic record of participatory mapping and timeline reconstructions in Buenavista and Nueva Venecia.
Collection and processing of the sediment core. Orange core was a pilot trial to prior to use UwitecR corer . Sediment core was mounted vertically to extract 1-cm thick slices as show in lower image-
Results
Palaeocological approach
Chore Chronology
The three radiocarbon dates retrieved from sediment samples along the core are reported in (Table 1). The basal date at depth 56 cm was 4731 years before present (yr BP), at 31 cm was 6546 yr BP, and the upper date at 11 cm was 2869 yr BP Sample 31-cm shows an inverted chronological order, thus has been excluded from the analysis until the age of the basal sample 102-cm is determined. Deposition of elsewhere reworked older sediments on top of younger layers, or fluid turbulence and slumping processes could explain the inverse order or the samples but this does not necessarily invalidate the palaeocological interpretation if other samples are coherent (Martinsen 2003).
Geochemical composition
A total of 100 samples were analyzed forPatterns in Ti, Fe and K concentrations were used as indicators hydroclimatic conditions associated to rainfall and fluvial sediment input, Ca as an indicator of salinity, and Ba/Ti as proxy for marine productivity (Haug et al. 2001, Davies et al.
2015, Salgado et al. 2020). Percentage of organic matter is a proxy for carbon accumulation and carbonate percentage is an indicator of CaCO3 accumulation (Bouillon et al. 2003). Such multi-proxy approach, provides a solid evidence base to characterize important hydroclimatic and geochemical features of the mangrove ecosystem across time.
Indicators for rainfall and erosion (Ti, Fe, K) (Fig. 4A) are high and homogeneous on the basal section (100 – 60 cm) and decline towards the upper (25 – 0 cm), whereas salinity (Ca) and marine productivity (Ba/Ti) show opposite trends (Fig. 4A). Temporal variation on geochemical composition shows a consistent section of increased variability (30-60 cm) across several indicators. Organic carbon values (1-8 %) fall within reported values for tropical
Sample Depth (cm) Age (yr. BP) s-error
D-AMS 039465 11 2869 23
D-AMS 039466 31 6546 36
D-AMS 039467 55-56 4731 28
D-AMS 039468 102 In process In process
Table 1. AMS - C14 radiocarbon dates expressed in years before present (yr BP), retrieved at different core depths (cm), with associated reported error. Sample from 102 cm is currently in process
mangroves (Bouillon et al. 2003) and display increased variability between ~ 35-60 cm, consistent with XRF signals, and a sharp increase on the upper 10 cm of the core. Carbonate (%) and Ca show low values along the core (100 – 10 cm), and both peak on the upper section (10 – 0 cm).
Demographic reconstruction based on archaeological 14C dates
Based on the archaeological evidence, the model of demographic variability (Fig. 4C) shows the estimated human occupation in the Caribbean coast of Colombia during the last 11000 yr BP Higher probability values represent larger estimated population sizes, arrows indicate peaks in population estimates occurring at 6000 yr BP, 2000 yr BP, ~1600 yr BP and ~800 yr BP The model is coupled with the sediment geochemical concentration of Ti (cps) retrieved through XRF analysis (Fig. 4B). Interestingly, from ~4500 - ~3000 yr BP the period of highest hydroclimatic variability, denoted by high amplitude fluctuations in Ti, co-occurs with the period of lowest probability of human occupation in the Caribbean coast of Colombia (Fig. 4 B,C), whereas after ~2000 yr BP frequent peaks in probability denote a sustained increase in estimated human population size, with the highest peak around ~500 yr. BP.
Figure 4. A. XRF-geochemical composition of selected elements expressed in counts per second (cps). Ti: Titanium, Fe: Iron, K: Potassium, Ba: Barium, organic matter and carbonate (%). Green shade: Period of high hydroclimatic variability. Purple:
Period of marine B. Concentration of Ti (cps) as indicator of hydroclimatic variability. C. Demographic variability across time in the Caribbean coast of Colombia, inferred from calibrated C14 dates of archaeological remains. Arrows indicate peaks in estimated human population size. Dashed line represents a period of high hydroclimatic variability and low human population size from ~4800 – 3000 yr BP
B.
C.
A.
P1 P2 P2
Participatory Historic Reconstruction
Focus group discussions at both stilt-house communities of CGSM, Nueva Venecia and Buenavista, provided insights into the temporal, spatial and contextual aspects of the degradation process during the last decades (Fig. 5 A,B)
The discussions emphasized the main changes in the western region of the CGSM (Fig. 5B), namely the mangrove systems within the Pajaral Complex (PC) and the floodplain of the Magdalena River, as these areas concentrate most of the daily activities of the inhabitants of the stilt-house communities. The starting point for the timeline reconstruction is 1947.
Agriculture and Extractive activities (Q 1-2) Crops and cattle ranching
Participants hold collective memory of active trade with land-based towns on the Magdalena riverbank, especially with towns of Remolino, Santa Rita and Salamina. Most of the commercial arrangements were based on exchanging dried fish for a variety of products including: corn, cotton, pumpkin, cassava, papaya, melon, watermelon, sugar cane local liquor (tapetusa) and chicken. Participants describe these exchange activities ceased to exist in the early 1990’s when significant changes in land-use patterns became widespread along the Magdalena floodplain, affecting crop diversity and navigation routes in the area.
By the decade of the 90’s and early 2000’s deliberate flooding pulses by landowners, extended north from Caño Remolino to Caño La Ceja for rice crops expansion (Fig. 5 A,B), which are highly water and space intensive. Interestingly, one council representative who participated in the focus group of Buenavista, argues that environmental authorities have modified La Ceja stream to benefit rice growers by installing water gates that promote flooding and freshwater accumulation in the crops (see below).
Cattle ranching remained small and scattered in the area until early 2000’s, but during the last 2 decades buffalo ranching has expanded throughout the Magdalena floodplain. Participants describe it as highly water intensive due to the need of buffalos to live in highly flooded areas.
Thus, buffalo herds have completely invaded marshes like Ciénaga de la Aguja in the south (Fig. 5B), and created water management issues like the setting of a large water gate in Renegado stream, which regulates freshwater flows from the Magdalena River.
According to participants, without such gate freshwater from the river would flow east through Renegado stream, flooding the invaded marshes and eventually would flow north into the Pajaral Complex. However, gate management is altering hydrologic connectivity as it allows just the enough flow to keep buffalo marshes flooded, but not enough to drain into the rest of the complex; therefore, compromising the access to freshwater for people in the stilt-house communities.
Mangrove use and exploitation
White mangrove (Laguncularia racemosa) forests around Pajaral Complex have been used as a subsistence wood source for construction of the stilt houses, canoes and fishing gear, and recently as a source for income given that a single mangrove pole costs about 240.000 Colombian peso (~64 USD).
However, the most significant exploitation of mangrove forests took place form 1970’s until mid 90’s, for commercial purposes by a company knows as “Láminas del Caribe”. According to participants, the company operated continuously throughout 20 years particularly around Ciénaga de La Ahuyama and La Luna (Fig. 5B) The company followed a staggered strategy, cutting the largest trees within a small area, then exposing poles to sunlight to accelerate the drying process of the wood, and finally moving on to the neighboring area to repeat the process.
After 2-3 months of this cycle, they came back to collect the dried poles and took them to their warehouse. For this purpose, they built an artificial stream in 1976 -78 known as “Caño del Indio”, nowadays “Caño Cobado” (Fig. 5A). When cutting, they used chainsaws to carve the trees as close as possible to the water surface, leaving almost no exposed trunks and thus hampering the tree’s ability to regrow. This modus operandi led to sustained degradation of (L.
racemosa), along with important landscape modifications in waterflows Massive Biodiversity Mortalities (Q 3-4)
Mangrove mortalities
Participants describe three significant massive mortality events during the last 50 years. There is consensus hypersalinization of water and soils has been the underlying factor in all the events, nevertheless participants highlight several -and compound- triggers for the die offs.
The earliest die off on which the community agreed happened in the early 60’s, very soon after the construction of the coastal highway, which marked the onset of the disconnection between fresh and marine water flows (Fig. 5A). The earliest sightings of local mortalities happened around Conchal and Alfandoque marshes southwards PC, and in La Ahuyama and Luna marshes northwards (Fig. 5B). In an attempt to control these early mortalities, environmental authorities established protected areas in the CGSM called “Vía Parque Isla de Salamanca”(1964) aimed at controlling the die-offs in the north, including La Ahuyama; and
“Santuario Fauna y Flora Ciénaga Grande Santa Marta” (1977) for southern forests. Despite, mangrove mortalities continued to spread along Pajarales Complex and by 1971 participants describe significant mangrove reduction in La Ahuyama, even before the onset of commercial extraction in the area.
Nonetheless, the most severe mortalities recalled by the participants happened between 1995- 1997, when droughts triggered by recurrent ENSO events (1992, 1995) compounded with acute
alterations to hydric connectivity, increased salinity to unprecedented levels such that fishermen describe entire mangrove forests disappearing whilst exuding salt grains from leaves and trunks, forming a white crust. Ever since, and despite constant management interventions, mangrove forests have continued to decay and while showing slight recovery periods, these are not enough to promote sustained regrowth because most trees die at the onset of every dry season.
Continuous mortalities have occurred within the PC such as that of 2010 which resulted from an extreme 3 year drought period, 2007-2010, that connected ENSO events on both years (Fig.
5A). The most recent mangrove mortalities took place in 2015, near Nueva Venecia and Buenavista, which according to participants, the area has experienced delays in mortality prevalence because of its location in the buffering zone between the aforementioned protected areas, thus it has better endured the degradation process than other areas.
Fish Mortalities
Participants agree 1995 as the year with the largest massive mortality of fishery resources in the history of CGSM (Fig. 5A). Although seasonal mortalities were common due to changes in water and wind conditions and they are acknowledged by participants as naturally occurring, 1995 mortality event has been the only one happening both within the Pajaral Complex and the main lagoon of the CGSM simultaneously,affecting freshwater and marine species, some of which fishermen had never recorded before (e.g: 30-40kg groupers and snappers).
Interestingly, participants highlight this event as a breaking point for fisheries resources in the ecosystem because some of the most valuable and, once, abundant species such as the New Granada sea catfish (Notarius bonillai) did not bounce back and were replaced by the Tarpon (Megalops atlanticus). The reason, according to participants, for the shift in fish dominance was the ability of the latter to adapt its diet to become a scavenger and feed on the decaying organic matter that remained after the die offs.
Mortalities started in La Rinconada, the oulet of Caño Hondo in the northwestern region of the CGSM and rapidly expanded eastward into the main lagoon and westwards into the inner streams and lagoons of the Pajarales Complex.
Recent die offs occurring in 2018-2019 took place between the outlets of Fundación and Aracataca Rivers in the south east (SE) of the CGSM, and participants associate them with chemical pollution coming from the extensive palm crops which use highly toxic substances for pesticides and fertilizers that are then washed into the main lagoon during the rainy season.
The main species affected during these events were Lebranche mullets (Mugil liza), different species of Mojarra (Diapterus spp., Eugerres plumieri), and even manatee (Trichechus manatus), tortoises and alligators.
Impacts of degradation on local communities (Q 5-7) Reduction of fisheries diversity and productivity
Marine fisheries are the most affected resource in the CGSM after 60 years of continuous degradation. Participants link this effect directly with the construction of the coastal highway which interrupted hydrologic connections, leaving just one inlet: “Boca de la Barra”, where high siltation prevents the entrance of marine fish species to the CGSM, including jacks, pompanos, grunts, sharks, and catfish from the genus (Cathorops spp.). Participants describe significant effects for the reproductive biology of the species that move between the sea and the mangrove during different periods of their cycle, including high value species for fishermen.
For example, the last record of sharks in the CGSM, according to participants, happened between 2008-2010. The loss in marine species diversity has happened over the 50 years following the highway construction, but participants agree species were critically reduced after the 1995 mortality.
Along with species diversity, fisheries productivity also decreased. Until the end of the 80’s decade, there was an active fish trade market and catches in the CGSM supplied enough volumes for subsistence and commercial revenue. Participants describe daily shipments of up to 15 canoes (known as “traveling canoes” (Table 2) transporting fish and invertebrate catches from stilt-house communities towards a landing site form where fresh product was shipped to Barranquilla’s market place, or exported internationally. One participant claims having caught 2 tones of M. liza in a single fishing trip. By 1992 productivity had decreased significantly and many fishermen had already quit fishing before the 1995 mortality event. That moment coincided with the privatization of Barranquilla’s market place, thus the community lost its main trading point and it was replaced by smaller and more disconnected markets, which significantly lowered volume demands. Nowadays just 1 canoe travels daily, and total catch fits inside a single cooler.
Participants agree the crisis in fisheries resources came before the 1995 mortality event, not afterwards, and highlight the installation of several natural gas pipes along the coastal highway (km 15- 32) (Fig. 5B), as one of most impactful factors driving frequent fisheries mortalities during the early 90’s. In 1993 the company used to blow up to 10 parallel dynamite 1 km lanes that produced frequent biodiversity killings and reduced the access to important fishing grounds for local communities.
To cope with the fisheries crises in the CGSM, participants describe emergence of destructive practices. The crab (Callinectes sp.) (Fig. 5 A,B) fishery became highly prevalent around year 2000, and is highly impactful, because the iron rings that support the fishing nets rapidly dissolve into the water, releasing ferric compounds that affect water quality and kill surrounding biodiversity. Additionally, traps are left in place for several days capturing high proportions of
bycatch larvae and trawled between fishing spots which, according to participants, disturbs benthic meiofauna around the trap including egg bundles.
Social and ecological impacts of sedimentation
Participants express strong disagreement with the management strategies undertaken by environmental authorities, as they highlight narrow understanding of the issues and thus limited solutions. Specifically, they refer to dredging works that aim to restore/maintain freshwater flow, but simultaneously promote increased sedimentation and formation of extensive beaches that displace water while sedimentation frontier advances towards the stilt towns. At the outlet of the Aguas Negras stream, accretion frontier has advanced ~2 km in the last 4 years, since dredging upstream Aguas Negras became more frequent. Participants describe loss of entire fishing grounds of tarpon (M. atlanticus) southwards Aguas to sedimentation, with significant economic impacts for fishermen. There is a strong emphasis on the systematic dismissal of local knowledge by the environmental authorities, as participants claim they approach the interventions with fixed strategies for solving the problems which ignore important details about the ecosystem functioning and therefore amplify the unintended consequences. Over time, this dynamic developed profound tensions and disagreement between local communities and environmental authorities, leading to poor legitimacy of management interventions.
Reference (baseline) descriptions of CGSM
Participants expanded on the characterization of some regular dynamics of the social-ecological system of CGSM before the major landscape transformations occurred. This category was not included as a separate topic during the focus group discussions, instead it emerged during the round of inductive coding as a recurrent topic.
Seasonality is the most important feature in collective memory when referred to natural dynamics of the CGSM. Participants describe clear effects of precipitation seasonality on species configuration, movements, and socioeconomic effects. An interesting example is the abundance of Equisetum plants during the rainy season in Alfandoque marsh, whose roots trapped large stocks of common snook (Centropomus undecimalis). At the onset of dry season high salinity conditions became dominant, causing widespread Equisetum die offs, and massive fish escapes into the main lagoon, yielding high volume catches for fishermen that significantly improved fishermen livelihoods at least twice a year, since it is a high market value species.
Nowadays, plants do not establish due to permanent hypersaline conditions in the water, and trawl fishing which kills the seeds; therefore accumulation and further escape of fish does not take place, decreasing fishing livelihoods. A similar dynamic was reported for “bocachico” fish (Prochilodus magdalenae), a freshwater species that accumulated in La Ahuyama and Alfandoque (Fig. 5B).
When describing these ecological cycles participants evidence clear local ecological knowledge because the warning signal for upcoming changes in salinity was the behavior of osrpey (Pandion haliaetus), that changed its flying pattern by doing circles on the same place,
identifying early weakened fish. This happened around 2-3pm when wind blew landward and saltwater streams further increased salinity. After observing the same pattern for several days in a row, fishermen knew it was time to head towards the main lagoon to fish. These “escapes”
were common until 1984-85 but ceased dramatically due to permanent hypersaline conditions that hindered the establishment of large “bocachico” stocks and Equisetum plants. Last record of this migration was during 2010-2011 La Niña phenomenon, which decreased salinity in the short term, but participants argue that stocks migrated back to the river even before brackish water came, as fish already “knew” low salinity conditions wouldn’t remain long (Table 2).
Water flow modifications
Participants agree that the environmental authorities responsible for the management interventions in the CGSM have done significant transformations on the hydrological connectivity and functionality of the system, often in connection to private users. This topic emerged when discussing “other changes” the community considered important and had not been explicitly addressed through the questions.
Participants gave detailed descriptions of 2 very recent channel opening (Fig. 5B )s not reported in the official cartographies. According to participants, a channel was built between 2017-2020 by the local environmental management agency, which, they argue, has not improved any of the challenges faced in the region. Participants describe it starts at “La Ceja” stream a tributary of the Magdalena River, and flows east connecting with “Márquez” stream and “Mendegua”
marsh. Then flows eastwards and crosses “Cobao/Indio” stream and ends in “La Caleta”, around km. 28 parallel to the coastal highway, with a that prevents water flow into the main lagoon of the CGSM. Participants describe it is “dead-end” and does not serve the purpose of delivering freshwater into the lagoon because as water flows backwards (East-West), fish populations cannot establish and they usually die overnight probably, to hypoxia.
Interestingly, participants describe that source stream “La Ceja” has been modified with closures and gates that concentrate water in rice plantations so it is wide and has strong water flow at the beginning, but as the channel extends eastwards, towards the stilt-house communities, it becomes weaker and unable to supply them freshwater. Therefore, participants argue environmental authorities are actively benefiting private actors over community’s wellbeing through waterflow modifications.
Community members in 2017 opened an artificial outlet from “Aguas Negras” stream, as a response to lack of freshwater and frequent fish mortalities around the stilt-house communities (Fig. 3AB). Following the installation of a gate up stream that regulates the flow, communities’
members organized collectively and opened an artificial outlet upstream the gate. It was done without any official planning or consultation processes, and even faced police confrontations but at the end the community pursued their goal. Another motivation was recovering the seasonality in salinity dynamics because the area has been continuously hypersaline for several years, but this did not happen. Instead, participants say that low salinity conditions dominate
now the area which is favorable for their daily life, but acknowledge the shrimp fishery has been lost to this change.
.
Fig. 5. A. Participatory temporal reconstruction integrating TEK from stilt-house communities of Buenavista and Nueva Venecia, highlighting main changes in the CGSM.
Fig. 5.B. Participatory reconstruction integrating TEK from stilt-house communities of Buenavista and Nueva Venecia highlighting the spatial distribution of main changes in the CGSM.
