by Brian Tomasik
First written: Jan.-Feb. 2016; last update: 3 Dec. 2017


Livestock grazing takes up 26% of the Earth's ice-free land area, making it one of the biggest impacts that humans have on wild-animal suffering. Several studies suggest that grazing reduces wild-insect populations (and hence, probably wild-insect suffering) on pasture fields, but some studies point in the opposite direction. Cattle poop supports large numbers of insects, which adds uncertainty to the overall analysis. The fields of organic cattle farms generally have higher insect densities than those of conventional cattle farms.

Whether cattle grazing increases or decreases total invertebrate populations on a given piece of land depends in part on whether the pasture is irrigated. For example, converting formerly dry land in the Western USA into green pastures would likely increase total invertebrate populations.

Note: This piece is one of a series of articles on the wild-animal impact of cattle grazing, and not all relevant considerations are discussed here.

A note on terminology

In this piece and in general, I use "cow" in a gender-neutral way as a singular form of "cattle".

Also, when I say "insect", I usually mean "any terrestrial invertebrate animal".

When I first wrote this piece, I tended to use more familiar if less technically correct words, like "cow" to mean "a domestic bovine of any sex" and "insect" to mean "invertebrate". In my later additions to this piece, I shifted toward the more proper terminology.


In "Least Harm: A Defense of Vegetarianism from Steven Davis’s Omnivorous Proposal", Gaverick Matheny contends (p. 509) that one argument against cattle grazing is that "by using land and resources formerly or otherwise available to wild animals, grazing cattle prevent some number of smaller wild animals from existing." This page echoes:

Grasses provide food, shelter, and even construction material for hundreds of desert animals ranging from jackrabbits to tiny insects, each of which is eaten in turn by other animals. Send in a wave of cattle to crop those grasses and we've diverted that ecological productivity to our own ends, depriving the local wildlife of food and habitat.

However, if we think wild animals on the whole endure more suffering than happiness in light of Cattle grazing off National Cycle Route 23 near Newchurchhigh infant-mortality rates and short lifespans prior to painful deaths, then this argument operates in reverse, favoring cattle grazing.

However, there remains an important empirical question: Does cattle grazing in fact reduce wild-animal populations? A priori this isn't obvious. For example, one might think that cattle poop is a better food source for insects than grass. On the other hand, if insects ate grassland vegetation directly, they'd derive more total energy from it than they would from poop, since a lot of energy is wasted on a cow's respiration. And much of the meat from cattle is not eaten by decomposers on the pasture land but by humans or other vertebrates, which may further reduce food available to insects.

The fact that plant food can be eaten either by insects or by cattle is highlighted in the following passage from a report on grasslands in the southwestern United States:

Hewitt (1977) and Hewitt and others (1976) stated that grasshoppers compete with cattle for forage and that both herbivores have similar preferences in grass species. Estimates of the biomass consumption on rangelands in the Western United States by grasshopper herbivory have been difficult to calculate, but typically range from 6 to 12 percent of available forage, to as much as 50 percent in certain areas (Loftin and others 2000). In grassland at the San Carlos Indian Reservation in Arizona, Nerney (1961) found that grasshopper consumption ranged from 8 to 63 per cent of the vegetation. However, Hewitt (1977) also reported a study from southeastern Arizona showing no correlation between grasshopper density and loss of forage.

Impacts on insect populations in temperate climates

Studies finding that grazing reduced insect numbers

Cole, 1946

This study compared invertebrates on grazed and ungrazed portions of woods. "These woods have been grazed by cattle for many years and are largely devoid of underbrush and second growth except for one small portion which has been fenced and protected from grazing since 1935" (p. 52). The study reported the following table:

The study explains (pp. 67, 84):

From Table 7 it is obvious that most of the forms were most abundant in the ungrazed portion of the woods. [...]

The amount of organic matter and the water-holding capacity of the soil was highest in the ungrazed portion of the woods with the exception of Area D which was in the grazed portion but placed on soil transplanted from the ungrazed region. [...]

Very few forms were most abundant in the grazed part of the woods. The difference with respect to the large carabids is of doubtful significance but the roaches, crickets, and the unidentified psammochorid wasp were definitely more abundant on the grazed land. [...]

The majority of the cryptozoic species reached their maximum abundance in the portion of the woods which had longest been protected from grazing by cattle and some evidence suggests that this was largely an association with relatively humid soil. A few animals, notably the orthopterans, were more abundant in the drier grazed portion of the woods.

Fielding and Brusven, 1995

This paper found that

Mean grasshopper density was higher on ungrazed sites than on grazed sites. [...]

Results indicate that livestock grazing during drought conditions tends to reduce grasshopper populations on southern Idaho rangeland.

Rambo and Faeth, 1999

This study found the following results:

Though the differences weren't statistically significant, the trend of much greater insect abundances on non-grazed land seems clear.

Kruess and Tscharntke, 2002

This study also found that grazing reduced insect numbers:

The authors note (p. 300):

the fourfold cattle density appeared to affect the plant–insect associations by disrupting the continuity of feeding by phytophagous insects. Presumably, insects of low abundance should experience most difficulty in persisting locally in a highly disturbed environment. Moreover, small-scale recolonization may be limited by the dispersal capabilities of different insects (Cornell and Lawton, 1992; Hanski and Gyllenberg, 1997).

Pöyry et al., 2004

This study notes:

Large differences in responses to grazing effects have been reported in studies of grassland insects. These discrepancies may be connected with dissimilarities in life histories among insect groups and species (Curry 1994). [...] In the majority of empirical studies in grasslands, grazing has been observed to cause negative effects on insect populations, resulting in modified and often impoverished community structures (Morris 2000, Swengel 2001). However, positive (Smith 1940, Holmes et al. 1979) or unimodal (Smith 1940, Hutchinson and King 1980) responses to increasing grazing intensity have also been reported.

That study itself found the following trends for numbers of individuals (pages 1656-57 and 1661, Table 1):

"old continuously grazed pastures" "restored pastures
with ~5 yr of reinitiated grazing"
"abandoned former pastures", "where grazing ceased >10 yr ago"
126.2 ± 57.9 126.0 ± 73.8 306.3 ± 141.8

So it seems insect numbers are at least twice as high on land that's not being grazed.

The authors wrote (pp. 1662 and 1666-68):

The simplest community measures, i.e., species richness and abundance (number of individuals), were highest in abandoned pastures. This result was similar for all species and for grassland species. A posteriori comparisons between pairs indicated that the main difference occurred between grazed and abandoned pastures, and that there were no differences between the two grazed types. [...]

The abandoned semi-natural pastures were found to possess the highest species richness and abundance of butterflies and moths. Grazing history was included in all our GLM models as the strongest variable explaining variation in patterns of species richness and abundance among the studied sites. These results are similar to the conclusions made in two recent reviews on the effects of management in structuring insect communities in grasslands (Morris 2000, Swengel 2001). [...] The decreased diversity of insects in pastures is usually attributed to direct and indirect disturbances caused by grazing animals, leading to decreases in structural diversity of vegetation and in the number of niches available to invertebrates (Morris 2000). Grazing can also disrupt food chains between trophic levels (Tscharntke 1997). In our case, a probable explanation for the observed decline of species richness and abundance in pastures is that many abundant species were largely confined to abandoned pastures, and they quickly diminished after the onset of grazing (J. Pöyry, S. Lindgren, J. Salminen, and M. Kuussaari, unpublished manuscript). [...]

Among invertebrates, reported cases of highest diversity or abundance at intermediate grazing intensities are much less common (Fuentes and Jaksic 1988). Nevertheless, despite the scarcity of studies comparing the effects of differing grazing intensities, such cases have already been reported for a number of taxa (Smith 1940, Hutchinson and King 1980, Roberts and Morton 1985, Bestelmeyer and Wiens 1996). Usually, an increase in insect species richness and abundance has been observed in grasslands with decreasing grazing intensity or increasing intervening time since last the grazing event (e.g., Kruess and Tscharntke 2002a, b). [...]

our result of the highest species richness and abundance in abandoned pastures clearly challenges the goal of enhancing species richness by intensive management.

The paper presents the following figure (p. 1667):

While the authors say that this shows a "unimodal relationship, in which the highest abundance in the year 2000 occurred at an intermediate vegetation height of ~40 cm", you can see that all of the highest abundances were in abandoned (non-grazed) pastures.

Debano, 2006

This study reported:

Vegetation-associated insect communities were found to be sensitive to livestock grazing. Overall abundance of these insects was lower on grazed grasslands, and certain insect orders appeared to be negatively affected by livestock grazing; beetles were less rich, flies were less diverse, and Hymenoptera were less rich and diverse on grazed sites. Conversely, Hemiptera were more diverse on grazed sites. [...] When compared to other studies conducted in areas of the United States that fall within the historic range of bison, this study suggests that invertebrates in areas outside this range may be more sensitive to grazing pressure.

Studies cited in Gudleifsson and Bjarnadottir, 2008

This piece reviews some studies regarding the impact of grazing on springtails (Collembola), which are some of the most numerous soil animals:

Management of grassland, including fertilization, grazing and cutting, creates a major disturbance to the sward structure and therefore to the microarthropods living there. [...] In a study of soil-living microarthropods, collected by a Berlese Tullgren funnel, Siepel and van de Bund (1988) found that nitrogen fertilization had a major influence on the species composition of Acari and Collembola populations, while mowing and grazing had only a minor influence. Petersen et al. (2004), on the other hand, found that grazing significantly reduced the abundance of total Collembola. [...]

Sigurðardóttir (1991) established that land reclamation with grass seeds and fertilizer and protection from grazing increased the number of Collembola, especially Isotomidae.

Studies finding that grazing increased insect numbers

Zhu et al., 2012

This study, surprisingly, found higher insect abundances with grazing, although the increases with cattle grazing specifically were modest:

Note that "control" means "no grazing". The authors explain:

Insect abundance was higher in grazed than ungrazed plots, possibly due to more available oviposition sites (Fisher 1994), more new growth to attract insects (Gao et al. 2008) and changes in microclimate favouring hatching of insects (Loftin et al. 2000) following grazing.

The study also helpfully summarizes some of the other literature on this subject:

Leaf-feeding insects can be depressed by large herbivores because they compete for the same plants (Bailey & Whitham 2003). Similarly, vertebrate herbivores reduce the abundance of caterpillars and grasshoppers (Huntzinger, Karban & Cushman 2008) and beetles (Dennis et al. 2008) because of limited food supplies and defensive structures of plants induced by grazing. In contrast, grazing by large herbivores may also facilitate insect diversity (Cagnolo, Molina & Valladares 2002). For example, bison grazing has been shown to result in higher grasshopper diversity by increasing plant species richness (Joern 2005).

(Keep in mind that as far as insect suffering, abundance rather than diversity is what matters.)

Studies finding mixed results

Wolcott, 1937

This study examined invertebrate numbers on three types of grassy fields: a meadow used to produce hay (no grazing), a lightly grazed pasture with "abundant vegetation", and a heavily grazed pasture with "scant vegetation".a The study found (p. 59):

A total of 6,843 invertebrates was collected from the hundred square of grasslands at Barneveld, New York, averaging 63.23 per square foot of meadow, 66.74 per square foot of pasture of scant vegetation and 86.74 per square foot of pasture of abundant vegetation.

This isn't an ideal comparison because the meadow was mowed for hay rather than laying idle. Reduced demand for beef would cause more grassland to return to being idle, not mowed for hay.

Helpfully, the study also confirmed the basic idea that as cows eat more vegetation, less is left over for invertebrates (p. 81):

In the pasture of abundant vegetation, the most important insects in the consumption of forage which the cows might have eaten were not those which specifically fed on grass or clover, but those which fed omnivorously on all sorts of vegetation: grasshoppers, crickets, white grubs and crane-fly larvae. On the supposition that they ate no more of the weeds than they did of grass and clover, they prove to be the dominant factor in indicating a consumption by insects in this pasture of approximately twice as much grass and clover by insects as by cows.

In the pasture of scant vegetation, the cows ate very considerably more of the grasses and clover than did the insects, relatively because the insects were less abundant, and absolutely because they were there long enough and in sufficient abundance to eat most of the surplus vegetation. This kept the vegetation short, so that this pasture was a less attractive environment for crickets, grasshoppers and leafhoppers, and also it was more attractive to robins, who foraged there in greater numbers and thus further reduced the number, not only of Carabid and Histerid beetles, but of all large insects. [...]

If grass and other vegetation is not eaten by cows and phytophagous insects, it eventually becomes available for the plant scavengers, such as snails, slugs, sowbugs and millipedes. The small number of such animals found in the pasture of scant vegetation, which is further greatly reduced if those occurring in dung are subtracted, is an index of the completeness with which these plants were consumed as living tissue, while the large number present in the other pasture shows how much more surplus vegetation was present than could be utilized by the cows and insects. The greatest amount of decaying vegetation was present in the meadow, where there were fewer insects, and the single mowing did not begin to make up for the absence of cows.

The study also explains (p. 3):

it was found that where there were few cows in the pasture, the insects ate more of the grasses and clover than the cows did. And the cows succeeded in obtaining a larger share of the pasturage only when they kept the pasture so short that it offered scanty protection to the crickets, grasshoppers and leafhoppers, and was more attractive to the robins which foraged there in greater numbers and further reduced the numbers of insects. In the pastures examined, the grasses and clovers constituted only a third of the total vegetation, the other two-thirds being weeds, most of which the cows refused to eat. But the insects feeding on the weeds were only a third as numerous as those feeding on the grasses and clovers, and most of them were insignificant in size. The orchard grass and the red clover of the meadow were hosts for comparatively few insects, which, together with a single mowing, removed only a small part of the plant tissue produced. Thus the plant scavengers: millipedes, sowbugs, snails, and slugs, flourished in greatest abundance.

Capinera and Sechrist, 1982

This study reported:

Total numbers of grasshoppers were significantly higher in high biomass pastures (ungrazed or lightly grazed by cattle), while numbers of grasshoppers in the subfamily Oedipodinae were significantly higher in low biomass pastures (moderately or heavily grazed). Grasshopper numbers in the phytophilous subfamilies Gomphocerinae and Catantopinae were positively correlated with grass and forb biomass, while numbers of the geophilous Oedipodinae were negatively correlated with all biomass components. It would seem possible to modify grasshopper abundance through regulation of cattle grazing intensity, but changes in abundance would also be accompanied by shifts in the grasshopper species complex.

WallisDeVries and Raemakers, 2001

This study examined very low-intensity grazing:

Changes in butterfly abundance were positive in grazed and ungrazed areas compared to cut areas. [...] No clear negative effects of grazing were observed, but species occurrence was not always positively related to the environmental characteristics associated with grazing. In the long run, even lower stocking rates might prove more beneficial to the butterfly community as a whole.

A caveat about soil insects

As far as I can tell from skimming and sometimes fully reading the above studies, many of the above studies focused on aboveground insects. But grassland plants also support bug populations underground. Might grazing increase underground bug populations?

If grazing reduces aboveground bug populations due to the mechanical disturbance that cattle create, then ignoring the vegetation eaten by the cattle themselves, there might be more remaining vegetation not eaten by aboveground bugs that will instead be eaten by soil decomposers, including earthworms, springtails, and so on. How cattle grazing affects the total amount of vegetative matter available for eating by cattle and decomposers is a complex issue, discussed more here.

Impacts on insect populations in tropical rainforests

Unfortunately, I haven't yet found studies on how grazing affects insect abundances in tropical forests. For now, here are some studies about how other kinds of land disturbance impact insects.

Species richness

Many studies of land-use effects examine species richness rather than abundance of individual organisms. Still, insofar as there's generally some positive correlationb between numbers of individual organisms and numbers of species, the following data points may have some value.

Schulze et al., 2004

This study examined species richness in Central Sulawesi, Indonesia:

Cattle grazing is not one of the land uses surveyed, but insofar as pasturing may be preceded by slash-and-burn clearing of more mature rainforest, the general decrease in species diversity with younger forests is relevant.

That said, the authors point out some counterexamples to the trend that more human disturbance of land decreases diversity (p. 1327):

Forest disturbance can also cause an increase of species numbers on a small-scale in insects. Davis et al. (2001) showed that small-scale species richness of dung beetles in logged forests may be higher than in primary forest due to the presence of overlapping species ranges that are usually spatially separated in primary forest. Trap-nesting bees and wasps may even become more diverse with increasing land-use intensity (Klein et al. 2002).


Ewers et al., 2015

This article examined the impacts of logging on vertebrates and invertebrates in rainforest. Once again, logging is not grazing, so the results don't transfer perfectly, but both are forms of disturbing the rainforest, so this information is weakly relevant. The paper found that logged forest is "associated with decreased abundance of key functional groups of termites, ants, beetles and earthworms, and an increase in the abundance of small mammals, amphibians and insectivorous birds in logged relative to primary forest." However, the paper's Figure 2 shows higher total invertebrate biomassc in logged forest. The authors explain:

total invertebrate biomass was doubled [...], in logged relative to primary forest. The increased invertebrate biomass is predominantly comprised of large herbivorous invertebrates [...]. These herbivores do not contribute directly to the ecosystem functions we examined, suggesting that herbivory may be one function that runs counter to the general trend for invertebrates to decline in importance in logged forest.

Combined with the increase in most vertebrates, this seems unsettling and suggests that pristine rainforests may not be at the peak of animal abundance.

Fortunately, it seems that habitat disturbance does generally reduce animal abundance, as this paper explains: "among pairs of disturbed and undisturbed sites globally, Lepidopteran species richness is on average 7.6 times higher in undisturbed than disturbed sites, and total abundance is 1.6 times greater [...]." This result was based on a meta-analysis where most studies showed a decrease in diversity in disturbed areas.

Does cattle poop support lots of bugs?

One way in which cattle grazing might increase insect abundances is if cattle dung is sufficiently more edible than raw grass that it supports more total bugs. Of course, raw grass has more energy than dung, since a cow extracts some energy for itself during digestion. But if the dung is more insect-friendly than the grass, it's possible the dung could support more insects. Following is a rough calculation aiming to assess this worry.

Decrease in bugs due to grazing

There are ~1018 bugs on Earth, and Earth's total land area is about 1.5 * 1010 hectares. This gives an average of 6.7 * 107 bugs per hectare. Some land regions, like desert and tundra, have very few insects, while others, like tropical forests and grasslands, probably have more than the global average. To be conservative, suppose a temperate pasture has only 107 bugs per hectare, and say that (just making up a conservative number) only ~1/5 of them are above ground: ~2 * 106. This could easily be off by at least an order of magnitude.

How much of pasture vegetation is actually eaten by grazing cattle?

  • This textbook says "The percentage of annual above-ground primary production utilized by herbivores varies greatly, but estimates generally range between 20 and 50% (Scott et al. 1979, Detling 1988)."
  • Mekonnen and Hoekstra (2010) mention (p. 37): "the harvest efficiency – the fraction of grass actually consumed by the animal compared to the standing biomass – is quite small. In a recent study in the USA, Smart et al. (2010) showed that, depending on the animal stocking density, harvest efficiencies reach between 14-38%."
  • This document suggests that a typical harvest efficiency (fraction of plants actually eaten by cattle) is 25% for "Continuous, Season Long" grazing.

Assuming the number of insects a pasture can support is proportional to plant productivity on that pasture, then grazing at, say, 25% efficiency would reduce insect numbers by 0.25 * 2 * 106 = 5 * 105 per hectare. If the grazing season is, say, 0.5 years long, that's a reduction of 0.5 * 5 * 105 = 2.5 * 105 insect-years per hectare grazed per year.

Here's an alternative calculation of the insect populations of grassland based on how many insects could be supported by how much primary production. It estimates that ~4.5 million housefly-sized bugs could be supported on a hectare of grassland, although this number includes primary production that's eaten by bacteria and other non-insects. Perhaps the number of actual insects supported would be closer to the 2 * 106 estimate used two paragraphs earlier.d It's nice to get roughly the same result using both a bottom-up calculation like this one based on productivity per hectare and a top-down calculation based on worldwide average insect densities.

Increase in bugs due to dung

Here are some estimates of cattle dung production:

  • Answers here are mostly in the range of ~24 kg to ~42 kg per day.
  • This page suggests that a 1000-pound beef cow produces 59.1 pounds of manure per day, which is 27 kg.

A typical density of cattle on pasture is 1.8 acres per cow, i.e., 0.73 hectares per cow. So one hectare can support 1.4 cows. If they each produce, say, 30 kg dung per day, that's 41 kg of dung per hectare per day.

This study created fake "cow-dung pats" of 1.5 kg each, so I'll assume that's a typical size. Then we have 41/1.5 = 27 cow dung pats per day per hectare.e

A pecked pat^ - - 1623003This study found an average of 101.82 insects per cow dung pat on conventional farms. And this study found that pats lasted "57–78 and 88–111 days in spring and summer, respectively". Approximate this as 90 days, i.e., ~0.25 years. Assuming 101.82 insects remain on the pats the whole time, that's 101.82 * 0.25 = 25 insect-years per pat.

(27 pats per day per hectare) * (25 insect-years per pat) = 700 insect-years per day per hectare. And again assuming that grazing lasts 6 months out of the year, that's (365/2 days per year of grazing) * (700 insect-years per day per hectare) = 1.3 * 105 insect-years per hectare per year of grazing.

This is exactly half the reduction in insect-years due to grazing calculated in the previous subsection. This is remarkable because I chose all of the input assumptions to these calculations before seeing the result and didn't fiddle with the numbers to make them come out right.


The values for lost insect-years due to plant-productivity appropriation vs. increased insect-years due to poop are basically equal relative to the level of noise in both estimates. That said, I weakly suspect that the reduction in insects due to cattle eating plants is a somewhat bigger effect because I was pretty conservative with my input assumptions for that calculation and because the studies cited earlier in this piece generally suggested decreases in insect populations with grazing. On the other hand, the estimate of 101.82 insects per cow pat might also have easily missed a bunch of small insects? Also, maybe the bugs that eat poop are on average bigger (and hence more morally important) than the average bug (given that most bugs in the world are quite tiny)?

Insect-years prevented per kg beef

If we take the numbers above at face value despite the high uncertainty, we would conclude that cattle grazing prevents 2.5 * 105 - 1.3 * 105 = 1.2 * 105 insect-years per hectare grazed per year. I assumed 1.4 cows per hectare, so this is (1.2 * 105) / 1.4 = 9 * 104 insect-years per cow-year. And assuming a US beef cow reaches 1200-1400 pounds (average = 1300 pounds) and lives 18-22 months (average = 20 months), that's (20/12) * 9 * 104 insect-years prevented (ignoring the fact that young cows eat less than adults) for ~1300 pounds of beef. That's 115 insect-years per pound or ~250 insect-years per kg of live-weight cow. This piece estimates that a 1200-pound steer might yield 490 pounds of boneless trimmed beef. So (250 insect-years per kg live weight) * (1200 kg live weight)/(490 kg edible beef) = 600 insect-years per kg edible beef produced. And then multiplying by ~0.5 as a "cumulative elasticity factor" gives ~300 insect-years per kg edible beef bought.

Of course, this calculation ignores other impacts of cattle grazing, like growing hay to feed the cattle during the ~6 winter months, grain cultivation for feedlot food, soil erosion, climate change, eutrophication, etc.

300 insect-years per kg edible beef bought is quite a bit lower than the median estimate of ~5 * 105 insect-years of suffering prevented per kg of Brazilian beef calculated in this piece. Probably this difference is partly accurate, insofar as (1) temperate grasslands are less fauna-dense than tropical forests and (2) US cattle have shorter lives than Brazilian cattle. But probably a lot of the difference results from my taking pretty conservative estimates for aboveground insect abundance in the preceding calculations. Indeed, in this piece, I assumed just ~2 million aboveground bugs per hectare on pasture fields, while in the Brazilian-beef calculations, I assumed (using that piece's terminology and parameter settings as of 2 Mar. 2016) this many bugs per hectare: u * N / A = 4.5 * 1018 / (1.5 * 1010) = 300 million, which is two orders of magnitude higher.

Other info on poop

This page notes that "Lignin is not digested by the rumen microbes."

This study "assumed that cattle and buffaloes excrete 35% and all other grazers excrete 25% of dry-matter feed intake (31, 32)". (I assume the remaining organic matter is respired as carbon dioxide, methane, etc.?) If poop is more nutritious per gram than undigested plant dry matter (is that true? maybe the opposite is true?), then maybe the nutrition left over in cattle poop is slightly higher than 35% -- perhaps closer to 50%? The above calculations also implicitly assumed that cattle poop supports about half as many bugs as the forage that cattle eat would.

This piece reports: "Manure, which generally increases invertebrate activity, has in some cases reduced the number of invertebrates in grasslands (Curry et al. 1980)."

Springtail-years prevented per cow

As discussed above, I'll assume that when a beef cow eats a gram of vegetation, half of the energy/nutrition is removed by the cow and half ends up in poop.

This book says: "The community of insects and other organisms in a dung pat is more complex than you might expect. Fungi grow in the dung and derive their nourishment from it. In turn, certain springtails and mites feed on the fungi [...]."

This page says:

Springtails live in air-filled pores mainly in the top 10 cm soil, in the leaf litter layer, in dung pats, compost heaps and any rotting animal and plant matter. Though they are tiny, they can occur in very high numbers – I once collected the equivalent of 1.5 million per square metre under a dead bird, but usually they are around 25,000 per square metre in grassland soil. Some live in vegetation, in leaf litter and dung or only in the deeper soil depths but they can move up and down the soil/litter/dung/vegetation profile depending on weather.

This piece estimates that each gram of grassland vegetation dry matter can create 3 to 6 springtail-years, plus life for other invertebrates. Per gram of detrital vegetation specifically, the number may be about twice as high. Since I'm assuming that a cow effectively removes half of the organic matter that it eats, I'll make the possibly conservative assumption that one gram of dry matter consumed by a cow prevents only 3/2 = 1.5 springtail-years from being created.

This page reports that a typical "dry cow" (i.e., a female cow toward the end of her pregnancy) consumes roughly 28-30 pounds (average = 29 pounds) of dry matter per day when eating "Green pasture".f This page says "a 1000-pound (454 kg) cow, with or without an unweaned calf, is one animal unit, with such a cow being assumed to consume 26 pounds (about 12 kg) of forage dry matter per day." I assume the dry-matter intake is roughly similar for beef cattle? This document gives as an example: "a 400 kg steer eating 10 kg of dry matter".

I'll use the figure of 29 pounds per day. Of course, the number will be smaller when the cow is younger and less big. If we assume that the cow's growth in body mass over time is a straight line and that daily feed is proportional to body weight (assumptions that aren't accurate but are unlikely dramatically off), then the average consumption per day will be half this amount: 29/2 = 14.5 pounds = 7 kg = 7000 g dry matter.

(7000 g dry matter eaten per cow per day) * (1.5 springtail-years prevented per g dry matter eaten) = 11,000 springtail-years prevented per cow per day!

This page reports that beef cattle in the USA are typically slaughtered at 18 to 22 months old (average = 20 months) and spend 4 to 6 months on a feedlot (average = 5 months), suggesting that about 20-5 = 15 months are spent grazing grass.

(15 months) * (~30 days per month) * (11,000 springtail-years prevented per cow per day) = 5 million springtail-years prevented per cow over its grazing lifetime. (And this isn't even counting other, non-springtail invertebrates prevented.)

Zooplankton killed by water use

Tomasik ("Water ...") estimates that withdrawing a liter of water from lakes, reservoirs, or rivers kills on the order of ~10 crustacean zooplankton (copepods, cladocerans, etc.). This number is rough and could easily be an order of magnitude higher, especially if we also count non-crustacean zooplankton like rotifers.

Beef production can famously use large amounts of water, mainly to irrigate the vegetation eaten by the cattle. Tomasik ("The Fine ...") calculates, based on Beckett and Oltjen (1993) figures, that raising a single bovine in the USA "requires an average of 742,000 liters of blue water over its lifetime." However, other sources in Tomasik ("The Fine ...") suggest that the blue-water footprint might be up to a few times higher depending on context, such as in dry California.

Maupin et al. (2014) say (p. 25) that in the USA in 2010, "Withdrawals from surface-water sources [...] accounted for 57 percent of the total irrigation withdrawals". Assuming that groundwater is mostly free of zooplankton, we have 742,000 * 0.57 = ~420,000 liters of surface water per bovine.g Assuming ~10 crustacean zooplankton killed per liter of surface water used, that's a bit over 4 million crustacean zooplankton killed per bovine.

How does this compare with ~5 million springtail-years prevented per bovine? I think the zooplankton mortality is somewhat less significant for a few reasons:

  • A springtail-year of suffering contains several springtail deaths because most springtails (especially early-dying offspring) live much less than a year. 5 million springtail-years might be more like, say, ~50 million springtail deaths, or something in that ballpark.
  • The zooplankton killed by irrigation would have died in some other way later on, so the total harm done by killing them now is less than it seems. In contrast, bringing a springtail into existence creates a full extra death (and suffering during life) that wouldn't have otherwise existed.

In regions where extensive irrigation of feed crops and pastures occurs, zooplankton mortality might be several times more than the above number and might begin to rival springtail the deaths in importance. Of course, in dry areas requiring significant irrigation, cattle grazing seems like it might increase invertebrate suffering anyway just because irrigation creates more total vegetation on formerly dry land. So both zooplankton killing and increasing plant growth both argue against irrigation-intensive beef production.

What if we count rotifers in addition to crustacean zooplankton? They could increase invertebrate deaths due to water withdrawal by perhaps an order of magnitude. However, if we count rotifers in addition to aquatic crustaceans, we should arguably also count terrestrial nematodes in addition to springtails, because nematodes and rotifers have similar numbers of neurons (Tomasik "Brain ..."). My general sense is that rotifers typically outnumber crustacean zooplankton in lakes and rivers by about an order of magnitude, or maybe two orders of magnitude; for example, see the "Snow (1974)" table in Tomasik ("Zooplankton Densities in Lakes and Reservoirs") or the "Orsi and Mecum (1986)" and "Thorp and Mantovani (2005)" results in Tomasik ("Zooplankton Densities in Rivers"). In contrast, the "Brady (1974)" and "Curl and Truelove (1986)" results in Tomasik ("Abundances ...") suggest that soil nematodes outnumber springtails by perhaps two or three orders of magnitude. So plausibly soil nematodes outnumber springtails more than rotifers outnumber crustacean zooplankton. Hence, cattle grazing's effects on terrestrial invertebrates still seem plausibly more important than its killing of aquatic invertebrates.

Nonetheless, the zooplankton carnage involved in beef production isn't negligible, and given the high uncertainty in this analysis, the zooplankton-killing consideration deserves further scrutiny.

What about the fact that it's not clear if cattle grazing actually does reduce rather than increase populations of terrestrial invertebrates? Well, it's also not clear if withdrawing surface water increases rather than decreases the total suffering of freshwater zooplankton. Assuming the uncertainty on these two issues is of a similar magnitude, then downgrading our numerical estimates to account for this uncertainty over the direction of the impact should apply similarly to both the springtail numbers and the zooplankton numbers. For example, if we think there's a 60% chance that the impact is in the postulated direction and a 40% chance that the impact is equal and opposite in the other direction, then the above numbers should be multiplied by 0.6 - 0.4 = 0.2.

Organic vs. conventional grazing

Organic grazing farms seem to have higher abundances of dung insects, which seems bad.

Hutton and Giller, 2003

This study found that "Organic farms had significantly greater beetle biomass, diversity and species richness compared to intensive and rough grazing farms. [...] Intensive agricultural management including the use of chemical fertilisers, veterinary drugs (e.g. ivermectin) and removal of herbaceous field boundaries could be detrimental to dung beetle biodiversity and dung decomposition." The following figure shows trends for three types of cattle-grazing farms ("I = intensive, R = rough grazing, O = organic") over time ("1 = spring, 2 = early summer, 3 = late summer and 4 = autumn"):

Geiger et al., 2010

This study also found higher insect abundances on organic pastures:

Effects on vertebrates

This piece has focused on insect numbers because total bug biomass is greater than total vertebrate biomass. However, the effects of grazing on vertebrates are also relevant to an overall moral evaluation.


Intuitively, more cattle grazing an area should mean fewer wild herbivores. A priori, the net change in the biomass of total (wild + domestic) big herbivores is unclear.

I haven't yet looked at studies of how grazing in Western countries affects big wild herbivores. However, here's one study of African grazing:

Wild herbivore decline in African savannas has been attributed at least partially to competition for forage resources with livestock (Ottichilo et al., 2000a; de Leeuw et al., 2001; Said, 2003). Competition between wild grazers and cattle is generally assumed (Lamprey, 1963; Casebeer and Koss, 1970; Field, 1975; van Dyne et al., 1980; De Bie, 1991), especially during low rainfall periods (de Boer and Prins, 1990; Voeten and Prins, 1999). During these periods cattle are often concentrated around the remaining water bodies, apparently displacing wild grazers and reducing their resource availability (Western, 1975; Coppolillo, 2000; de Leeuw et al., 2001).

Yet, our study provides no strong evidence of spatial displacement of wild grazers by cattle.


Predators may be killed to protect livestock. This report says of livestock farming: "resource conflicts with pastoralists threaten species of wild predators".

This page explains:

Keystone predators like the grizzly and Mexican gray wolf were driven extinct in southwestern ecosystems by “predator control” programs designed to protect the livestock industry. Adding insult to injury — and flying in the face of modern conservation science — the livestock industry remains the leading opponent to otherwise popular efforts to reintroduce species like the Mexican gray wolf in Arizona and New Mexico.

However, the total numbers killed are relatively small. The US cattle population is about 90 million. In contrast, the total animals killed by the US Department of Agriculture's Wildlife Services division was slightly over 4 million in 2013, and only a fraction of that killing was to preserve livestock. That said, probably some predators are killed privately or by other agencies?

Of course, reduction of predator populations has other effects on wild animals. On balance, these effects are plausibly net good, though further investigation is warranted.


This article explains that grazing can sometimes increase bird habitat:

“In the remnant (native) prairie, the exotic (grass) species move in and thistle comes with them,” Brite says. “We’re trying to maintain the integrity of the prairie for ease of maintenance and a diverse (mix) of waterfowl and songbirds. When brome and bluegrass take over, you get more of a monoculture. They mat over more quickly, and that’s not as attractive to the birds.”

The most common tools used to manage these grasses include grazing, haying and prescribed burning, which are followed by a period of rest.

Working with local ranchers, cattle are allowed to graze on certain Waterfowl Production Areas using a permit system. This grazing closely mimics the effects native bison once provided to stimulate plant growth.

This page notes that avicides may be used to protect livestock feed. In one instance, "USDA employees dispensed the poison in Griggstown, New Jersey, to kill an estimated 5,000 starlings that plagued feed lots and dairies on local farms."

Field animals killed by cattle

Grazing cattle step on and may sometimes ingest lots of invertebrates on pasture fields, as well as possibly a few vertebrates. Of course, large wild animals in native grasslands and forests, such as deer, would do the same.


  1. Here are details on the three locations (pp. 4-6):

    The meadow was the unwooded part of an old Methodist Camp Meeting Grounds and had been in continuous meadow for over fifty years [...]. The grass had been cut from the meadow every year, and the hay sold to the farmer visitors to feed their horses [...].

    The day pasture, which in this paper will be called the "Pasture of ABUNDANT Vegetation" [...] consisted of about 25 acres, a small part of which was in open woodland. It extended from the creek to the road, with woods on about half of both the other sides, but it was only in the open part between the woods that many observations were made.

    The night pasture, called in the present discussion the "Pasture of SCANT Vegetation" [...] was only about 15 acres, and although not over-pastured, was used to the limit of its carrying capacity, as the cattle passed through it to reach the day pasture, to which they might not be transferred until 9:00 or 10:00 A.M., whenever the farmer happened to get back from the milk station.

    Thus eight cows and one bull spent only 7 to 8 hours daily in the 25 acre pasture, while they were in the 15 acre pasture for the balance of the 24 hours, minus the time out for milking, or about 14 hours. Also, this pasture being adjacent to the barn, often had the horses turned out on it, with the result that the grass and even the weeds were very short indeed, while the grass in the night pasture was always longer, and weeds such as buttercups grew high. [...] so far as original soil conditions and elevation are concerned, all three fields are very much alike.


  2. For example:

    • "The stability-diversity hypothesis states that the more diverse a community is, the more stable and productive the community is. This hypothesis was formed from the basis that more stable and productive communities can use their resources better and more efficiently as compared to communities of less diversity." (source) If we assume that more productivity implies more animal abundance, then this hypothesis posits that biodiversity correlates with animal abundance.
    • "The species energy hypothesis suggests the amount of available energy sets limits to the richness of the system. [...] This hypothesis proposes the higher the net primary productivity the more individuals can be supported, and the more species there will be in an area." (source)

    That said, this correlation is far from perfect: "increased species richness over broad spatial scales is not necessarily linked to increased number of individuals, which in turn is not necessarily related to increased productivity.[6]"

    This textbook discusses an experiment showing that more plant diversity resulted in higher total plant productivity:

  3. Note that biomass was only measured for invertebrates bigger than 5 mm: "The total invertebrate biomass of samples was estimated by placing all invertebrates with body length >5 mm on blotting paper and weighing the blotted invertebrates. We excluded small invertebrates despite their high abundance because their small body size means they are likely to contribute less to total community biomass37, 38, and also to community-level energy fluxes37, 39, than the less abundant, but larger, organisms."  (back)
  4. If much less than half of primary production is eaten by bugs, then maybe the number of insects supported would be lower than 2 * 106 per hectare. On the other hand, houseflies are bigger than the average bug, which implies that the number of typical bugs supported might be bigger than this calculation suggested. A typical bug probably has a mass around ~3 mg, while this section estimates a housefly's mass at around 20 mg.  (back)
  5. This textbook reports that "A single adult bovine drops an average of 12 dung pads per day [...] (Waterhouse 1977)." Given 1.4 bovines per hectare, that implies 17 dung pats per day per hectare, which is somewhat lower than the estimate of 27 that I'm using in the main text.  (back)
  6. During the winter, the cows will eat hay, silage, etc. I'll assume that the amount of dry-matter biomass eaten per day during this time is the same as on green pasture.  (back)
  7. Not all of the 742,000 liters are used for irrigation, and the proportion of non-irrigation water that comes from surface-water sources may not be exactly 0.57. However, almost all of the 742,000 liters are irrigation water, so this technicality is negligible.  (back)