Global literature review - Quantitative findings

Spatial variation in the direct dried fish and general fisheries literatures

Table 1.

Continent	Dried fish sample		General sample
	percent	count	percent	count
Latin America and the Caribbean	29	2	7	7
Africa	248	20	5	5
Asia, Oceania, & Middle East	618	50	16	16
Subtotal: Global South	895	72	28	28
Europe	156	12	4	4
North America	54	4	5	5
Subtotal: Global North	210	16	9	9
Global/None	181	14	62	63
TOTAL	1105	102	37	100

For the dried fish literature, 53% of the direct references concerned the Global South and 17% focused on countries and regions of the Global North. This represents proportionately much greater attention to the Global South than the general fisheries literature sample, suggesting that the literature on dried fish more closely mirrors global population distribution than the general fisheries literature. [GU1]

Thirty percent of the direct dried fish literature has either no spatial focus or, in the case of just a few references, is global in scope. As Table 1 shows, the literature on dried fish in Asia is more than 40% of the total. Africa and Europe also each have substantial literatures, while that from the Americas is much smaller proportionately. There are likely, however, to be large literatures in Chinese, Japanese, Korean, Spanish, and Portuguese that the English focus of this literature review missed[BB2] . The overall geographical emphasis of the literature by continent would be unlike to shift dramatically in non-English language papers were included.

Thematic variation in the dried fish literature

Figure 1 presents the overall thematic shape of the dried fish literature in relation to the broader fisheries literature[1]. As indicated in the figure, the literature on dried fish shows a remarkable technical orientation[BB3] [DJ4] , with 55% of all direct references falling into this category. Many of the nutrition, health, and value chain references in the second and third ranked thematic categories are also tagged with the technical theme, which reinforces the dominance of the category in the dried fish literature. The dried fish literature is oriented towards economic and health improvement foremost. The remaining cluster of themes [BB5] that we argue are important to a holistic appreciation for the dried fish sector only represent a small minority in the literature. Unlike the technical literature, the non-technical contributions tend to more wide-ranging thematically. Attention to socio-economic and historical issues, even though relatively limited, is important as a rich starting points for further work.

[DJ6] [ET7]

Spatial variation in the direct dried fish and general fisheries literatures. SOURCE

The contrast of the dried fish literature with the general fisheries literature is striking. The least and second least most addressed themes in the dried fish literature, ecology and policy and governance, are the most addressed themes in the general fisheries literature. Technical concerns in the general fisheries literature are the third most important, but form a much lower proportion than in the direct fisheries literature. Culture is considerably better treated in the general fisheries literature, while attention to economic matters is the most comparable theme in terms of weight between the two literatures.

When the dried fish literature is examined from the continental, scale of analysis (Figure 2), divergences among world regions become evident. While the technical literature is universally dominant, it is especially pronounced in the Asian and African literatures. This technical orientation likely reflects histories of development-related poverty and health interventions. Africa's relatively greater focus on health and gender suggest that a development-driven agenda is the strongest in that continental context. Attention to the economic dimensions of dried fish is relatively less important in Asia as compared to Africa and Europe. Of all the continents, Europe's literature is most strongly oriented towards economic and change and history orientations. All continents show a surprising absence of attention to social questions, labour, gender, and particularly policy and ecology. This shows that the primary thematic focuses of the broader fisheries literature have yet to move downstream to influence the dried fish literature.

[DJ8] [ET9] [ET10]

Figure 1 Direct dried fish literature compared against the general fisheries literature by proportion of thematic tags to total tag set. SOURCE

Figure 2. Theme by continent[ET11]

3.1.3 Value chain segment coverage in the dried fish literature

The broader fisheries literature has long been characterized by a productivist bias (Tezzo et al. xxxx), as is underscored in Figure 3. Emphasis in the dried fish literature, in contrast, is on the processing segment of fisheries value chain (53%[DJ12] of all segment tags). In that sense, the dried fish literature goes part of the way to addressing the imbalance in the broader fisheries literature. As Figure 2 shows, however, the dried fish literature pays much less attention to other mid-stream value chain segments, so fails to rectify the problem of the 'missing middle' that characterizes much research on food value chains (Reardon et al. 20xx).

The dominance of attention to processing in the dried fish literature reflects the fact that the literature is centered on a product type, not fish as an organism. In the dried fish literature, processing is equivalent to the production segment, where the actual extraction from aquatic systems through fishing - the “production” segment in our value chain - is reduced to an input supplier. The processing-related focus of the large technical literature on dried fish makes sense given the literature's product focus.

The relative lack of variation in the proportions of non-processing segment tags in Figure 4 is striking, suggesting that the literature does make connections across dried fish value chains, and the beginnings of a complete food system perspective are available in the literature. Retail (7.5% of tags) is the least studied segment, despite its crucial importance in terms of access to dried fish and in terms of livelihood provision for low income participants in dried fish value chains, and women in particular (e.g. Hapke xxxx). Consumption has the greatest attention after processing, with 15.1% of the total segment tags. Trade also suffers surprising neglect in the dried fish literature (11.3% of total segment tags) even though, like retail, it is of critical importance for product distribution and employment.

[DJ13]

Figure 3. Proportional representation of attention to value chain segment in direct dried fish and general fisheries literatures

The global pattern of the missing middle in food value chain research is replicated universally at the continental scale (Figure 4). Three points of variation deserve flagging. First, there is a striking consistency across the continents of the general pattern of dominance in attention to processing. Of all the continents, Africa's literature is most balanced across non-processing segments. Although here, while is has the highest proportional attention to the retail segment (at 11% of all Africa segment tags), it has the comparatively least attention to the consumption segment. North America's dried fish literature is most skewed towards processing (63%), with no reference to retail whatsoever.

[DJ14] [DJ15]

Figure 4. Segment by continent

3.1.4 Attention to product type in the literature on dried fish

There is considerably ambiguity and inconsistency in how dried fish product types are defined. This is understandable in that dried fish product types may be produced using multiple different methods. Dried fish, for example, are often salted prior to drying. At the broadest possible level of product type categorization, however, all products may be categorized as either dried or fermented. Smoked fish, for example, are dried using smoke rather than sunlight and fish sauce is a by-product of fish fermentation. We have chosen a middle level of product type differentiation into six categories (Figure 6) that is sufficiently fine grained to capture important geographical distinctions.

The four main product types of dried, fermented, salted, and smoked fish each get substantial attention in the dried fish literature, suggesting that at the global level each processing type has an important functional place. The relatively even spread of attention to different product types at the global level obscures major continental and sub-continental differences, as shown in Figure 6. Asian attention is on dried and fermented fish; Africa's is on dried and smoked; and salted fish is most important in Europe and the Americas. At the lower sub-Asian regional scale level, there is a further major distinction between emphasis on South Asia on dried fish and emphasis in Southeast and East Asia on fermented fish. These major geographical distinctions in product type raise fascinating why questions for ecology, geography, and anthropology.

[DJ16]

Figure 5. Proportion of global literature on dried fish by product type

[DJ17]

Figure 6. Product type by continent

Figure 7. Product type by Asian region[ET18] [DJ19] [ET20]

Publication type and journal analysis

Table 2 lists the number of articles concerning dried fish published since 2005, by journal with three or more publications considered of direct relevance to the topic of dried fish.

This list is dominated by natural science journals, representing articles classified in our review as “technical” research. Only two of the 28 journals listed here publish research outside this domain: the Journal of Ethnic Foods, which includes a mixture of technical and descriptive reports; and the Journal of Maritime Archaeology. No social science and humanities journals are represented in this list.

Table 2. Publications by journal

article count	journal
21	Food Chemistry
20	Food Control
11	Journal of Aquatic Food Product Technology
10	International Journal of Food Microbiology
8	International Journal of Food Science & Technology
7	International Food Research Journal
7	Journal of Ethnic Foods
7	Journal of Food Processing and Preservation
6	African Journal of Food, Agriculture, Nutrition and Development
6	Journal of Food Engineering
5	European Food Research and Technology
5	Fisheries Science
5	Journal of Food Composition and Analysis
5	Journal of Food Science and Technology
5	Journal of Maritime Archaeology
4	Drying Technology
4	Food Research International
4	Food Science
4	Journal of Food Protection
4	Journal of Food Safety
4	Journal of Food Science
3	Fishery Technology
3	Food and Chemical Toxicology
3	Journal of Food and Drug Analysis
3	Journal of the Science of Food and Agriculture
3	LWT - Food Science and Technology
3	Modern Food Science and Technology
3	Nutrition & Food Science
3	Polycyclic Aromatic Compounds

[1] For comparative purposes, we conducted a literature review of the general fisheries literature using the search terms 'fisher'; 'fisheries', and; 'fishing' in Google Scholar. Every fifth record was selected until 33 records were selected per search term. We then tagged the resulting literature set following the guidelines we used for the dried fish literature, as explained in the methodology section above.

The graph is the nicer way to present this information. Suggest cutting the table. Also suggest dropping global/none from graph we can mention in the text [GU1]

[BB2]Its possible we also miss Spanish references for Latin America?

[BB3]This is partly a function of how we define technical - if we were to have a single 'socio-economic' category, it would look a lot larger than our individual social and economic categories (though still smaller then technical). If were to break 'technical' down into food processing technology, chemical composition analysis, etc, it would start to look a bit more variegated.

This is just to say that we need to provide some justification of our logic for calling articles 'technical', rather than it being a catchall for 'stuff we aren't very interested in'

[DJ4]I have made a preliminary effort to address this point, but we will need to return to it and refine the explanation.

[BB5]There are also often multiple themes of interest in a single paper, whereas technical papers tend to be purely technical

[DJ6]Change literature labels to Direct dried fish literature and General fisheries literature

Before seeing this comment [ET7]I applied the tags “Dried Fish Sample” and “General Sample”. I like the term “sample” but don't feel strongly so can certainly use “literature” if that is our preference. The logic to me is that these are two distinct samples from published literature, with defined selection criteria, that do not necessarily represent the complete body of literature itself. “Direct Dried Fish” seems a bit awkward to me; the “direct” is probably not necessary.

[DJ8]Formatting: I suggest ordering the bars by size from greatest to least on the technical row. I.e., Asia and Oceania first, followed by Africa.

Middle East should be combined with Asia, not Africa.

Try moving the continent labels below the graph to increase the size of the figure.

Wrap the text of the theme labels to increase the size of the figure.

Remove the tag row axis label.

[ET9] I suggest ordering the bars by size from greatest to least on the technical row. -a The figures are always sorted by y axis values but the order of categories (i.e., continents) is fixed for consistency across the graphs.

Middle East should be combined with Asia, not Africa. a DONE

Try moving the continent labels below the graph to increase the size of the figure. a DONE

Wrap the text of the theme labels to increase the size of the figure. a DONE

Remove the tag row axis label. a DONE

[ET10]Two options here: overall numbers (figure 2a) or proportion as percentage (figure 2b).

The third option is subgraphs

[ET11]This figure isn't too useful in its current format since there are too many categories. There are two options to deal with this: (1) conflate the regional categories (e.g., “Asia & Oceania”, “Americas”, “Africa”, “Europe”) or (2) generate a selective set of bar graphs for individual themes. We can probably exclude the themes like “ecology” that have almost no coverage anyway.

[DJ12]Give final figure when available.

[DJ13]Formatting: remove the tag row label.

Change literature labels to Direct dried fish literature and General fisheries literature

Order segments in following order: production, processing, trade, retail, consumption.

[DJ14]Change the segment order to match figure 4.

Remove tag label.

[DJ15]

[DJ16]Wrap fish powder and derivative products label.

Remove tag and count labels

[DJ17]Add Middle East to Asia

Category type should go in same order as in figure 6

[ET18]These are not up-to-date figures. When regenerating this graph, perhaps use the same categories above?

[DJ19]This figure still needs to be updated.

I find the organization of the labels in this graphical presentation creates a clearer figure that the newly generated figures. Perhaps also the vertical rather than horizontal bars also add to clarity. What do the rest of you think?

[ET20]This layout replicates the Excel graph design. The only substantial difference is that Excel puts a space between the columns in each set.