Discussions regarding the use of mucuna on steep slopes in Honduras
(7/29/02 - 8/12/02)

______________________________________________________________

postings: 10
countries: Canada, Honduras, Mexico, Nicaragua, USA (Hawaii, New York)
organizations/institutions: CATIE, CIDICCO, CIRAD/CIMMYT, Cornell University, IDRC

(Note: the last post in this discussion summarizes the first nine postings)

Date: Mon, 29 Jul 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Torsten Mark Kowal
Subject: Mucuna on steep slopes in humid zones

Colleagues:

I am involved in an impact study of an integrated development project that was implemented on the Honduran Atlantic coast. Rainfall in the region is often above 3000 mm/year, and farmers (many are migrants from other zones) practise basic grains cropping on moderate to steep slopes, often on extremely erodible soils, usually following cutting of fallow vegetation (both long and short duration), as well as primary forest.

As many of you will be aware, over the last 20 years or so, farmers on the north coast have found that Mucuna is an important addition to the basic grains farming system, and researchers have found relatively high levels of spontaneous dissemination.

However, during our field work, many farmers have mentioned that, while Mucuna works well on lower gradient slopes, when established on steep slopes and fragile soils, the accumulated biomass (simultaneous maize or follow-on bean crops + Mucuna), leads to far greater land slippages and rill erosion (compared to maize without Mucuna). Farmers give this as a reason for not using Mucuna in these settings, implying that net benefits are negative.

I'd be interested in opinions of mulch subscribers on the extent to which this is a proven limitation on the use of Mucuna, whether other CCs suffer this problem to a lesser extent on high gradients in humid zones, and whether there are other agronomic measures that can be used to extend Mucuna use onto higher slopes.

I'd also like your opinions on what makes up best practice sustainable hillslope agriculture in this region, given that this project worked more in attempts to diversify farm production (cacao, rambutan, coffee etc) than on improving basic grain production. For example, are their specific examples that show to what extent have frijol tapado and improved fallows been shown to be viable alternatives to swidden farming on the Honduran north coast?

Best regards,

Torsten Mark Kowal
(Honduras)

______________________________________________________

To: MULCH-L@cornell.edu (MULCH-L)
From: William Cook
Subject: Re: Mucuna on steep slopes in humid zones
Date: Wed, 07 Aug 2002

Torsten Mark Kowal,

Interesting report! But puzzles me that land slippage and rill erosion occur on steep slopes with plenty of cover... How do farmers manage the Mucuna? Do they not slash it in place, leaving trash, or do they pull the crop off and leave the soil bare?

As to other crops for diversification, what can these farmers sell to advantage?

Thanks much,

WmCook
Honolulu

______________________________________________________

From: Daniel Buckles
To: MULCH-L@cornell.edu (MULCH-L)
Subject: Re: Mucuna on steep slopes in humid zones
Date: Wed, 7 Aug 2002

Hello. According to research in Honduras from the early to mid 1990s, farmers believe that Mucuna managed as a sole crop in rotation with corn creates a loose surface area susceptible to slumping under heavy rainfall. It is true that the Mucuna-maize rotation eliminates virtually all other plant species, including trees and deeper rooting plants that might bind the soil at very depths. We found no hard evidence, however, that fields of the same degree of slope managed under Mucuna are any more susceptible to slumping that fields managed under other crops. We did find evidence that farmers use steeper slopes for Mucuna-maize rotations than would otherwise be put into a maize system, and that steeper slopes tend to slump more easily, practices that over time would create the impression that Mucuna fields are more susceptible.

For details and a full text copy of the reference see Buckles, D. et al. 1996. Cover Crops in Hillside Agriculture: Farmer Innovation with Mucuna: http://www.idrc.ca/acb/showdetl.cfm?&DID=6&Product_ID=486&CATID=15

Daniel Buckles
IDRC
Canada

______________________________________________________

Date: Wed, 07 Aug 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Harold van Es
Subject: Re: Mucuna on steep slopes in humid zones

Dear Torsten and others,

I have also heard about the increased risk of land slips with the use of mucuna. This concern was noted more than 10 years ago when we were doing an infliltration study on the mucuna-maize system in Northern Honduras. After Hurricane Mitch, land slides and slips were of course widely reported. The World Neighbors survey conducted immediately after Mitch showed that conservation practices did not show any positive effect on the potential to reduce landslides (although water erosion was generally reduced).

One of my former graduate students, Ben Zaitchik, did a study on predicting landslide potentials using GIS-based models near La Libertad, Districto Comayagua, a little ways South of the area you are referring to. He used the Mitch-generated landslides as validation for the model's performance (which was fairly good). This work is currently in press through two papers, one in the Soil Science Society of America Journal, and the other in the Journal of Soil and Water Conservation (the latter is more targeted towards the field professionals). According to a model sensitivity analysis, slope gradient, and relative soil wetness (mostly a function of the amount of converging water flow in the landscape) were the most important factors affecting the potential for slope failure. Soil type is an additional factor, with deep and poorly draining clay soils (also found on the hillsides in Northern Honduras) having high potentials for failure. However, we did not consider the effects of cropping patterns, and we focused on identifying locations in the landscape where landslides are likely to occur under extreme hydrologic conditions.

From a physical perspective, it makes sense that heavy mulches would increase the potential for landslides, because the increased water infiltration potential makes the soil more likely to become supersaturated, thereby weakening internal soil cohesion. There are several ways to reduce this risk. When landslides are shallow (< 1 meter), sparsely planted trees greatly reduce the potential for landslides, because their roots provide anchorage for the soil material. Sparse tree planting does not have to interfere much with crop production potential, especially in the tropics. Also, the trees may be planted only in areas where slips are likely to occur (based on soil gradient and wetness criteria).

A second, but untested, approach is to manage the mulch to encourage some water shedding rather than infiltration. For example, in the mucuna-maize system, the mulch material (especially maize) can be arranged in the up-and-down slope direction to obtain a thatched roof effect. Perhaps using tall grasses that lodge over in the downslope direction may provide similar results. The drawback is that we may be generating more runoff. Again, this is untested as far as I know.

If the landslides are deep (> 1m), however, the situation is different. Trees won't help, because the roots don't extend down to the failure plane. If fact, they may make matters worse by adding weight and increasing water movement to the deeper parts of the soil. The traditional approach to addressing such problems is to provide better drainage to the lower slope sections (e.g., by boring a hole horizontally into the slope and pushing in a drain tube). This is typically beyond the financial means of the compesino and tends to be used more in road construction projects.

So, in summary, landslides are difficult to prevent if steep hillsides are being cultivated in areas that receive a lot of rain, especially incidental tropical storms.

Cheers,

Harold van Es
Dept of Crop and Soil Sciences
Cornell University

______________________________________________________

Date: Thu, 08 Aug 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Proyecto Catie MIP Nicaragua
Subject: Re: Mucuna on steep slopes in humid zones

Torstan,
I think we need more of your observations about farmers comments. One might suggest that soils under mulch remain saturated for longer periods and are therefore more susceptible to slippage. However, the greater presence of rill erosion is contradictory evidence. Do you think what farmers are saying is accurate on a field by field basis? Tell us more about what happens to the residues during the year in the two different systems. How much does the situation vary from field to field?

Charles Staver
CATIE - Nicaragua

______________________________________________________

Date: Thu, 08 Aug 2002
From: Bernard Triomphe
Subject: Re: Mucuna on steep slopes in humid zones
To: MULCH-L@cornell.edu (MULCH-L)

Dear Torsten:

Just in case, this is something I wrote in my Ph.D dissertation 7 years ago, based in field work done 10 years ago. You can cite it alternatively as Buckles et al, 1998: Cover crops in hillside agriculture. Farmer Innovation with Mucuna. IDRC and CIMMYT).

" As much as 40% of farmers interviewed by Buckles et al. (1992) reported that the mucuna system might induce localized landslides in areas of very steep slopes (superior to 60-70%). Discussion with farmers confirmed that such landslides occur once in a while (not every year) during the peak of the rainy season (anytime between Sept and Nov.) under very heavy rainfall conditions (several hundred mm in a few days). A possible explanation would include a combination of the heavy weight of the wet mucuna biomass, a loosening of the upper soil profile as a result of shallow rooting habits of the mucuna plant, and/or a state of superstauration of the soil resulting from increased infiltration, inducing a lower shear strength and higher overburden weight (Van Es, pers. com.). Some farmers also indicated that landslides might result from the lack of deep rooting or anchoring caused by the substitution of the traditional bush-fallow rotation for one with a fairly shallow-rooted species such as mucuna, furthermore quite able to eradicate the few trees left in place by farmers if left unpruned.

None of these explanations is completely convincing. The landscape in the mountains of Northern Honduras appears to be geologically very young, having not yet fully stabilized. Hence there are numerous areas where mass redistribution continues to take place "spontaneously", and sectors of abrupt slopes are among the prime candidates for being affected by this gravitydriven redistribution process (whether such a landscape should ever be subjected to large scale cultivation is definitely a relevant question). Also, one could argue that when quantities of water pouring on any landform reach hundreds of mm in a few hours or days, something dramatic is poised to happen, and the actual role of the velvetbean cover in causing a landslide is probably insignificant compared to the role played by the sheer masses of water rushing their way downhill. This may explain why landslides, when they take place, affect lands under all kinds of land use, from virgin forest to pastures to fields cultivated without velvetbean, without obvious preferential impact on any one category of land use. This could be seen when 400 to 700 mm fell in a 15 hour period on October 31st, 1993 causing countless landslides in the hillsides.

The issue seemed important enough to address it in a general survey of the velvetbean system conducted in the summer of 94. Farmers were specifically asked about the occurrence of landslides prior to and after the introduction of the velvetbean rotation in their fields. Out of 34 fields having suffered from landslides (from a total of 44 fields included in the survey), 62% (21) had had similar problems before velvetbean was ever introduced. Furthermore, only 1/3 of the farmers incriminated velvetbean in the occurrence of landslides. Perceptions varied strongly from village to village: in Piedras Amarillas, where landslides are common, farmers blamed velvetbean for making things worse, whereas in San Francisco de Saco, where landslides are rare, most experienced velvetbean users vehemently opposed this view."

For more recent and more diverse experiences, you may want to contact World Neighbors and Eric Holt Gimenez to find out about the conclusions of their Post-Mitch study: most certainly, they have come up with interesting data about how different CC systems may contribute to soil conservation on steep slopes.

Good luck,

Bernard Triomphe
CIRAD/CIMMYT
Mexico

______________________________________________________

Date: Thu, 08 Aug 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Milton Flores-Barahona
Subject: Mucuna discussions revisted

Colleagues,
F or those of you who are interested, please visit the site http://ppathw3.cals.cornell.edu/mba_project/moist/mulchmail.html, of the TropSCORE Consortium where, thanks to the efforts of Lucy Fisher, you will find a record of the interesting discussions that took place during 1996-1998. Messages related to the Mucuna cultivation were sent by several mulch subscribers.

Salud y saludos a todos,

Milton Flores
CIDICCO-Honduras

______________________________________________________

From: Donald Kass
To: MULCH-L@cornell.edu (MULCH-L)
Subject: Re: Mucuna on steep slopes in humid zones
Date: Fri, 09 Aug 2002

Dear Milton, Lucy, Harold et al.,

When Milton and I reviewed the Central American literature on mulching for the 2000 ASA meetings symposium, I found a paper by Cassel and others in Tropical Agriculture who had measured some physical characteristics of northern Honduran soils with and without the mucuna mulch. When I looked at the data, it indicated what Harold and the farmers say, that the mulch does produce some discontinuities in soil physical properties which would make the surface soil more susceptible to slumping. However, for some reason, the authors looked at the data and concluded that this wouldn't happen. I understood that CIDICCO was advocating planting dispersed trees in the mulch to reduce the possibility of slumping. Well, I guess I am out of this now--but it would be nice if what has been written could be published so everyone won't go on reinventing the wheel.

Best to all from erosion-free Jamaica (New York) where they do make very good pupusas ( in the three Salvadorean restaurants about five blocks from the FDA Northeast Regional Laboratory)-- and even sell Salvavidas and Port Royal Beer (from Honduras) --so I get some of the niceties of Central America without its problems.

Don Kass
FDA Northeast Regional Laboratory
New York

______________________________________________________

Date: Sat, 10 Aug 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Proyecto Catie MIP Nicaragua
Subject: Re: Mucuna on steep slopes in humid zones

Torsten and colleagues,
The information has been great. I still dont understand why the farmers would say there is more slumping and more rill erosion. None of the comments so far have talked about the rill erosion.

Charles Staver
Proyecto Catie MIP Nicaragua

______________________________________________________

Date: Mon, 12 Aug 2002
To: MULCH-L@cornell.edu (MULCH-L)
From: Torsten Mark Kowal
Subject: Summary hillslopes Mucuna discussion

Dear Colleagues,

Thank you for your comments. I'd like to summarize and take up some of your comments.

William Cook wrote:

But puzzles me that land slippage and rill erosion occur on steep slopes with plenty of cover... How do farmers manage the Mucuna? Do they not slash it in place, leaving trash, or do they pull the crop off and leave the soil bare?
As to other crops for diversification, what can these farmers sell to advantage?

The major problem appears to be the interrelation between very friable soil types, excessive gradients of the cultivated slopes and the extreme nature of rainfall in the region, not so much Mucuna management as such, which follows usual practise. After maize harvesting, and take over of the site by Mucuna, the slash is left to decompose, and self-seeding occurs in the plot. The problem is then both shallow small scale landslips, as well as rill erosion of the accumulated organic matter.

In spite of the projects best efforts, only 4 tree crops have proved profitable: Rambutan (Nephelium lappaceum), Cacao, Mangostan (Garcina mangostana) and improved varieties of plantain / banana. Black pepper has also had limited uptake. The first two can generate substantial income from half acre plots.

Daniel Buckles wrote:

According to research in Honduras from the early to mid 1990s, farmers believe that Mucuna managed as a sole crop in rotation with corn creates a loose surface area susceptible to slumping under heavy rainfall. It is true that the Mucuna-maize rotation eliminates virtually all other plant species, including trees and deeper rooting plants that might bind the soil at very depths. We found no hard evidence, however, that fields of the same degree of slope managed under Mucuna are any more susceptible to slumping that fields managed under other crops. We did find evidence that farmers use steeper slopes for Mucuna-maize rotations than would otherwise be put into a maize system, and that steeper slopes tend to slump more easily, practices that over time would create the impression that Mucuna fields are more susceptible.

This appears to highlight the issue of farmer perceptions contrasting with research results. Also that the option of combining dispersed tree vegetation with Mucuna might alleviate the slumping associated with cropping on very steep slopes through a root binding effect.

In our field visits we saw an interesting experiment carried out by a farmer leader who had established 2 metre tall Gliricidia stakes at spacings of around 5 metres, within a 3/4 acre moderately sloping maiz-Mucuna field.

His idea was (1) to combine the high quality mulch from Gliricidia (pruned once yearly) with that of Mucuna; (2) to establish trees for a follow--on improved fallow; (3) and in third place to avoid landslips. However, at that wide spacing, and given the shallow rooting of trees established via stakes, he was unlikely to achieve the last objective. The fertility benefit of dispersed Gliricidias in a follow on fallow would be likely to be substantial. Observation of the effects on crops of live fence Gliricidia prunings, had led this particular farmer to carry out the experiment, which was interesting.

Other options for establishing trees would be to not burn fallow vegetation (these tend to be 3-4 years old at slashing), and to conserve appropriate trees in the following milpa. These could be left with crown intact, or pollarded at 2 metres or so, along the lines of the Quesungual system promoted by the FAO Lempira Sur project.

A further option would be to direct sow Ingas or Gliricidia etc. as an improved fallow, leading to a dispersed trees system following the end of the fallow cycle. Again not burning when cutting the majority of the trees in the fallow, would be vital.

This raises the issues of enhanced pests and diseases, which is one of the major reasons farmers give in the area for burning fallow residues (either for bean or maize cropping). Presence of thorns and greater labour requirements also factor in, as arguments against not burning. This makes up another challenge for sustainable agriculture development in humid zones.

Harold van Es wrote:

I have also heard about the increased risk of land slips with the use of mucuna. This concern was noted more than 10 years ago when we were doing an infliltration study on the mucuna-maize system in Northern Honduras. After Hurricane Mitch, land slides and slips were of course widely reported. The World Neighbors survey conducted immediately after Mitch showed that conservation practices did not show any positive effect on the potential to reduce landslides (although water erosion was generally reduced).

One of my former graduate students, Ben Zaitchik, did a study on predicting landslide potentials using GIS-based models near La Libertad, Districto Comayagua, a little ways South of the area you are referring to. He used the Mitch-generated landslides as validation for the model's performance (which was fairly good). This work is currently in press through two papers, one in the Soil Science Society of America Journal, and the other in the Journal of Soil and Water Conservation (the latter is more targeted towards the field professionals). According to a model sensitivity analysis, slope gradient, and relative soil wetness (mostly a function of the amount of converging water flow in the landscape) were the most important factors affecting the potential for slope failure. Soil type is an additional factor, with deep and poorly draining clay soils (also found on the hillsides in Northern Honduras) having high potentials for failure. However, we did not consider the effects of cropping patterns, and we focused on identifying locations in the landscape where landslides are likely to occur under extreme hydrologic conditions.

From a physical perspective, it makes sense that heavy mulches would increase the potential for landslides, because the increased water infiltration potential makes the soil more likely to become supersaturated, thereby weakening internal soil cohesion. There are several ways to reduce this risk. When landslides are shallow (< 1 meter), sparsely planted trees greatly reduce the potential for landslides, because their roots provide anchorage for the soil material. Sparse tree planting does not have to interfere much with crop production potential, especially in the tropics. Also, the trees may be planted only in areas where slips are likely to occur (based on soil gradient and wetness criteria).

A second, but untested, approach is to manage the mulch to encourage some water shedding rather than infiltration. For example, in the mucuna-maize system, the mulch material (especially maize) can be arranged in the up-and-down slope direction to obtain a thatched roof effect. Perhaps using tall grasses that lodge over in the downslope direction may provide similar results. The drawback is that we may be generating more runoff. Again, this is untested as far as I know.

If the landslides are deep (> 1m), however, the situation is different. Trees won't help, because the roots don't extend down to the failure plane. If fact, they may make matters worse by adding weight and increasing water movement to the deeper parts of the soil. The traditional approach to addressing such problems is to provide better drainage to the lower slope sections (e.g., by boring a hole horizontally into the slope and pushing in a drain tube). This is typically beyond the financial means of the compesino and tends to be used more in road construction projects.

So, in summary, landslides are difficult to prevent if steep hillsides are being cultivated in areas that receive a lot of rain, especially incidental tropical storms.

A very useful contribution. It appears that the problem is almost entirely small scale landslips, not deep, large-area mass wastage. In this case, trees could play a role in alleviating these Mucuna-induced shallow slips.

Erosion of organic matter is a different issue, and the extent to which tree roots could anchor top soil and decomposing Mucuna, would depend on spacings and tree management. In the Quesungual system in Lempira, research found that timber tree densities reached unexpectedly high average values of 400 trees/ha and pollarded trees could reach 900 per ha., for total basal areas of 6 m2 per ha. These data reflect the origin of field trees in previous fallows.

However, in this region, Mucuna had not been particularly heavily promoted or adopted, so observations on their potential complementarity were not possible.

Charles Staver wrote:

I think we need more of your observations about farmers comments. One might suggest that soils under mulch remain saturated for longer periods and are therefore more susceptible to slippage. However, the greater presence of
rill erosion is contradictory evidence. Do you think what farmers are saying is accurate on a field by field basis? Tell us more about what happens to the residues during the year in the two different systems. How much does the situation vary from field to field?

Daniel Buckles presented some research results that answer your questions. Unfortunately, given the restricted scope of our work, we don't have time to fully study these issues. These should be looked at if, and when, further project support in sustainable agriculture comes on stream.

Bernard Tromphie wrote:

Just in case, this is something I wrote in my Ph.D dissertation 7 years ago, based in field work done 10 years ago. You can cite it alternatively as Buckles et al, 1998: Cover crops in hillside agriculture. Farmer Innovation with Mucuna. IDRC and CIMMYT).

" As much as 40% of farmers interviewed by Buckles et al. (1992) reported that the mucuna system might induce localized landslides in areas of very steep slopes (superior to 60-70%). Discussion with farmers confirmed that such landslides occur once in a while (not every year) during the peak of the rainy season (anytime between Sept and Nov.) under very heavy rainfall conditions (several hundred mm in a few days). A possible explanation would include a combination of the heavy weight of the wet mucuna biomass, a loosening of the upper soil profile as a result of shallow rooting habits of the mucuna plant, and/or a state of superstauration of the soil resulting from increased infiltration, inducing a lower shear strength and higher overburden weight (Van Es, pers. com.). Some farmers also indicated that landslides might result from the lack of deep rooting or anchoring caused by the substitution of the traditional bush-fallow rotation for one with a fairly shallow-rooted species such as mucuna, furthermore quite able to eradicate the few trees left in place by farmers if left unpruned.

None of these explanations is completely convincing. The landscape in the mountains of Northern Honduras appears to be geologically very young, having not yet fully stabilized. Hence there are numerous areas where mass redistribution continues to take place "spontaneously", and sectors of abrupt slopes are among the prime candidates for being affected by this gravitydriven redistribution process (whether such a landscape should ever be subjected to large scale cultivation is definitely a relevant question). Also, one could argue that when quantities of water pouring on any landform reach hundreds of mm in a few hours or days, something dramatic is poised to happen, and the actual role of the velvetbean cover in causing a landslide is probably insignificant compared to the role played by the sheer masses of water rushing their way downhill. This may explain why landslides, when they take place, affect lands under all kinds of land use, from virgin forest to pastures to fields cultivated without velvetbean, without obvious preferential impact on any one category of land use. This could be seen when 400 to 700 mm fell in a 15 hour period on October 31st, 1993 causing countless landslides in the hillsides.

The issue seemed important enough to address it in a general survey of the velvetbean system conducted in the summer of 94. Farmers were specifically asked about the occurrence of landslides prior to and after the introduction of the velvetbean rotation in their fields. Out of 34 fields having suffered from landslides (from a total of 44 fields included in the survey), 62% (21) had had similar problems before velvetbean was ever introduced. Furthermore, only 1/3 of the farmers incriminated velvetbean in the occurrence of landslides. Perceptions varied strongly from village to village: in Piedras Amarillas, where landslides are common, farmers blamed velvetbean for making things worse, whereas in San Francisco de Saco, where landslides are rare, most experienced velvetbean users vehemently opposed this view."

This seems to present a clear picture of causes, and again the role of farmer perceptions.

A related issue would therefore be whether net benefits are likely to be negative; i.e. if a farmer is going to use a steep sloped field for a maize crop, does the real enhanced risk of greater soil erosion (of OM that would not otherwise have accumulated but for the Mucuna), or of landslips, outweigh the fertility etc. benefits of Mucuna?

Donald Kass wrote:

When Milton and I reviewed the Central American literature on mulching for the 2000 ASA meetings symposium--whatever happened to these papers, Lucy and Sheryl????, I found a paper by Cassel and others in Tropical Agriculture who had measured some physical characteristics of northern Honduran soils with and without the mucuna mulch. When I looked at the data, it indicated what Harold and the farmers say, that the mulch does produce some discontinuities in soil physical properties which would make the surface soil more susceptible to slumping. However, for some reason, the authors looked at the data and concluded that this wouldn't happen. I understood that CIDICCO was advocating planting dispersed trees in the mulch to reduce the possibility of slumping. Well, I guess I am out of this now--but it would be nice if what has been written could be published so everyone won't go on reinventing the wheel.

Donald's message confirms the contributions made by others.

Clearly there is an interesting rural development and farmer research agenda that could be developed by new projects, to study how to develop tree - Mucuna combinations. These would probably be based on working in fallow improvement as mentioned above.

However, almost no organisation has carried out substantive work in this area (either on the north coast or elsewhere in Honduras). I wonder why work on improved fallows and intensive dispersed trees systems has been so neglected by agronomists and foresters? (often in favour of alley cropping, despite the generally negative experiences with this system to date).

I can make available a draft paper (in Spanish) that I wrote on agroforestry research done on the Quesungual system in Lempira (for FAO). If any Mulch subscribers would like to review a copy, I can email this after 18th August. If interested, please send me an email.

Again, thanks for your contributions.

Best regards,

Torsten Mark Kowal