U of M > CBS > Cedar Creek > Research > Data >
We burned BioCON approximately every other spring. First we burned a ring from one side to another. Then we went through the ring and spot burned the areas that did not burn well during the first burn. We followed this procedure for each ring.
BioCON uses a unique Free Air CO2 Enrichment (FACE) technology to elevate the atmospheric concentration of CO2 in the experimental plots. The FACE system used in the BioCON experiment was developed at Brookhaven National Laboratory. It uses natural wind conditions to carry CO2 enriched air across the vegetation. Because the plants are outside in a more natural environment, the chamber effects normally created by enclosures such as greenhouses are reduced or eliminated.
We used 4 g NH4NO3/ M2/year and 0.474g 15NH415NO3 /plot/year. First, we filled a small tin full of soil (approximately 100g) and put it in a bucket. We did this 31 times for ring one. Each ring will have a varying number of tins full of soil based on the number of plots in that ring that are receiving the nitrogen application. We then added 3.28g 15NH415NO3 in the ring one bucket (each ring will have a different amount depending on the number of N addition plots in that ring). If the ring has 30 plots to receive the application 3.17g of 15NH415NO3 were added. If the ring has 32 plots to receive the application 3.39 g of 15 NH415NO3 were added. We mixed up the contents in the bucket and then placed a tin full of soil in each 2lb paper bag labeled with the plots for ring one. We then placed 15.69g of NH4NO3 in each paper bag and mixed the soil with the15N and the NH4NO3 as much as possible. We then repeated this with the appropriate number of tins of soil per bucket and appropriate amounts of 15N for each subsequent ring. When applying the fertilizer, we mentally created four equal squares in the plot and tried to place equal amounts of fertilizer in each of the four squares.
Weeds were placed in paper bags labeled with the plot number and date. Sometimes we sorted the weeds into categories; sixteen species weeds - weeds that were in the experiment but were not in the plot, and real weeds - weeds that were not in the experiment and found in the plot. All of the species that were not supposed to be in the plot were pulled. Weeding tools were used to try to get out all of the roots of each plant without damaging other plants surrounding it. The weeds were then put into one of the bags identified above and all of the bags were combined and placed into a main or "big" bag. If soil was disturbed during the process of extracting plants, we made sure to smooth it out afterwards. All weeds were dried and weighed.
These samples were taken at the same depths as the 5-year archival soil samples: 0-10cm, 10-20cm, 20-40cm, and 40-60cm. A 2" diameter schedule 40 PVC root corer was used for the 0-10 and 10-20 cm depths. A 2" diameter machined metal root corer was used for the 20-40 and 40-60 cm depth. Soil was sampled when moist. We made sure we didn't knock soil into the hole when inserting the root cores, so the amount of soil in the cores accurately reflected the density in the field. We cored every plot and put the respective depth in a paper bag and put these bags in the drying room. Once dry, we sifted the soil to remove roots and rocks and then we weighed the soil.
SPECIAL NOTE: This sampling was not done according to standard bulk density sampling protocols. I (Jared) conferred with Jean Knops and got our protocol approved, but never double-checked the standard protocols until after the sampling, sifting, and weighing. The first problem is the roots should not have been sifted out. However, their weight is a very small % of the total soil weight taken. I compared the bulk density sample soil weight to the June belowground harvest weight and got the following percentages: the 0-20 root weight is, on average, 0.197% of the 0-10 and 10-20 cm soil bulk density sample weights added together; the 20-40 root weight is, on average, 0.0284% of the 20-40 soil bulk density sample weight, no 40-60cm root samples were collected so I could not make a comparison at that depth. The second problem is that the volume was not determined for the rocks that were removed (this volume was supposed to be subtracted from the volume of the soil core taken). However, there are very few rocks in the soil at Cedar Creek, so hopefully this won't significantly impact the final bulk density calculation.
1 inch PVC tubes 20 cm long with one end filed to a point are used as the nitrogen mineralization tubes. First soil cores are taken and placed in a plastic sandwich bag, this is the initial sample. At the same time n-min tubes are placed in the ground with rubber stoppers on the tops close to where the cores were taken, but not the same exact spot. The soil in these tubes is collected after one month, and this is considered the final sample.
When the soil samples are taken (whether it is the initial sample or the final sample) it is then taken into the lab and two heaping scoopfuls are placed into a pre-prepared, preweighed, barcoded vial of 1M KCl, one for each plot. These are then weighed and placed on the shaker for 30 minutes. After they are finished shaking, they are placed in the fridge overnight. In the morning the contents are pipetted into smaller plastic vials that have the same label. It is important to not get any of the soil into the pipette and into the small plastic vials. These samples were then frozen until they were ready to be analyzed. The soil samples were also measured for soil moisture. They were put into preweighed tins when they were wet, put in an oven overnight to dry and weighed the next day.
Decagon Accupar light meters were used to take light measurements. For most of the measurements we took we used three places in the plot to place the light meter, all being in the percent cover frame area. First the light meters were calibrated. There was then one measurement taken above the canopy. The next three subsequent measurements were taken underneath the plant canopy but above last year's litter (still below this year's litter). The three below canopy measurements were averaged through the electronics of the light meter. We also did a light reading before the harvest in the clip strip. We only took one above canopy measurement and one below canopy measurement.
1 inch PVC tubes 20 cm long with one end filed to a point are used as the nitrogen mineralization tubes. First soil cores are taken and placed in a plastic sandwich bag, this is the initial sample. At the same time n-min tubes are placed in the ground with rubber stoppers on the tops close to where the cores were taken, but not the same exact spot. The soil in these tubes is collected after one month, and this is considered the final sample.
When the soil samples are taken (whether it is the initial sample or the final sample) it is then taken into the lab and two heaping scoopfuls are placed into a pre-prepared, preweighed, barcoded vial of 1M KCl, one for each plot. These are then weighed and placed on the shaker for 30 minutes. After they are finished shaking, they are placed in the fridge overnight. In the morning the contents are pipetted into smaller plastic vials that have the same label. It is important to not get any of the soil into the pipette and into the small plastic vials. These samples were then frozen until they were ready to be analyzed. The soil samples were also measured for soil moisture. They were put into preweighed tins when they were wet, put in an oven overnight to dry and weighed the next day.
We prepped for this process by labeling tins, sandwich bags, and vials with the appropriate labels. We then weighed the empty tins. We made up a 0.01 M KCl solution for the extraction. We pipetted 50ml of the KCl solution into each vial. After filling all of the vials we weighed them and placed them in the fridge.
Soil cores were taken in each plot after the biomass harvest. Metal soil cores (approx. 3/4 in. diameter) were used to take the cores. Cores were then taken at varying depths and placed into labeled plastic sandwich bags. If a core was needed below an existing soil layer (for example, 40-60cm is desired but is covered by 20-40cm of soil), the undesired soil depth was discarded before taking the appropriate depth. See the notes section of this protocol for more information on soil coring methods.
After all of the samples were brought in from the field, the soil was mixed in the bag and then two spoonfuls were placed in the vials of KCl. Soil was also placed in the tin until the bottom was covered, and then the bag, tin, and vial were scanned. The vials were weighed, placed on a shaker for 30 minutes, and then placed in the refrigerator. The tins were weighed then placed in a drying oven. After drying overnight, the tins were weighed again. This process was used for each sample.
The next morning the 50 ml vials were taken out of the refrigerator and disposable pipettes were used to extract samples from each vial. The scintillation vial was filled almost to the top with KCl liquid from the vial. When the scintillation vial was filled, the scintillation vials were labeled with the label from the specimen vials. The scintillation vials were then put into the freezer until they were analyzed.
Percent cover was done in June and August before the aboveground harvest in the permanent percent cover area in the center of the plot. Palm top computers were used to enter the data. Weeding benches were used to be able to look over the plot in order to increase accuracy. Before they began to actually collect data, the two people compared a few plots and calibrated themselves to each other. Then they each measured a plot separately. Each plant species was estimated for percent cover separately with bare ground and litter also included. In each plot the total of all estimates must equal 100%. The palm top computers were downloaded at the end of each day.
A root corer was used to create the root ingrowth sampling area in the soil. The core went to a depth of approximately 20 cm. A screen made of hardware cloth was molded around the root corer and inserted into the hole created by the first core. Soil from outside of the plot was sieved of rocks, live plants and root mass (which was discarded) and then placed in the hole. The soil was packed into the hole until it was level with the surrounding soil and approximately 1 cm of the hardware cloth in the hole was left above soil level.
In order to sample the coring area an initial core was taken to a depth of 20 cm and the contents were put into a labeled plastic bag. The hardware cloth screen was then pulled out of the hole and a reverse corer (a corer with the inside of the bottom edge filed) was inserted in the already existing hole to 25 cm. The contents removed from the hole with the reverse corer were discarded. The screen was then put back into the hole.
When we filled the sample holes back in, we used soil from within the buffer area of each plot. A root corer was pounded down approximately 40 cm in the buffer of the plot. This soil was then sieved through hardware cloth to remove plants and roots and was put into the hole that the RIC sample was taken from and packed down. This step was implemented so each subsequent core has the same amount of nitrogen as what occurs in the plot. Soil was then taken from outside of the ring, sieved through hardware cloth, and used to fill in the filler hole in the buffer. The roots were then processed.
Belowground harvest takes place during the week directly following the aboveground harvest. Root cores have been taken at 3 different depths. For root cores taken at a depth of 0-20cm and 20-40cm, a 2" diameter schedule 40 PVC pipe was used. Root cores at the 40-100cm depth were taken with a 2" metal corer. All of the cores from a single depth (either 0-20, 20-40, or 40-100cm) in a plot were combined. The holes that had been created in all of the plots were filled with soil from outside the ring that had been sieved through 1/4" hardware cloth. After the cores were washed they were sorted according to fine roots, coarse roots and crowns. Please refer to the root washing section to learn more about how the roots were processed.
June harvest:
The June harvest roots were only ground in 2001. We ground the roots from only the 0-20cm depth for all monoculture plots in BioCON. All roots from the 0-20cm depth were combined and ground to make 1 composite sample per monoculture plot.
August harvest:
This is a yearly protocol. Only the 0-20cm depth for all plots was ground. In every plot all roots from the 0-20cm depth were combined and ground to make 1 composite sample per plot.
BioCON plots were harvested twice during the field season, usually at the end of June and in the middle of August. We marked a clip strip with flags in each plot running north to south that was 1m x 10cm. We used hedge clippers (blade width of 10cm) to clip the strip from the plots. Plants were clipped approximately 1cm above soil level. Biomass from the clip strip that was collected for sorting includes: every plant rooted in the clip strip (not things leaning over) and all litter lying on the ground within the strip. Each plot is checked off and brought into the lab to be sorted.
The plots are then sorted according to the rules of the season. Some seasons we sorted all plots to species and some seasons we sorted some plots to green biomass and miscellaneous litter. In order to determine what miscellaneous litter was we followed certain guidelines. In years when BioCON was burned in the spring miscellaneous litter was considered to be anything that was unidentifiable or charred. In years when BioCON was not burned in the spring the guidelines for determining what is miscellaneous litter is as follows. Anything that was dark brown or unidentifiable was considered litter. Any plant material with a trace of green or that was connected to green biomass was considered this year's growth. The plots were bagged and then dried so they could be weighed later.
For both Amorpha canescens and Petalostemum villosum we counted the number of inflorescences on stalks that were rooted in a 0.25 m2 percent cover area. Only inflorescences that were greater than 1cm in length were counted. Stalks that had passed flowering, and had already gone to seed we counted only if they were greater than 1cm in length.
A full sampling takes 2-3 days depending on the health of the machines. A sampling of the subset of NICCR plots takes approximately 1 day. All samplings previous to 2007 were done using a LiCor 6200. In 2007 we incorporated the LiCor 6400 into the samplings as well. All vegetation was cleared from the inside of the pvc collar located in the northwest corner of every plot. The LiCor chamber was set on top of the pvc collar. A thermometer was placed in the soil next to the pvc collar to measure the soil temperature at the time of the measurement. After all the plots were measured, the machines were then downloaded and the data were checked.
Collars used for measuring were cut from 4" I.D. PVC. Each collar is 2" tall. One side was sharpened at a 45° angle on grinder. Using 10 cm clippers, a 10 cm x 10 cm square is clipped at the site of collar installation. Collars are inserted to a depth of .75" (range = 0.5" to 1").
BioCON plots were harvested twice during the field season, usually at the end of June and in the middle of August. We marked a clip strip with flags in each plot running north to south that was 1m x 10cm. We used hedge clippers (blade width of 10cm) to clip the strip from the plots. Plants were clipped approximately 1cm above soil level. Biomass from the clip strip that was collected for sorting includes: every plant rooted in the clip strip (not things leaning over) and all litter lying on the ground within the strip. Each plot is checked off and brought into the lab to be sorted.
The plots are then sorted according to the rules of the season. Some seasons we sorted all plots to species and some seasons we sorted some plots to green biomass and miscellaneous litter. In order to determine what miscellaneous litter was we followed certain guidelines. In years when BioCON was burned in the spring miscellaneous litter was considered to be anything that was unidentifiable or charred. In years when BioCON was not burned in the spring the guidelines for determining what is miscellaneous litter is as follows. Anything that was dark brown or unidentifiable was considered litter. Any plant material with a trace of green or that was connected to green biomass was considered this year's growth. The plots were bagged and then dried so they could be weighed later.
We performed SLA measurements on leaves from every species in monoculture and on leaves from Achillea millefolium, Bouteloua gracilis, Bromus inermis, Koeleria cristata, Lupinus perennis, and Poa pratensis in a subset of 16 species plots in the June and August harvest. Individual leaves were taken from plants in the clip strip sample. We tried to pick anywhere from 4-8 leaves that were a similar size and from various plants in the sample. They were scanned with the WinRhizo program at a resolution of 200 dpi. Next, the leaves/grasses were put into a large coin envelope and labeled appropriately. The leaves were weighed separately from the rest of the plot.
For Lespedeza capitata we measured reproduction in 2 ways. First we recorded the number of flowering stalks/unit area. We counted the number of flowering stalks rooted within a 0.5 m2 percent cover area of all plots containing lespedeza except for monocultures. In monocultures we counted the number of flowering stalks within the entire plot. Second, we recorded the number of flowering clusters/plant on a subset of stalks. In each plot, we began counting the flower clusters on the flowering stalks at the south edge of the permanent % cover area. We counted the number of flowers on the first 5 flowering stalks we came across in polycultures and the first 12 we came across in monocultures. If several stems were coming from 1 central location (indicating that all of the stems probably belonged to 1 individual), we only counted 1 of the stems, unless there weren't enough flowering stalks available in the % cover area. If the permanent % cover area didn't have 5 flowering stalks, then we had a smaller sample size for that plot. We didn't count flowers beyond the % cover area.
We measured Lupinus perennis reproductive effort in three ways. First we counted the number of flowering stalks per unit area. We counted the number of flowering stalks rooted within a 0.25 m2 percent cover area. Then we measured the total amount of biomass allocated to seeds and pods per flowering stalk. We also counted the number of pods and seeds produced per flowering stalk. We harvested lupine seed stalks from all plots with lupine. We picked stalks with a small amount of flowers at the top and fully formed pods on the bottom; the pods were beginning the dry at the seam. To pick the seed stalks, we started looking in the northeast corner of the plot (in the rear percent cover area) and moved west from there. We picked the first 5 stalks in polycultures and the first 10 stalks in monocultures from the northeast corner. We used this method to try and eliminate visual bias as much as possible. If there was more than 1 stalk coming from the same plant, we only harvested 1 stalk/plant. If we couldn't find enough seed stalks in the rear percent cover, we started searching for stalks in the buffer, starting at the west edge of the plot and working south, then east, then north, then west. All of the stalks from each plot were put together into a plot bag and then put in the drying room. After they were dry, the pods were counted and broken apart. A "pod" was only considered a "pod" if it was longer than 1/2", if it was shorter, it was not counted. Only a subsample of the seeds were counted, dried, and weighed because many of the seeds are so tiny. We did a subsample and used that data to estimate the rest. The pod remnants were also weighed. All 5 or 10 stalks/plot were combined.
For both Amorpha canescens and Petalostemum villosum we counted the number of inflorescences on stalks that were rooted in a 0.25 m2 percent cover area. Only inflorescences that were greater than 1cm in length were counted. Stalks that had passed flowering, and had already gone to seed we counted only if they were greater than 1cm in length.
We measured the pH of the samples that were taken for nitrogen mineralization. Soil samples were taken from each plot and put into a solution of 1M KCl, which was used to extract available soil nitrogen (ammonium, nitrate, nitrite). The supernatant was then pipetted off the soil/KCl mixture into vials. This solution was then analyzed for nitrogen content. The samples were then frozen until we measured pH. We take this value to be an approximation of soil pH.
First we thawed the samples. All pH measurements must be taken within 1 week from the day the samples were thawed. When not measuring pH, the samples should be kept in the fridge. We spread the samples out so that they could reach room temperature before we started the measurements. First, we calibrated the pH probe using pH standards (pH4, pH7, and pH10). We recalibrated after each set of 10 sample measurements. If the machine was not drifting, we recalibrated after every 25 samples. We made sure to rinse the probe with deionized water after every measurement.
For Solidago rigida we counted the number of bolts with and without flowers rooted within a 0.25 m2 percent cover area.
For both Amorpha canescens and Petalostemum villosum we counted the number of inflorescences on stalks that were rooted in a 0.25 m2 percent cover area. Only inflorescences that were greater than 1cm in length were counted. Stalks that had passed flowering, and had already gone to seed we counted only if they were greater than 1cm in length. For Andropogon gerardi and Sorghastrum nutans we counted the number of flowering stalks that were rooted within a 0.5 m2 percent cover area. For Asclepias tuberosa we counted the number of stalks in each stage of reproductive development that were rooted within a 0.5 m2 percent cover area. The stages were, budding, flowering, seeding, and not flowering. For Bouteloua gracilis, Koeleria cristata, Poa pratensis, and Bromus inermis we counted the number of flowering stalks rooted within a 0.25 m2 percent cover area. For Solidago rigida we counted the number of bolts with and without flowers rooted within a 0.25 m2 percent cover area.
For Lespedeza capitata we measured reproduction in 2 ways. First we recorded the number of flowering stalks/unit area. We counted the number of flowering stalks rooted within a 0.5 m2 percent cover area of all plots containing lespedeza except for monocultures. In monocultures we counted the number of flowering stalks within the entire plot. Second, we recorded the number of flowering clusters/plant on a subset of stalks. In each plot, we began counting the flower clusters on the flowering stalks at the south edge of the permanent % cover area. We counted the number of flowers on the first 5 flowering stalks we came across in polycultures and the first 12 we came across in monocultures. If several stems were coming from 1 central location (indicating that all of the stems probably belonged to 1 individual), we only counted 1 of the stems, unless there weren't enough flowering stalks available in the % cover area. If the permanent % cover area didn't have 5 flowering stalks, then we had a smaller sample size for that plot. We didn't count flowers beyond the % cover area.
We measured Lupinus perennis reproductive effort in three ways. First we counted the number of flowering stalks per unit area. We counted the number of flowering stalks rooted within a 0.25 m2 percent cover area. Then we measured the total amount of biomass allocated to seeds and pods per flowering stalk. We also counted the number of pods and seeds produced per flowering stalk. We harvested lupine seed stalks from all plots with lupine. We picked stalks with a small amount of flowers at the top and fully formed pods on the bottom; the pods were beginning the dry at the seam. To pick the seed stalks, we started looking in the northeast corner of the plot (in the rear percent cover area) and moved west from there. We picked the first 5 stalks in polycultures and the first 10 stalks in monocultures from the northeast corner. We used this method to try and eliminate visual bias as much as possible. If there was more than 1 stalk coming from the same plant, we only harvested 1 stalk/plant. If we couldn't find enough seed stalks in the rear percent cover, we started searching for stalks in the buffer, starting at the west edge of the plot and working south, then east, then north, then west. All of the stalks from each plot were put together into a plot bag and then put in the drying room. After they were dry, the pods were counted and broken apart. A "pod" was only considered a "pod" if it was longer than 1/2", if it was shorter, it was not counted. Only a subsample of the seeds were counted, dried, and weighed because many of the seeds are so tiny. We did a subsample and used that data to estimate the rest. The pod remnants were also weighed. All 5 or 10 stalks/plot were combined.
Weeds were placed in paper bags labeled with the plot number and date. Sometimes we sorted the weeds into categories; sixteen species weeds - weeds that were in the experiment but were not in the plot, and real weeds - weeds that were not in the experiment and found in the plot. All of the species that were not supposed to be in the plot were pulled. Weeding tools were used to try to get out all of the roots of each plant without damaging other plants surrounding it. The weeds were then put into one of the bags identified above and all of the bags were combined and placed into a main or "big" bag. If soil was disturbed during the process of extracting plants, we made sure to smooth it out afterwards. All weeds were dried and weighed.
