Arizona Game & Fish Dept. Texas Parks & Wildlife Dept.
555 N. Greasewood Rd. 114 Center Ave., #300
Tucson, AZ 85745 Brownwood, TX 76801
Carlos Hugo Alcalá-Galván Barry Hale
Instituto Nacional De Investigaciones Forestales New Mexico Dept. of Game & Fish
C. E. Carbό, Blvd. Del Bosque, #7, Col. 1 Wildlife Way
Valle Verde, Hermosillo, Sonoro, 83200 Santa Fe, NM 87501
Darrell L. Weybright Brian F. Wakeling
New Mexico Dept. of Game & Fish Arizona Game & Fish Dept.
1 Wildlife Way 5000 W. Carefree Hwy.
Sante Fe, NM 87507 Phoenix, AZ 85086
Len H. Carpenter Norris L. Dodd
Wildlife Management Institute Arizona Game & Fish Dept.
4015 Cheney Dr. 5000 W. Carefree Hwy.
Fort Collins, CO 80526 Phoenix, AZ 85086
On occasion over the years, desert wildlife and rangeland experts have told me that Allan Savory’s Holistic Planned Grazing (planned grazing) theories have been “disproved.” When I read this in your book, Habitat Guideline for Mule Deer, and finally saw specific studies cited, I decided to read them for myself. This paper is (1) a detailed examination of what I found (pp. 1-10; literature cited), (2) a correction of mis-perceptions about planned grazing (pp. 10-17), and (3) an analysis of why this most effective of range protocols is so fiercely resisted, and why this must change (pp. 18-22).
Every scientific citation contained herein appears in your book, Habitat Guidelines for Mule Deer, or was cited by, and necessary reading, to understand the authors whose studies you cite, with four additions (pp. 1-2 #1-4). Of these I am the only author you are not already citing.
In addition, I have relied on two authors (p. 2 #5&6) to shed light on the lack of objectivity of conventional range science in its evaluation of planned grazing.
The Circle Ranch Strategic Plan 2007-2011 by Christopher Gill, 2007. The Plan is a complete description of what we have done, and why, and how. It contains many references
, our grazing plans, planning methodology, forage results, wildlife (including mule deer) strategies, and much more. It is available to anyone, as indicated in the bibliography.
Holistic Management, A New Framework for Decision Making by Allan Savory and Jody Butterfield 1999 is the revised edition of Holistic Resource Management 1988 which your studies reference.
Holistic Management Handbook 2006 by Jody Butterfield, Sam Bingham and Allan Savory sets forth (1) theories and (2) specific techniques for financial planning, grazing planning, biological monitoring and land planning.
Twilight of the Mammoths by Paul S. Martin 2005. You rely on Dr. Martin for two citations (Heffelfinger et al. 2006, Martin et al.1984; Martin et al. 1998). One concerns the late Pleistocene (Quaternary) extinctions. In this book, Dr. Martin summaries his life’s work, and conclusions, on this subject.
The Structure of Scientific Revolutions by Thomas S. Kuhn, 1962, is one of the most influential books of the 20th century and a key text in the philosophy and history of science. In this book Kuhn explains that science does not progress via linear accumulation of new knowledge. Instead it undergoes periodic revolutions which he calls “paradigm shifts.”
John Preston’s Kuhn’s The Structure of Scientific Revolutions 2008, presents and analyzes Kuhn’s main themes in the context of forty-six years of examination and discussion.
Excessive Herbivory, page 11, states in part: "(Allan) Savory (founder of Holistic Management International, or, HMI) claimed that by grazing pastures intensively and moving stock frequently the range could actually be improved while simultaneously increasing the stocking rate. On some ranches it was even claimed that stocking rate could be doubled or tripled with improvements to range and livestock productivity. Researchers during the last few decades have shown these claims to be invalid (Heffelfinger et al. 2006; Holechek et al. 2000).”
To the contrary, this is precisely the outcome we have experienced. After eight years under planned grazing at our 32,000-acre high-desert mountain Circle Ranch in Hudspeth County, far-West Texas, we take almost triple the animal days of grazing (AD's) possible from conventional stocking rates recommended by NRCS and Texas Parks & Wildlife. Conventional practice dictates a herd of 250 head for 365 days: 250 X 365 = 91,800 AD's. This year we are running 1000 head for 240 days: 1000 X 240 = 240,000 AD's; plus, 50 head for 180 days: 50 X 180 = 9,000. This totals 249,000 AD's, 270% of conventional results, virtually the very result you say has been proven invalid. Records of grazing results over eight years demonstrate we are
experiencing consistent improvements in habitat, which we measure by changes in forage production. This year we will graze about half the ranch, moderately, as the rest recovers from last year's grazing (Gill 2007). We consider AD's to be the most appropriate benchmark by which to measure herbivory. With respect to stocking rate alone, we are at about 400% of the recommended level, and about 800% given we are grazing only half the ranch. Sometimes, depending on paddock size, we are at 250 times your recommendations! This is before adjusting
high animal densities and numbers to reflect that the herd is only in a particular pasture for a few days. Time management of grazing is the central practice of planned grazing (Gill 2007; Savory
1983; Savory et al. 1999; Butterfield et al. 2006). Most ranchers and range scientists would consider grazing at these stocking rates and animal densities to be “intensive.” Our animals will
be in about 20 paddocks for about 14 days each: most ranchers and range scientists would consider two weeks grazing every-other-year to be “short duration” (Gill 2007). But we are not running an SDG at Circle.
I. Right answers wrong questions Your conclusion has two parts: (1) planned grazing does not increase range productivity, and, (2) the increased hoof action of a large number of cattle associated with planned grazing consistently compacts soil and thereby decreases water infiltration (Heffelfinger et al. 2006). Both
conclusions may be correct with respect to short duration grazing systems (SDG's). Neither is valid with respect to planned grazing. No study cited, or relied upon, tested the outcomes of planned grazing (Holechek et al. 2000). Faulty premise faulty answer Bad assumptions cannot be fixed by regression analysis or technical language. If assumptions are wrong, so are the conclusions that rely on them. The Holechek study defines planned grazing as a SDG wagon wheel system, with shared water, no fewer than eight paddocks, 5-days-or-less grazing with 4-weeks-or more-nonuse: about 10 full cycles per year using a schedule of predetermined moves (Holechek et al. 2000).
Here is what Allan Savory said about such SDG systems in the August 1983 issue of Rangelands. "As a general rule, ... the government prescribed stocking rates can safely be doubled in the first year of operation as long as adequate time control is bought into the grazing handling (emphasis added)." "(Planned grazing) is not a grazing system. Anyone describing it as a grazing system is merely indicating that he has not yet understood the holistic approach to the management of all resources simultaneously, with constant monitoring and adjustment to achieve a goal." "Some say it (planned grazing) is a 'cell system' or the 'wagon wheel system'. Again this is totally erroneous and can only lead ranchers to costly error if they believe it and apply it as such." "Yet another serious misconception being spread, in Texas in particular, is that SGM (Savory Grazing Method, aka holistic grazing, aka planned grazing) is Short Duration Grazing (SDG). I recently inspected two grazing cells in Zimbabwe where SGM had been applied at a high and very successful level... over many years The operators (changed) ... and have now operated for four years under Short Duration Grazing, i.e., they operated as a grazing rotation with short grazing periods of 1 and 2 days length. Both cells have suffered as a result of the
inevitable over-grazing which has been very severe indeed. One currently contains no livestock and the other is desperately short of forage. In addition to the over-grazing the soil surface has reverted to a low successional level which has led to a considerable amount of new erosion..." "Ranchers should be very wary of Short Duration Grazing as even with development to the 30-and-40 paddock levels, and grazings as short as one and two days, severe over-grazing will occur." "Short grazing periods such as used in Short Duration Grazing which are not 'time-controlled' can only prevent over-grazing if long rest periods are constantly given (emphasis added) regardless of the growth rates of plants." "Short Duration Grazing must not in any way
be confused with (planned grazing). The Texas A&M claim ... that (planned grazing) is the Savory way of doing Short Duration Grazing is totally incorrect. ... I am extremely familiar with this (SDG) grazing system as I developed it. Texas A&M copied it from Zimbabwe.... I am not
prepared to go backwards to something I have found does not work and which I know they will eventually find out."(Savory 1983).
The foregoing statements of Savory’s are unambiguous. Yet, in 2000, citing Savory’s 1983 paper while dismissing its objections (which are consistent with hundreds of Savory's later writings and observations), including Savory’s statement that he had both developed, and later rejected SDG’s as unworkable, Holechek equates planned grazing with SDG systems anyway. According to him that is the “common conception” (Holechek et al. 2000). Proceeding from that re-definition he tests not Savory’s actual principles, but a mischaracterization. This 2000 paper is the one on which you base your conclusion that Savory advocates SDG’s, and that planned grazing been disproved (Heffelfinger et al. 2006). Holechek relies on studies summarized below (Holechek et al. 2000, Savory 1983). These studies share certain common deficiencies, primarily the equation of planned grazing with wagon-wheels-or-other unplanned SDG systems. These systems have frequent, short grazing periods and consequently shortened recovery periods (Savory 1983; Butterfield 2006 et al.). The grazing tests Jung examines a 105-day growing season program incorporating five grazing/rest cycles through 8 pastures (paddocks), with moves according to a pre-determined schedule. Grazing periods were 2-1/2 days and recovery periods 18-1/2 days long (Jung et al. 1985). Rotations were not determined by actual plant conditions. Recovery periods are inadequate: in West Texas, weather warms in late April but rains seldom arrive before July. Since weather cools again in mid-September, the effective growing season is 75 days. At our Chihuahuan Desert ranch two full seasons are usually necessary for complete recovery and sometimes on parts of the ranch three to five years may pass without adequate rain for plant recovery. So cattle moves and grazing according to monitoring actual plant conditions is critical. By definition (see page 12) any spring grazing causes over-grazing. Even if, and after, rains begin, an assumption of full plant recovery after only 18-1/2 days, let alone that there will be five such recoveries annually, postulates the impossible. This system overgrazes for all three reasons contained in the definition of over-grazing found at page 12. To avoid repetition, consider the foregoing comments as implicit in any repetition of the next statement: “This was not a test of planned grazing.” Pitts tested an SDG described as 118 acres in 16 paddocks, 7.3 acres each, under year-round stocking of between 6-9 animals in the whole herd, against a continuously stocked comparison. Grazing period of the SDG was 2-7 days followed by 30-60 days rest. SDG was no better than continuous grazing (Pitts 1987 et al.). Using an average of 3.5 days grazing and 45 days rest indicates 7.3 cycles per year. Deficiencies include lack of grazing planning, pre-determined moves, spring grazing, inadequate recovery times, too many grazing periods. Together, these cause over-grazing. This was not a test of planned grazing. Planned graziers would say both protocols tested will harm plants.
Anderson 1988 examined different grazing approaches in the Jornada del Muerto of southern New Mexico. He found: “… flexible rotation produced more uniform use of green tobosa compared to grazing with fixed intervals between rotations. Flexible rotation improved plant
quality. Conservative continuous stocking during the growing season, the traditional method, is the (worst). … Conservative stocking and slow rates of plant decomposition are two factors that contribute to the accumulation of excess standing dead forage in arid ecosystems. Fixed interval rotation of animals among paddocks also may not provide high quality forage …. Rotations with pre-determined fixed grazing periods produced overuse in some and underuse in other areas. Due to the many variables simple analysis is difficult (Anderson 1988). None of these excellent points demonstrate deficiencies of planned grazing: Flexibility produces better outcomes. Low stocking through the growing season is the worst of all choices studied. Unplanned rotations
produce over-rest and over-grazing. Reductionist analysis is not successful in the deep deserts of the southwest. These comments could be notes from an HMI planned grazing seminar. Thurow studied different grazing methodologies on the Edwards Plateau. He found plant cover declined under SDG. Several measures are offered for this phenomenon but all reduce organic cover and above ground biomass. Cover and above ground biomass decline during the dormant season and drought for all strategies. The SDG pasture had significantly lower organic cover when there was less rain. The exclosure did not increase until the second year of favorable rainfall (Thurow et al. 1986). The system relied on 14 pastures, one herd, four days of grazing and 50 days of rest. However, the system was changed during the study. The supposition of complete recovery during 50 days in, for example, the winter may cause over-grazing as does spring grazing. If plants are dormant during the winter (not the case everywhere), then multiple grazing periods during the winter do not imply over-grazing. If the plant was fully recovered upon the commencement of winter dormancy, multiple grazing events during the dormant season don’t harm the plant. In the planned grazing Aide Memoire planning book, we refer to these dormant season grazing events as “selections.” At Circle we could graze each pasture more than once if we wanted to, taking a percentage of the forage on hand with each selection. However, in big, rugged country, this is usually logistically challenging. The only reason to do this is for the sake of animal performance, since multiple, shorter grazing periods are usually better than one long grazing period. In the high rainfall brittle tropics — productive and easier to manage than big deserts — this is often the approach, since it is so hard to maintain animal performance in these environments. In lower rainfall desert environments with much higher quality forage, it’s easier to maintain animal performance with only one dormant season selection. Because of this, and because of the logistical challenges mentioned above, we’ve never considered multiple dormant season selections at Circle. This is a fairly detailed topic but
the key point is a plan that addresses the unique particulars of each ranch (p. 15). The shorter the grazing period, the shorter the recovery period and inadequate recovery periods always cause over-grazing because they put animals back on the plants before the plants have fully recovered. Does not test planned grazing. White 1991 compared a nine paddock SDG cell with a continuous-grazed pasture over a five-year period in south central New Mexico. He found little difference between the two methods. SDG was no better than continuous. Basal cover was somewhat higher under SDG but the differences were small. Because of varying conditions in different parts of the test (reductionist based) conclusions are highly subject to the interpretation of the interpreter (White et al. 1991).
This is a SDG study. There is no planning process described. While neither grazing nor recovery period data are included, probably recovery periods were inadequate, given our experience in similar 15-inch rainfall deserts. Recovery periods in arid and semi-arid
environments are generally inadequate if shorter than 1-2 full growing seasons. Planned graziers would say that continuous stocking, or any other protocol that returns animals to plants before they have fully recovered, or which grazes with animal numbers insufficient to achieve uniform grazing in a short period of time, will be harmful. Not a test of planned grazing. Manley et al. 1997 conducted a 12-year study on rangelands near Cheyenne, Wyoming. This compared continuous, to SDG methods. They concluded no effect on above or below biomass. He relies on Holechek’s representation that SDG’s and Holistic Resource Management (planned grazing) are the same thing. The SDG paddock was subdivided into eight paddocks. No data
was furnished on grazing/recovery period. Both methods caused total plant cover including lichens and mosses to decrease sharply in the first few years of grazing. This is because this plant community had been ungrazed for forty years prior to the beginning of the grazing. Root biomass increased under those protocols that involved some deferral (recovery) as compared to no deferral (recovery) whether lightly or heavily stocked. The authors conclude that increased forage is partly the result of better management, i.e. more efficient utilization of forage. Cross-fencing and water development are important to achieve this improved utilization but they must be installed on large areas and the installation of the fencing must be accompanied by adequate water development. Management of the forage base is the key to improved livestock performance. This must include monitoring of the vegetation. This does not need to be elaborate but must be implemented consistently. The use of proper monitoring tools will allow the producer to adjust stocking rate and time and length of grazing to reflect a forage capability of that particular year (Manley et al. 1997). While this is a study of SDG’s and not of planned grazing, many excellent points are made that are consistent with planned grazing predictions and experience.
Dormaar et al. A test of the hypothesis that animal impact can improve the nutrient and water status of the soil and promote grassland succession. A 17-pasture SDG system was established in Alberta, Canada. 18 inch rainfall. Average grazing periods 2 ½ days, average rest 40 days. Longer grazing and rest periods were given in the dormant season. Plants and soil moisture worse than in ungrazed exclosures. Savory states (Savory 1983) that stocking rates can be doubled or tripled regardless of range condition. Short-duration grazing caused degraded
infiltration of water and degraded condition of plants (Dormaar et al. 1989). This is a study of an SDG, year- round system: not of planned grazing. Longer recoveries help growing plants, not dormant ones. Planned graziers state this system will harm plants and infiltration.
Taylor et al. tested SDG’s in three forms and concluded that none worked very well. They caution against using range protocols developed in moist environments on dry rangelands. They state that it is necessary to control the frequency and intensity of plant harvest. “Grazing management directed toward solving grazing distribution problems in determining optimum plant harvest are two positive practices on semi-arid rangelands which can increase livestock production within economic bounds. We feel that long rest periods are needed during the major part of the growing season to allow the more productive mid-grasses to recover from grazing. Grazing systems are certainly worth the trouble when they facilitate the implementation of
biologically and economically sound grazing management principles. The appropriate system will vary from one ranch to another, depending on the goals and objectives of the manager and the resources of the land (Taylor et al. 1993).” These are all excellent points with which planned graziers agree.
Heffelfinger'>Infiltration Studies Heffelfinger: “Long-term changes resulting from over-grazing include undesirable changes in the plant community, decreased mulch cover, decreased water infiltration, compacted soil, increased water run-off, decreased plant vigor and production of a dryer micro-climate at ground level (Heffelfinger et al. 2006; figure 13, Severson and Medina 1983: 24).” This statement is correct. Not a test of planned grazing.
Heffelfinger: “The increased hoof action of a large number of cattle was shown consistently to compact the soil rather than increase water infiltration as claimed by Savory (1988) (Heffelfinger et al. 2006).” This is perhaps correct if you mean SDG’s. Wrong if you mean planned grazing; this was never tested. The infiltration tests Holechek concludes that several studies prove that Savory’s claims that short duration grazing will increase water infiltration into the soil compared to continuous grazing are disproved … and that several studies have been quite consistent in showing that hoof action from having a large number of animals on a smaller area for short time periods reduce rather than increase infiltration (McCalla et al. 1984; Thurow et al. 1986; Weltz et al. 1986; Warren et al. 1986; Pluhar et al. 1987). These same studies have also been consistent in showing short-duration grazing increase erosion compared to continuous or season long grazing (Holechek et al. February 2000).
Short-duration grazing (SDG) is not planned grazing (Savory 1983; Savory et al. 1999; Butterfield et al. 2006). Let’s look at each nevertheless. McCalla tested SDG’s. McCalla found that “livestock grazing can alter infiltration rates of rangelands soils by removing protective plant cover and by trampling. … Reduced infiltration
rates as a result of livestock grazing have been attributed to (1) loss of vegetation cover, (2) decreased mulch cover, (3) decreased amounts of vegetation standing crop and mulch, (4) increased bare ground and (5) increased bulk density as a result of trampling” (McCalla et al, 1984). These observations are correct and predicted by planned grazing protocol. SDG causes over-grazing and degradation of plants.
The runoff plots were pre-wet until they would not take any more water, then allowed to rest for 24 hours. Then the moist soil was tested for relative infiltration (McCalla et al. 1984). McCalla’s methodology fails with respect to the most important aspect of planned grazing and soil trampling: the breaking of the crust that otherwise keeps rainfall from initially soaking into the top layer of soil (Thurow et al. 1986). Observers of rains in deserts know these generally fall on dry soils where capping, if present, sheds water. Once capped soils become wet in the desert,
water soaks in more. Testing uncapped wet soils to evaluate how capped dry soils absorb sudden rains is not a test of real conditions. Nor is it a test of the most important aspect of hoof
action which is breaking the soil crust. Planned graziers will stipulate to the self-evident: infiltration varies directly with plant community health. It is the foliated plant, its root mass, liter cover, the associated insect tunnels, microorganisms, organic soil content, and lack of capping that slows, traps, absorbs, and allows water to soak into the soil instead of running off
(Savory 1983; Savory et al. 1999; Butterfield et al. 2006; Heffelfinger et al. 2006; Thurow et al. 1986).