More Forest Thinning Science

1. You need Water for Ecohydrology

I previously wrote about the effects of thinning Southwestern ponderosa pine forests on forest hydrology.  AE Brown et al (Journal of Hydrology, 2005) summarized the last 50 years of hydrology research around the world on exactly this question.  Their conclusion was that yes, thinning increases water availability by decreasing evapotranspiration (ET).  However, at drier sites there is less of a difference.

In the figure above, the difference between the grass and forest curves represents the change in mean annual water yield for 100% conversion of one vegetation type to the other.  Partial conversion (i.e. thinning) was shown to have a proportional partial response.  The lack of difference between grass and forest in drier climates (below 500mm or 20 inches precipitation/year) indicates that most ET is actually just evaporation in these areas.  Therefore, because transpiration does not play a large role, reducing transpiration via thinning would not be expected to generate a large increase in water availability.

2. Don't Miss the Forest for the Trees

This classic forestry study found that thinning ponderosa forests increased growth of the remaining trees, but decreased total wood production.  In other words, the increase in vigor didn't compensate for the decrease in trees.  This even includes the decrease in disease (bark beetles) in thinned forests. So the question becomes, do you want a healthier forest or more wood?  

Data is from the The Level-of-Growing-Stock (LOGS) study on thinning ponderosa pine forests in the US West: A long-term collaborative experiment in density management.  A 2020 follow up provides a summary review of this study that started in 1962.  The The AZ portion of the study was conducted at Fort Valley experimental Forest just north of Flagstaff.   PDF with much more info is available from https://www.fs.usda.gov/rm/pubs/rmrs_p055.pdf.

Finding the Greenest Place in Arizona

 I love plants, and I love places where plants are happy.  But how can I find the greenest places, especially when I live in Arizona?

In the Southwest deserts, plant phenology and growth are dependent on intermittent rains.  But the rains can be variable between nearby areas. 

Rainfall totals and percent of normal can be viewed using https://water.weather.gov/precip/ This uses observed (radar) precipitation. 

Displaying Last 90-Day Percent of Normal Precipitation from Sep. 20.

However, more important than total rainfall over the growing season is timing of rainfall.  A single deluge that brought 3 inches last week to an area that hasn't seen rain in 6 months, is not as effective in stimulating plant growth as regular weekly or biweekly 0.5 inch storms.

A better approach to locating areas of high plant growth is using NDVI (Normalized Difference Vegetation Index), or observed "greenness" from satellites.  Data updates every week or couple of weeks, depending on the satellite.

But some areas, like forests, will always look greener than other areas, like deserts.  What we really want to locate are areas of anomalous greenness.

This UA webpage calculates departures from last week, last year, and from the average year.

VIIRS NDVI difference from average from Sep 6.

Other resources include the USGS Vegetation Drought Model, VegDRI.  The Vegetation Drought Response Index (VegDRI) shows seasonal drought impacts on vegetation. The weekly index is produced using a model trained on the self-calibrated Palmer Drought Severity Index using MODIS satellite-derived greenness anomaly integrated with precipitation, land cover, soils, and other biophysical data sets. Updated every Monday by 10 AM. 

VIIRS NDVI as shown by the USGS VegDRI model.

Clicking on the map generates a time series, which can be useful to compare seasonal and yearly changes at a location.

Graph of VegDRI drought index for a location in Yavapai County over the last 4 years.

The United Nation's FAO also maintains an NDVI anomaly drought mapper that uses the European satellite AVHRR.  However, the data is only provided at the scale of entire nations.  

So where is the greenest place in Arizona?  As you may have noticed from the screenshots above, that depends on who you're asking!  I hope to "ground-truth" some of these models to learn which are most reliable for greenness-hunting.

2000-2021 Drought in the Southwest

 

The list of Drought Impacts:

D0  Forage is limited; soil is dry

Fire risk increases

D1 Plants are stressed; hillsides are unusually brown

Stock ponds and creeks are nearly dry; some springs are dry

D2 Water and feed are inadequate for livestock

Fire danger is high; fire crews are mobilizing

Little forage remains for wildlife; pine trees are losing needles

D3 Ranching operations are affected

Fire preparedness increases; fire restrictions are implemented early

Skiing tourism is low; snowpack is extremely low

Wildlife encroach on developed areas in search of food and water

Native plants are stressed

Livestock do not have adequate water; runoff is short; conditions are dusty

D4 Fire restrictions increase; large fires occur year-round

Vegetation green-up is poor; native plants are dying

Lakes, ponds, and streams are dry

Ponderosa Mortality After Fire

The best data comes from Hull Sieg, Carolyn, et al. "Best predictors for postfire mortality of ponderosa pine trees in the Intermountain West." Forest Science 52.6 (2006): 718-728.

This graph shows probability of death based on both “crown consumption” (actual burning of the crown) and “crown scorch” (browning of needles due to heat injury). It shows that probability of tree death increases above 50% when Crown Scorch increases past 75%, whereas tree death is above 50% when crown consumption is only 25%. A combination of the two is determined to be the best predictor of tree mortality.

Here are some diagrams of crown scorch and crown consumption. In general, if there are brown needles on the tree that is scorch, if they are black or missing it is consumption.



(from Tree and forest restoration following wildfire by Peter Kolb)

Bark Beetles

The Threat

Records dating back as far as 1750 indicate a consistent record of intermittent outbreaks of aggressive bark beetles in various forest types across the West. Over the past 10-15 years, however, the frequency, severity, and extent of bark beetle outbreaks have increased.

The current bark beetle outbreaks differ from previously recorded infestations because of:
Their intensity — bark beetles are killing trees in larger numbers, at a faster pace, and over longer time periods
Their extent — bark beetle outbreaks are occurring in numerous ecosystems from Alaska to northern Mexico
Their synchroneity — bark beetle outbreaks are occurring concurrently across western North America


The Characters: all beetles in the subfamily Scolytinae

Figure from Bentz, 2005.

Dendroctonus sp (Bark beetles) Outbreaks are usually associated with drought (Page, 1981). The beetle also responds to fire-damaged trees, but not those killed by fire.

Ips sp. (Pine engraver beetles): Species within the genus Ips, such as the piñon ips and
Arizona fivespined ips, also can kill their hosts, although typically they are not considered major disturbance agents. In recent years, however, elevated population levels of a number of Ips species have coincided with drought, resulting in large areas of mortality, particularly in piñon and ponderosa pine forests of the southwestern U.S.

The above beetles are distinguished from Ambrosia beetles as they only attack and consume live trees, whereas Ambrosia beetles can bore into heartwood even after a tree is dead and dry.

Ambrosia beetles (several species): Classed as wood borer beetles that attack weakened, dying, and
recently cut or killed trees. They can attack freshly cut lumber and lumber in decks before it is dried, and they can cause pinhole defects and dark staining in the outer wood. Galleries are formed in the sapwood or heartwood and damage the wood. Because ambrosia beetles tunnel into the wood, they are considered wood borers rather than bark beetles.


Management Considerations

When live trees are blown down, their phloem can remain suitable for bark beetle development up to a year later.  Stressed pine trees emit volatile compounds (terpenes). Bark beetles have evolved to detect these compounds and use them to identify suitable host trees. 

Bark beetles are also attracted to freshly cut wood.  Freshly processed chips emit the same volatile compounds (terpenes) as susceptible host trees. These chips will attract bark beetles during their active periods. The bark beetles cannot utilize the chips as a food source, but the attracted bark beetles may then colonize suitable host trees adjacent to the chips piles.

Fresh dead and down material (slash) can be a refuge for bark beetles, allowing them to breed, and possibly colonize nearby healthy trees. Lop and scatter can let slash dry out, killing any beetles in the bark and preventing new beetles from feeding on it.

Wood that has been debarked is not suitable for colonization by bark beetles. Only freshly cut, logs or slabs that have not been debarked are at risk of colonization.

Dead trees that do not have bark beetles in them and that do not pose a safety hazard can be left in the forest to be used by wildlife.  Dead trees do not necessarily pose more of a fire hazard than live trees.

Figure from Bentz, 2005.

Additional Resources

Diana L. Six, Ryan Bracewell, Bark Beetles, 2015

Bentz, Barbara.  Bark Beetle Outbreaks in Western North America: Causes and Consequences.  2005.

A New Precipitation-Evapotranspiration Index to Map Global Drought

Standardized precipitation–evapotranspiration index (SPEI) is precipitation minus potential evapotranspiration (Vicente-Serrano et al 2010).  It is distinguished from other drought indices because it accounts for temperature through modelled evapotranspiration. This website maps global drought:

Drought in the US 2010-2011.

SPEI has recently been used to predict pine mortality in SW forests.  When the SPEI is below -1.68 for at least 11 months, pinyon and ponderosa pines cannot grow and mortality soon follows. (Kolb, T.E., 2015. A new drought tipping point for conifer mortality.Environmental Research Letters, 10(3), p.031002.)

Long-term drought graph for NM.

2015 Summer Monsoon Totals

The NWS maintains a report of monsoon totals for the ABQ metro area.

Percent of normal monsoon moisture received in July, August, and September 2015.

Moisture during the summer was spotty, but usually above average.

Four Forests Restoration Project

The draft EIS for the first 500,000 acres has been released.

Thinning work is already ongoing, but is behind schedule:

"The company said it is now thinning about 30 acres a day, which works out to about 625 acres a month. That’s a significant increase in the pace of operations since January, but still far behind the schedule established for the project nearly four years ago.  Ultimately, the company’s 10-year contract with the Forest Service requires it to clear 40,000 acres annually. In the nearly two years the company has had the contract, it has cleared about 3,700 acres. That puts the company about 70,000 acres behind the original schedule."

In my experience, in the Jemez, the major time lags are for completing NEPA and EIS and waiting for good prescribed fire weather.  Cutting the trees is fast and easy, and if they skip the fire (unnecessary and possibly environmentally detrimental) nothing should slow them down.  One of my other main criticisms from the Jemez is that they're not thinning enough trees to reduce the basal area to the most beneficial levels.  I know the ideal density and pattern of trees has been argued about ad infinitum... it seems they are trying to avoid conflict by cutting less trees, which totally defeats the purpose of preventing catastrophic crown fire.

What stand density is optimal in Ponderosa pine forest restoration?

 In my humble opinion, there is really only one chart needed to answer the question of what density is best for Ponderosa pine trees:

Source: The role of stand density on growth efficiency, leaf area index, and resin flow in southwestern ponderosa pine forests.  McDowell, Adams, Bailey, and Kolb.  (link)

 The lower the basal area, the more trees grow.  Growth is especially important in drought-stressed trees because lower growth can weaken tree defenses.  The lowest treatment in this study reduced trees to 7 m2 ha-1, and trees in this treatment had the highest leaf area per tree, indicating that they were healthiest at the time of the study.

Leaf area of understory plants was also highest at the lowest stand densities:

Source: ibid.

However, in the interest of full disclosure, this chart also shows (in panel B) that total forest resource utilization, measured by total leaf area, is highest at medium tree densities.

Predicting Plant Phenology

Plants in the Southwestern deserts respond to water availability and temperature:

Map from AHPS Precipitation Analysis from January and February

NM is doing good on water so far this year!

Growing degree day map from PNWPest.org

On average NM is about 6 growind degree days (GDD) behind 2014, but 11 days ahead of 2013 and 6 days ahead of normal.

Tamarisk Beetles Continue To Spread

Tamarisk beetles defoliate the invasive tree throughout its range:

Herbaceous Vegetation in Southwestern Pine Forests

Herbaceous forage in southwestern pine forests can be few and far between.  Especially in unnaturally-overgrown thickets, there is simply not enough light and too much acidic leaf needle litter to grow robust grasses and forbs.  The high carbon content and resistant nature of conifer bark and needles means that few nutrients are available to growing plants.  

Even when forest thinning restores natural stand densities, the recalcitrant accumulated biomass often seems to limit production.  There is simply too much carbon and it is clogging up microbial turnover of nutrients.  Sure, you'll find plenty of aboveground mushrooms and mycelial hypahe at work in the soil, but I have to wonder if there isn't a more productive alternate-stable state.

What is the limiting resource?  I've theorized that light may still be limiting when forests aren't thinned enough to create light gaps in the canopy.  But it also seems that the overabundance of carbon may be soaking up nitrogen.  In these nutrient-poor systems we typically find 'tolerator' species like sedges, that can eke out a frugal living in acidic conditions.

What about blueberries?  In some parts of the world, conifers and blueberries go together like apples and pie, but in many of the southwestern mountain ranges we have no native Vaccinium.  I'm still not sure why, except that they seem to prefer colder (moister?) climes than New Mexico can provide....

There are two species of Vaccinium in the southwest, with Vaccinium myrtillus much more common.  However, south of Santa Fe even this species only occurs at scattered locations in mixed conifer and subalpine forests.  These high-elevation forests are being rapidly lost to stand-replacing forest fires.  Even with current climate change, it is unlikely that extensive tracts of cooler forests will be able to regrow.

What about legumes?  Fabaceae are often able to supply their own nitrogen requirements and eventually supplement total ecosystem N.  But I'm surprised by how rarely I find good legumes in the forest.  Thermopsis is surprisingly rare, as are Lotus and I almost never find clovers growing in conifer needle duff.  

Robinia neomexicana seems to be one legume shrub that has found its way into a diversity of habitats, growing almost like a weedy in mountainous areas throughout the southwest.  

Seeding mixtures in forests typically use annual grasses, but I wonder if there isn't a Fabaceae that could dramatically enhance production to increase ground-cover, forage, pollinator, and wildlife habitat?  Should standard thin-and-chip treatments be supplemented with seeding efforts?

What about increasing disturbance to disrupt pine duff accumulation?  These thick, undecomposed layers and inhibit germination of many forbs...

Good Monsoons so far in New Mexico!

Almost the end of August and the climatological "monsoon" in the Southwest should be wrapping up around the first or second week of September. The last two months of rains have been sporadic, but significantly above average in the Land of Enchantment. Most areas in central NM are above 150% of normal for the season, and there are many areas that have tripled their average summer rainfall!

Two Views of Recent (2000-2012) Forest Fires in the American Southwest

Imaging and recording forest fire extent serves a number of important roles in forest ecology and is also very useful for planning vacations!  (Don't plan to camp in a recently burned forest)  There are now two ways of mapping recent fire locations.

The USFS ForWarn Forest Change Assessment Viewer (FCAV) can visualize fires that burned from 2000-2012 using the Fire Perimeter Layer.

 The Global Forest Change map (GFC) (Hansen et al) is a new tool that shows global forest extent (green),  deforestation (red), and aforestation (blue) and reforestation (purple).  In the high-resolution image above, one can match USFS fire perimeters to the actual burn scars, as well as visualize total forest extent.  Beetle die-off is apparently not apparent in this view.

But why isn't the 2011 Horseshoe Two fire visible? This large stand-replacement fire burned more than 70% of the Chiricahua mountains (look for a green tear-drop shape near the Southern edge of the map above).  It is clearly visible as a red blob in the USFS image, but not in the Global Forest Change map.   It is clearly visible on Landsat images:

Image source.

Zoomed in images show the disparity:

 The USFS viewer shows the entire range as burned area.

But hardly any forest loss is recorded on the GFC map.  Why??

Perhaps the discrepancy is due to burn severity?  Fires do not burn evenly across a landscape, and many regions within a burn perimeter may be only lightly or moderately burned.  

Here is the Horseshoe Two fire burn severity map, clearly showing large areas of high and moderate intensity fire within the burn perimeter:

...

An important note about the colors shades on the GFC map:  they indicate the percent forest (defined as vegetation over 5m) per grid cell.

This view of the Rodeo-Chediski (2002) fire shows different shades of red, and green.  No regrowth is apparent.

Zooming in all the way to individual pixels shows the different shades.  Assuming the forest loss shades are proportional to the initial forest cover shades, the lightest red colors indicate areas that were 75-100% forest (and now are not forest or are 0-25% forest), the light grey-red indicates areas that were 50-75% forest (and now are not forest or are 0-25% forest), the darker grey-red indicates areas that were 25-50% forest (and now are not forest or are 0-25% forest), and black  indicates areas that are (and continue to be) 0-25% forest.  

What about the 2002 Biscuit fire in the Klamath-Siskyou region....

Purple in the GFC map above may indicate some regrowth (and/or replanting) in the 10 years following the fire.

Four of my Favorite Characteristic New Mexico Shrubs

Chilopsis linearis, Desert Willow.  In landscaping, this large-flowered shrub will flower for almost half the year.  It is a dependable riparian tree that can survive in washes that are too hot and dry for cottonwood.

Forestiera pubescens.  New Mexico Privet.  Native shrub in the Olive family can form thickets at the base of canyon walls and along upper benches above rivers.  Birds prize the berries almost as much as the non-native Russian Olive.

Lycium pallidum.  Pale Wolfberry.  Lyciums are widespread in the Southwestern deserts, and Lycium pallidum is the most widespread.  Goji berries are Lycium barbarum (nonnative), and Lycium pallidum has been cultivated for its fruit for 1,000s of years.  The trumpet-shaped flowers attract Sphyngid Hawkmoths and Hummingbirds.  

Robinia neomexicana.  New Mexico Robinia, Fabaceae.  Despite its rapacious thorns, this plant made my list for its tenacity and abundance across a range of habitats.  From riparian area in the South to Ponderosa hillsides in the North, Robinia provides abundant pink blossoms and herbaceous cover. 

What does a natural forest fire look like in the Southwest?

What is "natural"?  Ecologists use the range of natural variability to define proper functioning condition (PFC) or minimally disturbed condition (MDC), etc (in the jargon of ecologists -- basically, what was it like before we mucked it up?)  The question is surprisingly hard to answer.

Tom Swetnam has been quoted drawing a distinction between burn severity (high-intensity burns are natural) and extent (but high-severity burns across large landscapes are not natural).  These statements are part of an ongoing debate that, like many debates in ecology, may be dependent on local factors.  

In Arizona, Dr. Wally Covington became famous for championing open park-like Ponderosa forests.  Indeed, many landscapes are dominated by frequent low-intensity grass fires that clear out underbrush and young trees.  But at longer time scales it seems equally clear that these same forests can catastrophically burn.  This has been pointed out by Dr. Baker's work in Arizona, and by Dr. Grant Meyer in Sacramento Mountains of New Mexico.  More evidence is not hard to find.  For example, a study of the geomorphology of the Rio Cebolla river in the Jemez Mountains noted extensive sediments with large burned logs indicative of high severity fire in what must have been extensive headwater areas. 

Dendrochronologists  often focus on the last 1,000 years, whereas geomorphologists have a much larger picture, often spanning not just this interglacial, but also the last several ice ages.  (CF Great Basin Riparian Ecosystems by Miller and Chambers et al)

Fires are transformative, but both more and less impactful than commonly thought.  They accelerate and are accelerated by climate change.  Most large fires are not as bad as news reports indicate -- they are bananzas for wildlife and fire-dependent trees and shrubs like aspen and oak.

Most "megafires" aren't really managed with a full suppression strategy, but a containment strategy -- still not 'let-burn' but pretty close.  Unfortunately, suppression and containment efforts can be as damaging as fire.  

Whether or not fires are natural, the inevitable question is raised as to whether humans can restore the environment better or faster than natural succession would?  This is a loaded question, because of the salvage logging controversy.  One logger expressed confusion about why thinning before a fire is justified, but salvage logging after isn't.  But either may be ineffective in preventing the next conflagration.  If ERC is high enough and there is enough ventilation, anything will burn, even the Olympic Peninsula rainforest! (Cf Dr. Gavin's ESA talk)

Forest Fires versus Forest Restoration

At ESA this year, one discussion turned to the elephant in the room of forest restoration thinning projects: the certainty that thinning will have some deleterious effects versus the probability that a thinned area would actually encounter a forest fire during the approximately 25-year window of restoration effectiveness.  Forest restoration typically involves thinning forests to lower stand densities to discourage the spread of stand-replacement crown fires in historically low-intensity fire regimes.  Yet,  absent the return to presettlement semi-annual burn patterns (i.e. an end to the total suppression paradigm), tree seedlings will recruit and form high density stands within a couple decades.  

What are the odds that a given patch of forest will encounter a mega-fire over a 25-year time span?  According to some participants in the discussion, the odds are quite low.  But that may depend on the forest.  Take, for example, the fire-prone semi-arid mountains of the southwest:

Representative map of burned areas in SE Arizona and SW New Mexico over the last decade.  The Wallow Fire (538,000 acres in the White Mtns) and the Horseshoe Two (223,000 acres in the Chiricahua Mtns) fire, are visible as large yellow areas.

Shaded areas indicate official MTBS 2000-2009 burned area polygons.  Yellow areas are provisional "fire detection" areas from MODIS.  This map was generated using the U.S. Forest Change Assessment Viewer (FCAV), which can map a large number of forest disturbance types, utilizing historical to contemporary GIS layers and current satellite imagery.

Perhaps the ESA discussion participants should read:


Rhodes, J.J. and Baker, W.L., 2008. Fire probability, fuel treatment effectiveness and ecological
tradeoffs in western U.S. public forests. Open Forest Science Journal, 1: 1-7.

Best Map of Tucson and Mexico

Check out species lists and maps for flora and fauna.  Data portal is powered by Symbiota, which also has fungi and South American biota.

Solar Insolation in the West

Lithic Mulch

I have done field ecology throughout New Mexico and Arizona -- but no matter how long I live in the Southwest it seems there is always more to discover...I recently discovered the existence of lithic or cobble mulch gardens while perusing a report by Richard D. Periman featured in "Desired future conditions for Southwestern riparian ecosystems: Bringing interests and concerns together," produced by the Rocky Mountain Forest and Range Experiment Station, and have spent an enjoyable morning reading some of Dale Lightfoot's papers about sites in Northern New Mexico and the Safford Valley of Arizona. Both are areas where I have spent considerable time, and I am now wondering if I have unknowingly visited some of these prehistoric gardens. I am especially interested to learn how thoroughly these type of sites have been studied...from my natural history work I understand that sometimes the best preservation strategy for valuable sites is to keep them undocumented, but I wonder if there is a GIS layer of inventoried sites? These archeologists deserve congratulations on bringing this remarkable phenomenon to light; discoveries like these may help transform how we ecologists think about the Natural Range of Variation (NRV) in these areas.

"Lithic-mulched features have not been thoroughly studied, in the USA or elsewhere. Details are spotty...The research by Maxwell and Anscheutz in the Ojo Caliente Valley in northern New Mexico is very good. The Fish’s work on Hohokam Arizona is good. Berlin et al. 1977 are quite thorough, as is Engel’s work in the Andes. Much remains to be done at all other sites associated with lithic mulching. I understand your comment about leaving things undocumented. It’s best, sometimes, though in cases where features are on the verge of being obliterated, as were many of the gravel mulched plots I worked on in 1988-89 south of Santa Fe (1990 dissertation and subsequent articles), it’s best to record what you can, when you can (i.e. salvage archaeology).

I know of no GIS inventory related to lithic mulching anywhere. I have no plans on the immediate horizon to work on this topic in the U.S. Southwest, but I didn’t have any plans to work on the Safford project, either, until Bill Doolittle approached me about joining his project and I could clearly see a way to contribute to the project."