Farm Household Head Characteristics and Perceptions of Factors Related to Sustainable Management of Fogera Wetlands in Five Kebeles of South Gondar, Ethiopia

Abstract

Policies regulating common-pool resources (CPR), such as wetlands, should take into account community socio-economic realities and community perceptions. The study aims to examine whether policies regulating common-pool resources, such as wetlands, should take into account community socio-economic realities and community perceptions. Specifically, it characterized Ethiopian farm households in the Amhara Region and household-head perceptions of six factors related to sustainable wetland management. Surveys were given to 385 respondents from five administrative units or kebeles (kebele is the smallest administrative unit). Focus-group discussions followed. All respondents were Orthodox Christians, and men constituted 93% of household heads. Kebeles differed significantly for household-head age and education, but overall education level was low. Most households had three or four children. Overall, the mean farm area was 0.89 ha, but within kebeles, mean areas ranged from 1.35 to 0.80 ha. Kidest Hana was the most successful kebele for farm size and yield. Significant differences existed among kebele farms for percentages of harvested crops sold, suggesting some were less food-secure. Data suggested male household-head farms had twice the area of female-head farms, older household heads had more land than younger ones, heads with more education had more children, and married heads had more land than divorced or unmarried heads. None felt that people outside their kebele should access their natural resources. Overall, 85% thought land ownership was communal, and 15% thought it was private. A plurality (44%) did not know who oversaw land management. None knew of any organizations working on sustainable wetland management, but most (75%) wanted to attend meetings on the subject. No farm used mechanized agriculture. Most (87%) had not abandoned alternative crops to produce more rice. Therefore, as a conclusion, every effort needs to be taken to address this socio-ecological and development challenge that faces Fogera wetlands and larger national and international impacts to achieve balanced and sustainable development. Hence, our study highlights the need for better education, leadership, and policies regarding sustainable wetland management.

1. Introduction

Wetlands are among the world’s most biologically diverse systems and provide a number of critical social, economic, and environmental ecosystem services [1,2,3,4]. For example, they play vital roles in filtering pollutants and controlling water flow [5] and provide diverse sources for food and employment [6]. More recently, wetlands have provided new economic opportunities through recreation and ecotourism [7]. In many societies, they hold important aesthetic, cultural, educational, and spiritual values. Because of the essential ecosystem goods and services that wetlands provide to support the livelihoods of so many, policies should be in place to ensure their sustainable management and to prevent degradation [5]. Unfortunately, wetlands have been lost and degraded for thousands of years due to human activity, and currently, wetland degradation is occurring worldwide at alarming rates [8,9]. Primary indirect drivers of wetland degradation include population growth and encroachment, and land and economic development. Direct drivers include infrastructure development, such as industrial enterprises and settlements, agricultural expansion of crop and grazing lands, water withdrawal and diversion, pollution, over-exploitation of natural resources, including fish and wildlife, and invasive species [8,10].

In Ethiopia, wetland flora has provided diverse economic goods for centuries [4], including food, fiber, fuel, and medicinal and dietary supplements [11]. Papyrus from Lake Tana, the largest lake in Ethiopia and the source of the Blue Nile, has been used to make fishing boats for hundreds of years [12]. Wetland flora in Ethiopia is also used for agriculture, including large-scale rice and vegetable production, fishing and pisciculture, and livestock pasture, especially during the dry season [1,13]. Dense wetland vegetation can also be harvested as fodder for cattle and sheep and for use as raw material for thatching, fencing, and artisanal crafts that have important ceremonial roles.

There are already examples of Ethiopian wetlands that have been destroyed, most notably Lake Haromaya in the east-central Oromia Region of Ethiopia [14,15]. Abebe et al. [16] found that the lake’s demise negatively affected the livelihoods of farmers who depended upon its ecosystem services. More than 70% of Ethiopia’s population is employed in the agricultural sector. Those dependent upon Lake Haromaya wetlands lost income from a reduction in overall crop and livestock production, including decreased production of vegetables, chat, irrigated crops, and inter-cropped fields. Abebe et al. [16] recommended alternative policies to reverse the lake’s demise, and efforts towards that end are now underway.

Agriculture provides the greatest source of income in the Fogara wetlands, which are located in Ethiopia’s South Gondar zone, part of the Amhara Region (Figure 1) [17]. Agricultural production in South Gondar is constrained by numerous factors, including a shortage of arable land, poor soil fertility, and land and water degradation. As is often the case in Sub-Saharan Africa, increasing demographic pressure has forced resource-poor farmers to meet food needs by reducing fallow periods and expanding agricultural production into marginal lands or sensitive forest and wetland areas rather than by increasing yield. This expansion has caused major ecological damage [18,19].

In Ethiopia in general, and in the Fogara wetlands in particular, different stakeholders among the local population, as well as governmental and non-governmental organizations, use natural resources in uncoordinated and unregulated manners. This has led to a Tragedy of the Commons [20], a well-known phenomenon in which poor government and management of common-pool resources (CPR) lead to their degradation. Better approaches to wetland CPR management and governance are needed [21,22]. Addressing the lack of appropriate policy and effective institutional oversight is key to reversing the degradation of Ethiopian wetlands [23]. Kassa [15] concluded that, in order to prevent environmental destruction and to rehabilitate depleted wetland resources such as Lake Haromaya, it is imperative to pay particular attention to both the action and inaction of social entities and to set restorative policies and programs that maintain a balance through interactions with community stakeholders as well as with officials.

Systems-based research, which links ecosystem management to economic, institutional, governance, and socio-cultural realities, is of central importance when analyzing land degradation and finding pathways towards sustainability [24,25,26,27]. Sustainable management of agroecosystems, and, in particular, those that rely on CPR use, should be studied within the context of the entire system, including farmer communities and their perceptions regarding CPR and other stakeholders as components of those systems. Socioeconomic attributes of farming communities should be taken into account, as well as household perceptions towards factors related to sustainable land management. All of these are required to gain active community participation in problem definition with a view toward testing potential solutions to mitigate land degradation [25]. Therefore, studies on policies regulating common-pool resources (CPR) such as wetlands should take into account community socio-economic realities and community perceptions. Thus, this study was motivated to address the previously mentioned research gaps by characterizing farm households for a number of socioeconomic attributes, as well as household-head perceptions towards six factors that influence sustainable management of the Fogara wetlands along the eastern shore of Lake Tana, and providing possible recommendations based on those characteristics and perceptions.

Conceptual structure pro-environmental behavior (PEB) is defined as any active response to the preservation or conservation of natural resources, according to Eilam and Trop [28]. As a result, individuals who exhibit PEB engage in responsible acts that do not harm the environment. The current study focuses on wetland resources, which are under growing threat and require PEB from those who use them in order to be sustained. However, because human behavior can be influenced by a variety of factors, it is a hard task to attempt to fully comprehend it [29]. Academics have long developed many frameworks and models in an attempt to comprehend behavior in the environment.

The Theory of Planned Behavior (TPB), put forth by Icek Ajzen, is a behavioral paradigm that has the backing of multiple empirical studies in a variety of environmental situations. TPB serves as the foundation for the current study’s understanding of the variables thought to influence (pro-) environmental behavior. The investigation, however, did not only stay inside this framework. Giford and Nilsson [30] pointed out that it is challenging to adequately account for variation in PEB in this model; hence, more personal and social components could be added to the model. Ajzen [31] further confirmed that finding pertinent background elements in the behavior domain of interest can be used to supplement our understanding of the determinants of TPB.

PEB to utilize and manage wetland resources is postulated in this study to be a consequence of environmental attitude, involvement intention to wetland management, and knowledge about wetlands and their ecosystem services—all of which Blankenberg and Alhusen [32] refer to as “psychological factors.” Prior to PEB, one must have environmental knowledge or an understanding of the environment. In this sense, Adem [33] and Lawson [34] argued that greater awareness of the environment or resources will lead to a greater appreciation of their values and the development of more ecologically conscious conduct. According to Ajzen [31], the TPB also made the assumption that human beings usually behave in a sensible manner such that they take account of available information. In addition to environmental knowledge, it is assumed that people’s PEB is influenced by their attitude or their favorable or unfavorable assessment of engaging in a certain action of interest. Ajzen [31] asserts that a behavior is most robust when it receives positive feedback. However, the TPB acknowledges that the presence of another, more immediate factor, the intention to engage in that activity, mediates (attenuates) the influence of knowledge and attitude on a particular behavior. It is believed that participation intention, or the desire to engage in an activity, captures the driving forces behind conduct.

Therefore, according to the Theory of Planned Behavior (TPB), the most significant immediate determinant of an action is its intention to be performed; the stronger the intention, the greater the likelihood of performance [31,35]. The theory put forth by the TPB on the mediating role of intention appears to be at odds with empirical research [36,37], which showed that knowledge and attitude have a significant direct impact on behavior. The current study attempted to investigate the direct influence of attitude and knowledge on behavior as well as the intermediary role assigned to intention in light of these seemingly contradictory principles. Aside from the assessments of environmental knowledge, attitude, and intention, additional contextual factors are anticipated to impact the way wetlands are used and managed.

Ajzen [31] acknowledges the possible significance of these aspects by stating that various background circumstances impact an individual’s conduct. But it is hard to tell which elements should be taken into account because there are so many that could be important. Therefore, empirical data can serve as a guide when selecting pertinent aspects for research of an interest in behavior. In their study, Getacher and Tafere [38] attempted to divide background influences into three categories: economic, biophysical, and demographic. The characteristics (demographic and socioeconomic) of individuals or HHs and the biophysical/environmental parameters (distance from the wetlands location) are considered background factors in the present study that may have an impact on behavior regarding wetland management (Figure 2).

Accordingly, the purpose of this study was to characterize farm households for a number of socioeconomic attributes, as well as household-head perceptions towards six factors that affect sustainable management of the Fogara wetlands: (1) land ownership, (2) responsibility for resource utilization, (3) responsibility for development, (4) ways to support sustainable management, (5) intensity of rice production systems, and (6) loss of crop diversity. Based on the purpose of this study, the overall question was as follows: what are the farm households’ characteristics of socioeconomic attributes and perceptions of factors influencing the sustainable management of Fogera wetlands? The primary objective of this research was to investigate the socioeconomic characteristics of farm households and their perspectives on the factors affecting the sustainable management of the Fogera wetlands. The following are the specific objectives of study: (1) to determine the degree to which the common pool wetland resources’ land ownership is understood (2) to assess how households view their role in resource usage, (3) to investigate how households view their share of the burden of development, (4) to comprehend the family system or methods for assisting with sustainable management, (5) to examine how intensively rice cultivation techniques affect wetlands, and (6) to determine whether crop diversity is declining.

2. Methodology

This study followed a quantitative (descriptive) and qualitative approach. Descriptive research is a quantitative method that focuses on describing the characteristics of a phenomenon rather than asking why it occurs. Doing this provides a better understanding of the nature of the subject at hand and creates a good foundation for further research. In addition to this, the focus group discussion is frequently used as a qualitative approach to gain an in-depth understanding of social issues. So, it helps for research that the method aims to obtain data from a purposely selected group of individuals rather than from a statistically representative sample of a broader population.

Study Area. The Fogera wetlands, the most prominent wetlands in the Amhara region, are situated in the northern Ethiopian highlands between 11°46′ and 11°59′ N and 37°33′ and 37°52′ E. Elevation ranges from 1774 to 2410 m, and mean annual rainfall is 1216 mm, with a bimodal distribution. Most of the Fogera wetlands flood during the summer rainy season and are dry during the winter season. The Gumera and Rib rivers provide water year-round and deposit clay and silt eroded from uplands through seasonal flooding. Rice is cultivated in the floodplain lowlands where there is sufficient water [39]. In higher elevations, and when flood water recedes, fields are irrigated using diverted water from the two rivers [40]. While the area is currently known mainly for rice production, many other crops are also grown in the surrounding areas.

In recent years, wetland hydrology has been significantly altered through settlement encroachment, installation of eucalyptus plantations, and construction of dams and other infrastructure for water diversion and drainage to promote intensive, irrigated agriculture. Intensive agriculture has been associated with additional environmental challenges, such as increased use of pesticides and fertilizers [41]. Moreover, during the dry season, the wetlands are increasingly subject to more intensive grazing by livestock.

Sampling Procedure. The study area consisted of 10 Woredas or districts, which are divided into 137 Kebeles or administrative districts. Five Kebeles in the Fogera Woreda were selected for household surveys (Figure 1). The number of households chosen for questionnaires in each Kebele was equal to approximately 3.8% of the total number of households (

Table 1. Stratified sample population and total household number for sampled Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia.

No	Name of Kebele	Total Household Number *	Sampled Households **
1	Nabega	2283	87
2	Shina	2136	81
3	Kidis Hana	1790	69
4	Shaga	1515	57
5	Wagetera	2398	91
Total		10,122	385

* This one asterisk indicated the total household number is found from the local Kebele administrative offices. ** These two asterisks indicated in the table can show the sample household in each kebele is found through calculation based on the formula presented in the above.

). Individual households within a Kebele were selected using a systematic random sampling method described by Yemane [42]. This method uses the following formula to select the number of households from each administrative unit:

$$n={\displaystyle \frac{N}{1+N({e)}^2}}$$

(1)

where n = total household sample number, N = total population, and e = error term, which we set at 0.05. Thus, our total household sample number for all five kebeles in the study was determined from

$${\displaystyle \frac{\mathrm{10,122}}{1+\mathrm{10,122}({0.05)}^2}}=385.$$

The formula for determination of the needed number of households in an individual Kebele was

$$S={\displaystyle \frac{TNH \times TSH}{THS}}$$

(2)

where S is the number of needed household samples for each Kebele, TNH is the total number of households in a Kebele, TSH is the total number of sample households, and THS is total households of all sample Kebeles. Thus, for Nabega Kebele

$$\frac{2283\ast 385}{\mathrm{10,122}}=87.$$

Data sources and collection methods. We used both primary and secondary data sources. Primary data were collected through (1) questionnaires submitted to heads of households, (2) interviews with heads of households, and (3) focus group discussions (FGD) (Figure 3a).

Household contact information was provided by local Kebele administrative offices. Household heads were asked to provide age, gender, marital status, education, and religion of the household head, as well as information on number of children, the area farmed, and on crops, including species, yield, and amount consumed and sold. Household heads were then questioned to gain understanding of their perceptions of (1) who owns the land (public, private, or community); (2) who is responsible for governing or utilization of wetland resources (community, Kebele leader, development agency, cooperative, Woreda wetland association, or do not know); (3) knowledge of organizations working on wetlands management [nongovernmental organizations (NGOs), governmental organizations (GOs), or do not know); (4) preferred method of promoting sustainable use of wetland resources (expose illegal users, participate in regular meetings on wetland use, or participate as a member of local committee promoting sustainable wetland use); (5) whether non-mechanized (Figure 3b) or mechanized rice production system was used; and, finally, (6) whether any alternative crops had been taken out of production due to the introduction of rice, as had been found by Abebe et al. [16] for Lake Haromaya. Data from interviews and local government reports were cross-tabulated among Kebeles using the XTAB module of SYSTAT v. 13.1. Standard deviates were calculated from (Observed Value-Expected Value)/SQR (Expected Value) and used to determine deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele, from chi-square tests of association, using SYSTAT version 13.1 (Grafiti LLC, Palo Alto, CA, USA). Analysis of variance was performed when it existed for quantitative data, and significant differences among Kebeles were tested using Tukey’s hsd in SYSTAT 13.1.

Regarding participants in focus group discussions and key informant interviews, the selection was made based on the purpose of the study. In other words, participants for focus group discussions and key informant interviews were selected purposively complement the quantitative data collected through household surveys. Accordingly, participants with relevant knowledge and experiences were selected purposively to participate in focus group discussions and key informant interview sessions. A total of five focus group discussions was administered for the study. Six participants (two female and four male) participated in each kebele, and the participants were from different stakeholders in the study kebele (the stakeholders such as kebele administration leader, cooperative leader, youth, women, traditional and culture groups. (“Equb” “ ̂እቁብ̃/ ̂ቁብ̃” Amharic A rotating credit association in which each of its members contributes money periodically and the amount collected at each period is provided to one of the members often on a lottery system.), “Edir” (Edir “እድር” Amharic Edir is a traditional community organization in different parts of Ethiopia whose members assist each other during the mourning process.), and Senbetie, “ሰንበቴ”, is a religion that dictates that groups of people drink local alcohol (local name: “Tila”) every month on Sunday, prepared in turns, for their soul)). Moreover, one female and two males interviewed farmers/key informants who were also selected based on their life and work experience in the wetland areas and expertise working in this field. Selected households were interviewed in each kebele.

3. Results

Kebele Household Profiles. There were similarities among Kebele household heads in terms of religion, gender distribution, and other characteristics, but there were also important differences. All 385 respondents from the five Kebeles identified as Orthodox Christians. Household heads were mostly men, with 94% headed by men and only 6% by women (

Table 2. Gender distribution among household heads in five Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of kebele. p-values indicate significance of differences among kebeles. Tables were generated using Systat 13.1’s XTAB module.

				Standard Deviates
Kebele	Male	Female	Total	Male	Female
Kidest Hama	65	4	69	0.015	−0.06
Shaga	76	5	81	−0.018	0.073
Shena	54	3	57	0.055	−0.22
Wagetera	85	6	91	−0.061	0.242
Nabega	82	5	87	0.022	−0.087
Total	362	23	385	p-value = 0.998

). Kebele household heads were also similar in marital status, although Shena household heads tended to have a lower divorce rate (

Table 3. Marital status of household heads in five Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia. Positive r negative signs are used to indicate deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Tables were generated using Systat 13.1’s XTAB module.

					Standard Deviates
Kebele	Married	Unmarried	Divorced	Total	Married	Unmarried	Divorced
Kidest Hama	65	3	1	69	0.015	−0.22	0.334
Shaga	76	4	1	81	−0.018	0.001	0.173
Shena	54	3	0	57	0.055	0.112	−0.77
Wagetera	85	5	1	91	−0.061	0.24	0.056
Nabega	82	4	1	87	0.022	−0.142	0.101
Total	362	19	4	385	p-value = 0.999

).

Household heads differed significantly (p = 0.000) among Kebeles for age distribution (

Table 4. Household age distribution among household heads in five Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence.

Kebele	Under 20	20 to 29	30 to 39	40 to 49	50 to 59	60 to 69	70 to 79
Kidest Hama	0	12	8	21	10	18	0
Shaga	0	23	18	12	3	25	0
Shena	0	23	17	6	1	10	0
Wagetera	1	33	24	9	5	19	0
Nabega	0	24	23	12	12	13	3
Total	1	115	90	60	31	85	3
Standard Deviates
Kebele	Under 20	20 to 29	30 to 39	40 to 49	50 to 59	60 to 69	70 to 79
Kidest Hama	−0.423	−1.897	−2.024	3.125	1.885	0.709	−0.733
Shaga	−0.459	−0.243	−0.215	−0.175	−1.379	1.683	−0.794
Shena	−0.385	1.448	1.007	−0.967	−1.676	−0.729	−0.666
Wagetera	1.571	1.116	0.591	−1.376	−0.86	−0.243	−0.842
Nabega	−0.475	−0.39	0.59	−0.423	1.887	−1.416
	p-value = 0.000

and Figure 4). Kidest Hana had fewer household heads in the age group of 20 to 40 years but more in the age group of 40 to 69 years. Shaga had more household heads aged 60 to 69 years old compared to other Kebeles and fewer that were 50 to 59 years old. Shena had more household heads in the 20- to 39-year-old age groups but fewer aged 40 and older. Similarly, Wagetera household heads tended to be younger, particularly in the under-20 to -29 age groups, and to have proportionally fewer people in older age groups. Nabega tended to have more heads in the 50- to 59-year-old age group and fewer in the 60- to 69-year-old age group but was the only Kebele to have household heads (three) in the 70- to 79-year-old age group. Kidest Hana had the highest mean age of 51.2 years, and Shena had the youngest mean age of 42.9 years. The box-and-whisker plot of age distribution among Kebeles (Figure 4b) further illustrates differences among Kebeles for age distribution.

There were also tendencies for the level of education differences among Kebele household heads (

Table 5. Education level among household heads in five Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

Education					Standard Deviates
Kebele	Illiterate	Literate	Primary School	Total	Illiterate	Literate	Primary School
Kidest Hama	60	8	1	69	0.546	−1.199	0.11
Shaga	71	10	0	81	0.661	−1.139	−1.026
Shena	47	10	0	57	0.119	−0.021	−0.86
Wagetera	69	20	2	91	−0.553	0.98	0.753
Nabega	65	20	2	87	−0.655	1.182	0.819
Total	312	68	5	385	p-value = 0.293

). The p-value of 0.293 (Table 4) suggests that the model of independence between Kebele and education did not deviate quite as strongly from independence as that for Kebele and age. Even so, Wagetera and Nabega tended to be the best educated, with proportionally more literate household heads than other Kebeles. Wagetera household heads tended to have more primary school education. Kidest Hana and Shaga tended to be the least educated, with higher proportions of illiterate household heads and lower proportions of literate household heads. None of the 385 household head respondents from the five Kebeles had any secondary or tertiary education.

There was some evidence that Kebele and the number of household children were not independent, with a p-value of 0.850 (

Table 6. Number of children in households in five Kebeles of the Fogera Wetlands of South Gondar, Amhara Region, Ethiopia. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

Number of Children
Kebele	1	2	3	4	5	6	Total
Kidest Hama	0	2	41	25	0	1	69
Shaga	2	2	44	32	1	0	81
Shena	1	1	35	20	0	0	57
Wagetera	1	2	50	35	3	0	91
Nabega	1	2	49	31	4	0	87
Total	5	9	219	143	8	1	385
Standard Deviates
Kebele	1		2	3	4	5	6
Kidest Hama	−0.947		0.305	0.279	−0.124	−1.197	1.939
Shaga	0.924		0.077	−0.306	0.349	−0.527	−0.459
Shena	0.302		−0.288	0.453	−0.255	−1.088	−0.385
Wagetera	−0.167		−0.087	−0.245	0.206	0.807	−0.486
Nabega	−0.122		−0.024	−0.069	−0.231	1.63	−0.475
	p-value = 0.850

). Data suggest that most household heads had three to four children, but there was a slight tendency for Kidest Hana and Nabega households to have larger (>4 children) families.

Kebele Farms. The total area farmed by the 312 households in the five Kebeles was 340.8 ha. Of this amount, 263.3 ha were sown to rice, 36.0 ha to grass pea, 21.8 ha to teff (Eragrastis tef), 8.5 ha to onion, 3.8 ha to tomato, and 2.5 ha to chickpea. Nearly all farmers grew rice during the rainy season, irrespective of Kebele (Figure 5). Depending upon the Kebele, 24–39% of farmers grew teff, and 16–52% grew grass pea (Figure 5). Onion, garlic, and tomato, which are typically grown during the dry season in fields distant from rice fields and require irrigation, were only grown by a small minority of farmers. Only a very small proportion of farmers grew chickpeas, and all of these were in either Shaga or Nabega kebeles (Figure 5).

There were significant differences (p = 0.000) among Kebeles for total farm size, with Kidest Hana having the largest mean farm size of 1.35 ha and Nabega having the smallest with a mean area of 0.80 ha (Figure 6). Kidest Hana farmers seemed to be the most successful compared to other Kebeles, with the greatest area farmed to rice (p = 0.000), grass pea (p = 001), onion (p = 0.001), and garlic (p = 0.000) (Figure 6). They tended to have the greatest area sown to teff as well (p = 0.135), but variability for teff area among farmers within Kebeles was high, as reflected by error bars (Figure 6). Kidest Hana farmers had by far the greatest yield for rice, or 3416 kg ha⁻¹, followed by Shaga, Wagetera, Nebega, Shena, and final Nabega, where mean rice yield was only 1884 kg ha⁻¹ (Figure 7). There were no statistically significant differences for yields of teff (p = 1.75) or grass pea (p = 0.242), again due to high within-Kebele variability (Figure 7), but nonetheless, the trend was for higher mean yields in Kidest Hana farms. Nabega had the greatest yields for onion (p = 0.000), while Nabega and Wagetera had the greatest yields for tomatoes (p = 0.000). Kidest hana had the greatest garlic yields (p = 0.000) and Nebega had the greatest chickpea yields (p = 0.000) (Figure 7).

In general, the cash crops tomato, garlic, onions, and, to a lesser extent, chickpea were sold at much higher proportions of 0.9 or more (Figure 8). Much lower proportions of the staple food and fodder crops (rice, teff, and grass pea) were sold, with values ranging from ~0.3 to ~0.7. Nonetheless, there were significant differences among Kebele farms for percentages sold for rice (p = 0.000), teff (p = 0.007), chickpea (p = 0.000), tomato (p = 0.016), and garlic (p = 0.002), but not for grass pea or onions (Figure 8). Kidest Hana sold the largest proportion, or 0.50, of its rice, followed by Shaga at 0.46, Wagetera at 0.38, Nabega at 0.35, and Shena at 0.32. In contrast, Kidest Hana sold the smallest proportion (0.41) of its teff, followed by Shena (0.43), Shaga (0.47), Wagetera (0.54), and lastly, Nabega, which sold the largest proportion (0.73). There were no significant differences for grass pea (p = 0.135) due to high within-Kebele variability, as reflected in error bars, but the trend was that the highest proportion (0.63) was sold by Nabega, followed by Kidest Hana (0.58), Wagetera (0.54), Shena (0.50), and Shaga (0.49).

Among the three Kebeles who did raise tomatoes, Nabega sold 0.96 of its produce, while Wagetera sold 0.93, and Shaga sold 0.88. Only two households produced chickpeas in Shaga, and they consumed the entire crop. In Nabega, where 11 households produced chickpea, 0.81 of the produce was sold. Kidest Hana sold the most or 0.97 of its garlic crop, whereas Nabega sold the least or 0.89 of its garlic crop. There were no significant differences among Kebeles for the proportion of onions sold, but mean values ranged from 0.80 for Nabega to 0.97 for Kidest Hana.

Overall, these trends may suggest that Nabega and Shena were less food-secure and had lower incomes, for they kept more of their rice, sold more of the relatively expensive teff, and raised more cash crops to sell for income.

When respondent data were pooled among Kebeles, interesting differences were found among groups, particularly for farm size. Male household heads had approximately twice the land area as women (Figure 9). Differences were highly significant (p = 0.000), with mean values of 0.45 ha for female household heads and 0.91 ha for male heads. A half-dozen males, shown as outliers in the box-and-whisker plot, had relatively large land holdings. Overall, males made up 93% of household heads. Moreover, older household heads tended to have more land than younger ones (Figure 10), with a handful of older farmers, likely men (Figure 10), having large holdings of two or more ha. A perhaps non-intuitive association existed between the number of children and level of education—those groups with more education tended to have more children than those with less (Figure 11). Married household heads had more land than those divorced or unmarried (Figure 12). Differences were highly significant (p = 0.00) despite the low numbers of divorced (only four) and unmarried (nineteen) household heads. Still, again, there was a half-dozen outlier married household heads, likely older men (Figure 9 and Figure 10), with large holdings of two or more ha. Mean areas were 0.91 ha for married heads (n = 362), 0.45 ha for unmarried heads, and 0.44 ha for divorced heads. There were no widowed household heads among respondents. Total farm size was significantly correlated with area cultivated to teff (p = 0.022) (Figure 13).

Household Head Response to Wetland Management Questions. As with household profile data, household head responses to questions concerning sustainable wetland management had both similarities and differences among Kebeles. Regarding perceptions on who owned the wetland, not a single respondent in any of the Kebeles felt that the land was publicly owned in the sense that anybody outside the Kebele could use its natural resources (

Table 7. Farm household head perceptions of Kebele households on ownership of wetlands. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

Standard Deviates
	Private	Communal	Total		Private	Communal
Kidest Hana	6	63	69	Kidest Hana	−1.363	0.574
Shaga	4	77	81	Shaga	−2.348	0.989
Shena	13	44	57	Shena	1.506	−0.634
Wagetera	11	80	91	Wagetera	−0.732	0.308
Nabega	24	63	87	Nabega	3.009	−1.267
Total	58	327	385	p-value = 0.000

). Overall, 85% of respondents thought wetland ownership was communal or a common-pool resource available to the Kebele, and only 15% thought the land was privately owned. But chi-square tests of association for Kebele and ownership showed clear differences among Kebeles for their perceptions of land ownership, with a p-value of 0.000. Shena and Nabega household heads tended to perceive more strongly that land ownership was private, while Kidest Hana and Shaga tended to perceive it more strongly as communal.

A plurality of 44% of respondents did not know who was responsible for overseeing land management (

Table 8. Farm household head perceptions of who is responsible for sustainable management of the wetlands. Colors are used to indicate positive (blue) or negative (red) deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

Standard Deviates
	Community	Kebele Leader	Development Agent	Do Not Know	Total		Community	Kebele Leader	Development Agent	Do Not know
Kidest Hana	17	9	14	29	69	Kidest Hana	0.501	0.199	−0.227	−0.297
Shaga	16	8	16	41	81	Shaga	−0.398	−0.6	−0.35	0.838
Shena	11	3	14	29	57	Shena	−0.407	−1.501	0.488	0.732
Wagetera	22	12	20	37	91	Wagetera	0.481	0.267	0.086	−0.538
Nabega	18	15	19	35	87	Nabega	−0.225	1.344	0.056	−0.586
Total	84	47	83	171		p-value = 0.803

). About 22% thought the community was responsible for land management, and an equal number thought a development agency was responsible. Only a minority, or 12%, thought that the Kebele leader was in charge. There were tendencies for Kebeles to perceive this differently, with Shena viewing the role of the Kebele leader as particularly weak, and Nabega viewing it as particularly strong.

Remarkably, respondents were unanimous about unfamiliarity with any governmental or non-governmental organizations working on the development or sustainable management of the wetlands. Irrespective of Kebele, age, level of education, or gender, all responded that they did not know of any organization devoted to wetland wellbeing.

Regarding perceptions on ways in which individuals might contribute to improved sustainable wetland management, by far the largest response (75%) was to attend meetings on the topic (

Table 9. Farm household head perceptions of ways they could participate in improving sustainable wetland management. Positive or negative signs are used to indicate deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

	Expose Illegal	Participate in Meetings	Committee	Total		Expose Illegal	Participate in Meetings	Committee
	Users		Member			User		Member
Kidest Hana	10	52	7	69	Kidest Hana	0.347	0.004	−0.375
Shaga	11	61	9	81	Shaga	0.148	−0.002	−0.152
Shena	7	42	8	57	Shena	−0.148	−0.143	0.518
Wagetera	12	69	10	91	Wagetera	0.053	0.055	−0.195
Nabega	10	66	11	87	Nabega	−0.386	0.058	0.261
Total	50	290	45	385	p-value = 0.999

). Fewer people were interested in punitive measures against illegal users (12%) or in actively participating in a committee charged with improving management (12%). There were no strong tendencies for differing perceptions on this among Kebeles.

Respondents were unanimous on their use of non-mechanized agriculture in their rice systems irrespective of Kebele, age, education, etc. This implies the use of manual labor for the broadcast application of seeds and fertilizers and for the application of herbicides or pesticides, as well as the use of animal traction for tillage. Given the small parcels of land even for the largest farms, i.e., those in Kidest Hana (Figure 4), the lack of mechanization does not seem surprising.

Responses to the last question, on whether household heads had taken other crops out of production to produce more rice, are shown in

Table 10. Farm households’ responses on whether they had replaced alternative crops with rice. Positive or negative signs are used to indicate deviation from the model of independence, which assumes responses to a particular survey question were independent of Kebele. p-values indicate significance of differences among Kebeles. Color shades accentuate the magnitude of deviation from independence, i.e., light red and blue indicate slight deviation from independence, and darker red and blue indicate stronger deviation from independence. Tables were generated using Systat v 13.1’s XTAB module.

	Yes	No	Total		Yes	No
Kidest Hana	8	61	69	Kidest Hana	−0.271	0.104
Shaga	10	71	81	Shaga	−0.104	0.04
Shena	10	47	57	Shena	1.011	−0.387
Wagetera	10	80	90	Wagetera	−0.438	0.168
Nabega	11	76	87	Nabega	−0.03	0.012
Total	49	335	384	p-value = 0.829

. By far, the most common response, i.e., by 87% of respondents, was no. There was a light tendency for Shena respondents to say yes compared to those of other Kebeles and a very slight tendency of those in Wagetera to say no.

There was no association of the various respondent groups for their responses to perceptions on sustainable wetlands management, i.e., among gender, education, number of children, marital status, etc., except for the first question on who owned the land (Table 7), which was associated with total farm size (Figure 4). This, however, would seem to be an artifact of the responses of household heads in Kidest Hana and Shaga having both the largest farm sizes and tendencies to perceive the land as communally owned.

Regarding perceptions on ways in which individuals might contribute to improved sustainable wetland management, by far the largest response on wetland management strongly agree (57.9%), agree (34.6%), or neither agree nor disagree (7.5%), and none of them disagree or strongly disagree, respectively (

Table 11. Farmers’ perception of wetland management.

Kebeles	Respondents	Strongly Agree	Agree	Neither Agree nor Disagree	Disagree	Strongly Disagree
Kidest hana	69	45 (65.2%)	19 (27.5%)	5 (7.3%0	0	0
Shina	81	48 (59.3%)	27 (33.3%)	6 (7.4%)	0	0
Shaga	57	31 (54.4%)	23 (40.4%)	3 (5.2%)	0	0
Wagetera	91	50 (54.9%)	33 (36.3%)	8 (8.8%)	0	0
Nabega	87	49 (56.3%)	31 (35.6%)	7 (8.1%)	0	0
Total	385	223 (57.9%)	133 (34.6%)	29 (7.5%)	0	0

). There were no strong tendencies for differing perceptions on this among Kebeles.

4. Focus Group Discussion

The discussants revealed that household-head perceptions towards six factors affect sustainable management of the Fogara wetlands: (1) land ownership, (2) responsibility for resource utilization, (3) responsibility for development, (4) ways to support sustainable management, (5) intensity of rice production systems, and (6) loss of crop diversity were performed by the local community. Most participants agreed that the local communities benefited from the wetland resources. Most of them described that respondents thought wetland ownership was communal or a common-pool resource available to the Kebele, and responsibility for resource utilization of wetlands of Fogera was not clearly known by farmland holding due to the sense of common pool resources around them. This could have increased pressure on the wetland area resources for livestock grazing and agricultural expansion, especially from intense rice production systems. Discussants emphasized the responsibility for the development of wetland management and equity in sharing of resources. There was some concern over the loss of crop diversity in discussion groups, even though most questionnaire respondents had not reported replacing alternative crops with rice (Table 10). Some of the discussants noted that, previously, they used to obtain different wetland resources like reeds, papyrus, and fish, and there may be a minimal threat to the production of these wetland products. However, at present, there was a perception that the negative effects of human activity encroachment were on the increase. As a result, even though there was increased income from rice production, some of the discussants were dissatisfied with the effects of intense rice production. They considered the deterioration of wetland resources as a limiting factor in improving their livelihood. Some of them also stated that the government had been responsible for restricting their access to resources in the area by other kebeles, and some further claimed forced relocation. They also felt that other kebele members do not like neighboring communities around the study kebeles’ boundaries. Nonetheless, most discussants had a positive attitude towards sustainable wetland management for its importance in attracting tourists, job opportunities during drought, the enjoyment derived from viewing birds, and preserved wetlands’ value to future generations.

Discussion

This study endeavored to characterize the socioeconomic attributes of farm households and household-head perceptions towards factors related to sustainable wetland management with a view to finding better paths towards reversing land and water degradation.

With regard to farm household characteristics, all respondents practiced the Christian Orthodox faith, and men constituted 93% of household heads. Kebeles differed significantly for household-head age and education, but overall education level was low, with only 5 of 385 respondents having a primary school education and 81% describing themselves as illiterate. Most households had three or four children. These are smallholder farms, with a mean farm area of less than one ha and a mean farm area varying among kebeles from 0.80 to 1.35 ha. Not a single farm was mechanized. Kidest Hana was the most successful kebele in terms of farm size and yield. Practically all farmers grew rice. This is confirmed by Alemu [43]; all household members (men, women, and youth) play important roles in rice production, from land preparation to processing. Depending upon the kebele, 24–39% grew teff, 6–52% grew grass pea, and much smaller percentages grew the cash crops onion, garlic, tomato, and chickpea. Again, Alemu [43] indicated that the introduction of a cultivated rice variety resulted in significant changes in the local farming systems as rice cultivation expanded and grazing land used for Fogera cattle and other crop production (niger seed, chickpea, wheat, and oats) decreased. According to focus group discussions with local farmers, some perceive such negative trends.

Significant differences existed among Kebele farms in terms of percentages of harvested crops sold, suggesting some had less food to sell in the market. Male household-head (MHH) farms had twice the area of female household-head (FHHs) farms, older household heads had more land than younger ones, and married heads had more land than divorced or unmarried heads; these results confirmed that the average area cultivated by farmers during the 2008/09 and 2009/10 rain seasons was 2.25 acres, with MHHs cultivating larger farms than FHHs (2.41 vs. 1.75 acres, p < 0.05) [44]. Larger land holdings seemed to be concentrated among about a half-dozen older, married men. These are major farm household socioeconomic characteristics to be considered when devising better wetland management plans, including any measures intended to address apparent inequities associated with gender, age, or the distribution of land and water resources. Marital status influences the capacity of poor rural women in Ethiopia to innovate in agriculture and, more broadly, in their lives [45].

Because these wetlands constitute a CPR, better paths towards reversing land and water degradation would ideally include as many of the eight “design principles” that Ostrom [46] found among long-enduring CPR groups that sustainably managed natural resources. The importance and robustness of these principles were later confirmed by subsequent studies of long-enduring CPR groups by Cox et al. [47]:

• Individuals or households who have the right to withdraw resource units from the CPR must be clearly defined, as must the boundaries of the CPR itself;
• Congruence between appropriation and provision rules and local conditions;
• Most individuals affected by the operational rules can participate in modifying the operational rules;
• Monitors who actively audit CPR conditions and appropriator behavior are accountable to the appropriators or are the appropriators;
• Appropriators who violate operational rules are likely to be assessed as graduated sanctions (depending on the seriousness and context of the offense) by other appropriators, by officials accountable to these appropriators, or by both;
• Appropriators and their officials have rapid access to low-cost local arenas to resolve conflicts among appropriators or between appropriators and officials;
• The rights of appropriators to devise their own institutions are not challenged by external governmental authorities;
• For larger CPR groups, nested enterprises. Appropriation, provision monitoring, enforcement, conflict resolution, and governance activities are organized in multiple layers of nested enterprises.

Our survey results help elucidate household head perceptions towards factors associated with sustainable management of the Fogara wetlands and the current status of management within the context of Ostrom’s [46] design principles. Since the Fogara wetland Kebeles do not constitute “larger CPR groups”, principle #8, on multiple layers of nested enterprises, would not be applicable.

The fact that 15% of respondents thought the land was privately owned and 85% thought it was communally owned (Table 7) suggests a level of confusion that would interfere with sustainable land management because, apparently, households do not know who has the right to withdraw resources from the CPR or, for that matter, where the CPR’s boundaries lie (principle #1). Moreover, it implies a lack of clarity over who sets appropriation and provision rules (principle #2), who can participate in modifying the operational rules (principle #3), who can monitor CPR conditions (principle #4), who sanctions appropriators who violate operational rules (principle #5), who mediates conflict (principle #6), and what defines the relationship between the appropriators and external government authorities (principle #7). The fact that the percentage of respondents who believed land ownership was private ranged from as little as 5% in Shaga Kebele to as much as 28% in Nabega (Table 7) underscores this confusion and lack of clarity.

Irrespective of Kebele, age, level of education, or gender, not a single respondent had any familiarity with any organization, whether governmental or non-governmental, working on issues of sustainable wetland management. Similarly, 44% of respondents did not know who was responsible for overseeing land management (Table 8). Another 22% thought the community was collectively responsible, 22% thought a development agency was responsible, and only 12% thought that the Kebele leader was in charge (Table 8). These underscore a lack of communication and education and, together, collectively further undermine the first seven of Ostom’s [46] principles. There is confusion on who could serve as a knowledge resource and a complete lack of clarity on who is in charge. Once more, differences in perceptions among Kebeles on who is in charge further underscore the confusion, with Shena viewing the role of the Kebele leader as particularly weak and Nabega viewing it as particularly strong. Such a situation can only further the Tragedy of the Commons that is occurring in these wetlands.

Despite confusion about ownership, responsibility, and knowledge resources, household head responses on willingness to participate in measures towards improving sustainable wetland management would seem to provide room for some optimism (Table 9). A large majority (87%) of respondents were interested in participating in meetings or committees, and only a minority were interested in punitive measures against offenders, which might lead to conflict. Moreover, there was little difference among Kebeles for this perception (p = 0.999). Such committees and meetings, particularly if led by qualified government or non-government facilitators trained in conflict management, may offer mechanisms for much-needed education on the ownership and responsibility of CPRs. They could provide a mechanism to collectively define (a) CPR boundaries and rights as far as who can withdraw resource units (principle #1), (b) congruence between appropriation rules and local conditions (principle #2), (c) pathways through which operational rules can be modified (principle #3), (d) define or recognize monitors and their accountability (principle #4), graduated sanctions against violators (principle #5), conflict-resolution mechanisms (principle #6), and determine, in consultation with local government, the rights and abilities to organize (principle #7). Birhan [48] has also supported this idea.

Finally, the observation that only 13% of respondents had taken alternative crops out of production to produce more rice gives hope that the Fogera wetlands are not yet nearly so degraded as was the case with wetlands around Lake Haromaya, described by EWNRA [14] and Kassa [15]. It suggests that, contrary to what Abebe et al. [16] had found among farmers near Lake Harmony, most Fogera wetland farmers had not decreased production of vegetables, irrigated crops, or inter-cropped fields.

5. Conclusions

The study analyzed farm household characteristics for socioeconomic attributes and their perceptions of factors influencing the sustainable management of Fogera wetlands. This research has shown that there are many stakeholders using the wetlands and that they have varied perceptions, motivations, and interests for their participation and operate at various levels, with community-level stakeholders being the least active ones. The second major finding was that no single stakeholder category is responsible for the degradation and conversion of wetlands, although the degree and scale of their contribution differ, with rice farmers contributing more than other activities. Results further suggest a level of confusion that would interfere with sustainable land management because, apparently, households do not know who has the right to withdraw resources from the CPR. There is a lack of enabling laws and policies to help guide the conservation and restoration of wetlands. There is no concerned body regarding wetland management, leaving the management of this critical resource solely in the hands of those who want to use it to benefit themselves rather than satisfy societal needs. Most of the stakeholders supported wise use in principle but lacked the capacity to implement it. Taken together, these results suggest that the majority of the stakeholders are concerned about the current state of wetland degradation in the Fogera wetlands and blame the central government for not doing enough to halt damaging activities using the available laws and policies. However, driven by the desire to create wealth and improve the standards of living for the majority of citizens, it is likely to remain a big challenge for the government to conserve wetlands as it strives to develop the country. Every effort needs to be taken to address the socio-ecological and development challenges that face Fogera wetlands and have national and international impacts affecting balanced and sustainable development. Our study highlights the need for better education, leadership, and policies regarding sustainable wetlands management. Further research is needed to determine the value of lost wetland area versus increased overall revenue for farmers, as well as the win–win scenario for future development while considering the area’s wetland resource sustainability.

6. Recommendation

• The Fogera floodplain needs to be regulated based on a wetland management plan.
• Encourage the preservation of indigenous knowledge and practices so that wetland systems and rice cultivation can co-exist in a win–win arrangement.
• Environmentally friendly, non-intensive production processes and the marketing of regional products should be encouraged even more to sustain wetlands.
• Raising public awareness, developing independent wetlands policies, enhancing organization-to-organization cooperation, creating buffer zones, and ongoing monitoring are key policy and institutional issues that need to be practiced to save Fogera wetlands and Lake Tana.
• Establish a stronger Traditional Ecological Knowledge framework and policy; the research community should conduct more studies at national and regional levels. As a result, policymakers and stakeholders must collaborate on wetlands management vis-à-vis rice production intensification.
• Hence, proper management of Fogera wetland resources is required because, without proper management, these valuable resources will soon be lost or will no longer be able to play their required role in the socioeconomic development of the area.