Source: PENNSYLVANIA STATE UNIVERSITY submitted to
RELATIONSHIP BETWEEN POLLINATION DEFICIENCY AND DECLINE OF BLACK CHERRY REGENERATION IN THE ALLEGHENY NATIONAL FOREST
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
TERMINATED
Funding Source
Reporting Frequency
Annual
Accession No.
1020583
Grant No.
(N/A)
Project No.
PEN04721
Proposal No.
(N/A)
Multistate No.
(N/A)
Program Code
(N/A)
Project Start Date
Oct 1, 2019
Project End Date
Sep 30, 2021
Grant Year
(N/A)
Project Director
Hoover, KE.
Recipient Organization
PENNSYLVANIA STATE UNIVERSITY
208 MUELLER LABORATORY
UNIVERSITY PARK,PA 16802
Performing Department
Entomology
Non Technical Summary
Black cherry is the highest value timber species in the Northeastern US, with significant production in the Allegheny National Forest (ANF) of Pennsylvania. PA grows the highest quality black cherry in the world, and sales of cherry lumber and products account for a significant portion of the value of timber sales in the state. However, seed production and regeneration have been declining since 2000 in the ANF, threatening this important industry. Since black cherry requires insect pollination to produce viable seeds, the observed reductions suggest that pollination in these regions may be compromised. Declines in wild pollinator species have been broadly documented in agricultural regions, but we have little understanding of pollinator communities within forested landscapes, which are necessary for production of flowering timber trees such as black cherry, and the habitats that support these pollinators. Many factors that impact pollinator species (changing climate, site conditions, and stand structure) will also impact tree health and productivity. Our goal is to understand the relationship between habitat factors that support a robust community of pollinators and production of abundant viable black cherry seeds for regeneration. Armed with this knowledge, we will offer recommendations to the PA Bureau of Forestry and the US Forest Service for increasing black cherry seed production, and in turn, improving regeneration of black cherry in PA forests.Our objectives are to: (1) comprehensively assess the pollinators of black cherry trees and identify the most important pollinators, and (2) examine the degree to which viable seed production in the ANF is determined by the presence and abundance of suitable pollinator species, and how pollinator species and viable seed production relate to important biotic and abiotic factors in the landscape.In Objective 1, we will identify the key pollinators of this important timber species through a series of detailed observational and sampling studies, using easily accessible trees in landscapes that support diverse pollinator communities. In Objective 2, we will conduct larger landscape scale studies in the ANF to determine if the presence of these pollinator species and/or overall pollinator community abundance and diversity influence fruit and seed production (yield). We will simultaneously assess other aspects of landscape quality, stand structure, and tree health. We expect that the percentage of successful fruit set will correlate with the abundance and diversity of key pollinator species, but we also expect these parameters and the density of flowers/tree will vary with a number of biotic and abiotic factors.The results of these studies will help forest managers develop methods for assessing the quality of pollinator communities and landscapes for supporting viable seed production of black cherry in the ANF. Our findings will also provide recommendations for practices to improve pollination, viable seed set, and regeneration. New information will allow forest managers to determine which sites are most likely to respond favorably to interventions, such as thinning or shelterwood cutting.
Animal Health Component
0%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1250620107050%
2110620107050%
Goals / Objectives
Our goal is to understand the relationship between habitat factors that support a robust community of pollinators and production of abundant viable black cherry seeds for regeneration. Armed with this knowledge, we will offer recommendations to the PA Bureau of Forestry and the US Forest Service for increasing black cherry seed production, and in turn, improving regeneration of black cherry in PA forests.Our objectives are to: (1) comprehensively assess the pollinators of black cherry trees and identify the most important pollinators, and (2) examine the degree to which viable seed production in the Allegheny National Forest is determined by the presence and abundance of suitable pollinator species, and how pollinator species and viable seed production relate to important biotic and abiotic factors in the landscape.
Project Methods
Objective 1. Comprehensively assess the pollinators of black cherry trees and identify the most important pollinators. We will use 3 sites with at least 5 black cherry trees with accessible canopies (using a crawler lift) on a forest edge that showed significant fruit set in August/September 2018. Because forest edges have significant pollinator biodiversity, these sites represent trees that should be serviced by diverse and abundant pollinator communities. At each site, we will conduct the studies outlined below on each of 3 trees in spring of 2020.(a) Identify the insect community present in the canopy of flowering black cherry trees. For these studies, we will use two methods for passive sampling. As in 2019, we will use the standard sampling approach of hanging bowl traps on platforms near flowering racemes in the trees. The bowls will be hoisted into the tree canopy and placed next to the flowering racemes using two sets of bowls per tree. Bowls will be set out for 5 days before flowers begin to open, 5 days during peak flowering, and ~5 days after flowering has finished. We will collect trap contents and refill the bowls with fresh soapy water solution daily.Because our preliminary results indicated that the insects sampled by the bowl traps did not always correspond to the insects observed visiting the flowers, we will use colored sticky cards in addition to bowl traps. Sticky cards will be coated with Tanglefoot, and specimens removed by dissolving the adhesive in xylene for identification. All collected insects will be identified to species.(b) Identify insects actively visiting flowers in the canopy of black cherry trees. We will use protocols developed for orchard trees to identify insect species, and their relative abundance, which are directly visiting and interacting with black cherry flowers. Using a crawler lift, an observer will be placed in close proximity to the flowering branches. A 1 x 1 m2 plot within a sunny area of the canopy will be marked off; the insects on the flowers within this area will be observed for a 15 min period at 3 different time points during the day and the nature of their interaction recorded. After this initial observation period, a 15 min collection period will begin. The number of open flowers in this area will be recorded. All sampled insects will be identified to species. After fruit set, the number of fruit in each observed area will be counted as a measure of pollination services.(c) Determine the floral fidelity of different insect species moving black cherry pollen. In Obj. 1b, individual insects will be collected into different vials, allowing us to collect the pollen grains on each individual and evaluate the pollen species of origin. Each insect will be scored for the presence of black cherry pollen and relative frequency of "foreign" pollen. We will statistically compare results from the most abundant fly and bee species to determine which is the more efficient pollinator.Objective 2. Examine the degree to which viable seed production in the ANF is determined by the presence and abundance of suitable pollinator species, and how pollinator species and viable seed production relate to biotic and abiotic factors in the landscape. Obj. 1 will evaluate the diversity and efficiency of pollinators that are attracted to black cherry flowers at the forest edge. Pollinator diversity is particularly high at forest edges, due to the diversity of nesting habitats and flowering plant species near edges. However, even small forest clearings can still support diverse bee communities (Roberts et al. 2017). In Obj. 2, we will determine if these pollinator species are present within black cherry stands in landscapes typically used for timber production, and if their presence and abundance is correlated with fruit set and viable seed production in black cherry. We will also investigate the potential relationship between other biotic and abiotic landscape features outlined below.These studies will be conducted along 8 transects in the Kane Experimental Forest (KEF) in 2020 and 2021 (pre-scouted in 2019). Transects will be selected such that there are at least 5 black cherry trees per transect (focal trees) and include a distribution of stand structure, understory plant communities, and levels of open canopy. All of these factors can impact pollinator communities, tree health and potentially reproduction (see for example studies on black cherry tree mortality (Long 2017) and nutritional stress in sugar maple (St. Clair et al. 2008)).Insect species abundance and diversity. We will sample the insect populations in the canopies of focal trees during maximal bloom as for Obj. 1a. In 2020, we will use bowl traps and sticky traps, and transition to only sticky traps if our results from Obj 1 continue to suggest that these provide the best measure of the insects actually visiting the flowers.Quantities of viable fruit produced. After blooming is completed, for each focal tree, we will collect seeds every 2 weeks from June through November using seed trapping methods developed by Bob Long of the US Forest Service (2017). We will test seed viability, which is dependent on successful insect pollination, using tetrazolium staining based on protocols developed by Verma and Majee (2013); this method has worked well in our hands and is used by the USDA National Seed Lab. There are at least 3 major periods after drupe set begins that black cherry routinely drops developing seeds (Long 2017).Understory composition. We will characterize the understory plant composition using a 3-m fixed-radius circular plot centered at each focal tree. Within each plot, we will record the number and taxa of seedlings and saplings, as well as the estimated percent cover of understory shrubs and grasses. We will also record cover estimates for snags, leaf litter, exposed soil and rock.Tree health. Parameters for tree health will include live crown ratio, crown density, tree height, and dbh. Two increment cores per tree will also be taken and analyzed to determine tree age and vigor.Open canopy. We will also estimate the percentage of open vs. closed canopy surrounding our focal trees; we can obtain ultra-high resolution (gigapixel) photos from under the canopy and image side branches using GigaPan cameras. Climate. For each focal tree we will collect general weather conditions (rainfall, temperature, wind speed, humidity) from the weather station at KEF.Soil conditions. We will measure soil moisture near the root zone of each focal tree using a tensiometer. Two soil samples from each site will be tested by the Agricultural Analytical Laboratory at Penn State for soil pH, since black cherry prefers a slightly acidic soil for proper development. From the same samples, the test will also provide data on the levels of phosphorus, potassium, calcium, and magnesium.Using multivariate statistical approaches, we will evaluate the explanatory power of each measured biotic and abiotic factor, including weather (temperature, rainfall amounts and duration, growing degree days), canopy cover, tree health, age and size, pollinator community abundance and diversity, soil characteristics, and understory cover and composition on our response variables (amount of fruit set and amount of viable seed set). Fruit set and viable seed set will likely depend on pollinator community abundance and diversity, which may be most dependent on the abundance of the key pollinator species identified. However, we anticipate many of the other factors we measure will also be significantly correlated with the response variables. We also expect pollinator community diversity and abundance will be correlated with other features of the landscape (understory composition, proportion of open canopy and bare ground) and weather.

Progress 10/01/19 to 09/30/21

Outputs
Target Audience:Target audiences are foresters, land managers, USDA Forest Service, PA Bureau of Forestry, pollination researchers, and the general public. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?A graduate student was trained on this project in pollination research and statistical methods. She gave several presentations at scientific conferences and was trained in how to write up her results for publication. A graduate student learned how to use aerial imagery and machine learning during this project and gave a presentation about his work at a national conference. How have the results been disseminated to communities of interest?Results were presented at several conferences, to the PA Bureau of Forestry, and a copy of the publication from this project will be provided to the USDA Forest Service and PA Bureau of Forestry. Other researchers have read about our study and the results in newsletters from the Center for Biodiversity and the Pollination Research Center at Penn State University. Two MS theses were completed on this project: Reimer, Tyler. A UAV Imaging Application to Monitor Black Cherry flowering and Regeneration in Allegheny National Forest in Pennsylvania. Sept. 2021. McLaughlin, Rachel. Insect pollinators of black cherry (Prunus serotina) in the Northeastern United States. May 2021. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? We established study sites in Centre County, PA, and on the Allegheny Plateau in Northwestern PA at the Allegheny National Forest (ANF) and Kane Experimental Forest (KEF). At all study sites, we identified the insect communities present in the canopy of black cherry trees and quantified viable seed production. To identify the insects actively visiting the canopy and for measuring floral fidelity, we only used the trees in Centre County since they were shorter with more accessible flowers. For field relevance, we tested the influence of tree characteristics on pollinator communities and viable seed production at sites in the ANF and KEF. In general, we found that viable seed production was related to the abundance of andrenid bees, which carried the greatest amount of black cherry pollen. Edge trees produced more viable seed than interior trees, and seed viability declined slightly with DBH and tree age. For Centre County, we selected three sites in 2019 and added a fourth in 2020. To study the insect community that is attracted to black cherry flowers, we deployed two canopy traps per tree in three trees at each site in both sampling years. Each canopy trap supported threebee bowls of different colors on a platform, one that was painted fluorescent blue, one fluorescent yellow and one white to attract a diversity of potential pollinators. Bowls were filled with soapy water and deployed for five days before the trees bloomed, for the duration of the bloom period, and for five days after bloom. Bowl trap contents were collected every 24 hours. We collected a total of 5,525 insects over the two years of trapping. We found that total abundance of insects trapped differed by site. During the bloom period in both years, the average number of insects collected was greater in ground traps than canopy traps, but there was not a difference before and after the bloom period. We completed a total of 8.25 h of observations of insects interacting with black cherry flowers in 2019 and 13.5 h in 2020 for a total of 271 insects over two years. We also hand collected 200 insects interacting with open flowers; 48 of these were andrenid bees, and they consistently carried more black cherry pollen than any other taxa. To measure seed production, we deployed four seed collection buckets per tree at the start of August in 2020; contents were gathered twice a month to the end of October, and all resulting fruit was tested for viability using a tetrazolium assay. The average viability of seeds in fruit collected at Site #1, Site #2, Site #3, and Site #4 were 4.77%, 52.70%, 80.77%, and 59.3% respectively. A higher proportion of viable seeds in the fruit was collected from trees at sites with greater insect abundance. Greater abundance of hand-collected andrenids was predictive of a higher proportion of viable seeds. In the ANF and KEF, we established 8 sites in 2019, and 9 more in 2020, for a total of 17. Each site consisted of two paired black cherry trees: one on the edge of the forest and one in the interior. Canopy and ground traps were of the same design as those deployed in the Centre County sites. In 2019, bowl contents were collected every 48 hours. Due to heavy winds and rain events leading to the loss of samples in 2019, we increased the frequency of trap collections to every 24 hours in 2020. We collected a total of 3,679 insects over two years and total insect abundance was greater in traps on the forest edge than in the interior. Ground traps collected more insects than canopy traps and flies were more abundant at interior trees than edge trees, while more andrenids were collected on edge than interior trees. Trees produced significantly more seeds and a greater proportion of viable seeds in 2020 than in 2019. The total number of fruits produced increased with increasing total insect abundance and as the number of andrenids collected increased. Edge trees produced a higher proportion of viable seeds than interior trees. Total insect abundance was not a significant predictor of the proportion of seeds that were viable, but this proportion did increase with the number of collected andrenids. The proportion of seeds that were viable was slightly less as trees aged and DBH increased, but these two variables were correlated with each other. Flies were also abundant, but carried a significantly lower amount of pollen than andrenids. Results suggest that the abundance of the solitary bees of the family Andrenidae can improve the production of viable black cherry seeds. Given that edge trees had more andrenids and more viable seeds, management activities that increase the amount of open canopy surrounding mother trees could benefit black cherry regeneration. This would be consistent with the shelterwood cuts prescribed several years ago for promoting black cherry health in the Allegheny National Forest. Machine learning algorithms were developed for aerial images using a drone and YOLO v3 and v4 models to identify black cherry tree flowers and their slower stages. Image data sets were collected including images of flowering, non-flowering, and post-flowering racemes in the black cherry trees in the spring of 2020 and 2021, during April, May, and June in Lederer Park in State College, Pennsylvania (40.7965° N, 77.8403° W), Kane Experimental Forrest located near Kane, Pennsylvania (41.5980° N, 78.7656° W), and Allegheny National Forest located near Kane, Pennsylvania (41.6488° N, 79.0061° W). Performances of deep learning models based on image input size were compared to investigate the efficient algorithm in terms of detection accuracy while reducing model training and detection time. Impact statement: Black cherry trees in the Allegheny Plateau in Pennsylvania have experienced a steep decline in numbers over the past few years. The black cherry trees' decreased reproduction has been proposed to be associated with pollinator asynchrony, which may be linked to climate change. The outcomes of this study can be used to explore the potential reasons for decreased regeneration of black cherry trees by identifying the locations of black cherry flowers and the timing of flowering to study pollination of the flowers.

Publications

  • Type: Journal Articles Status: Published Year Published: 2022 Citation: McLaughlin, R., J. Keller, E. Wagner, D. Biddinger, C. Grozinger, and K. Hoover. 2021. Insect visitors of black cherry (Prunus serotina) (Rosales: Rosaceae) and factors affecting viable seed production. Env. Entomol. DOI: 10.1093/ee/nvab141.


Progress 10/01/19 to 09/30/20

Outputs
Target Audience:The audience for this project includes the USDA Forest Service, State foresters, and private forest managers. Also the general public may be interested in how results affect management of private lands to encourage pollinators to visit flowering trees. Changes/Problems:Because of the coronavirus, we could not use a crawler lift this year to reach 80 feet into flowers at the KEF. As a result, we are using other data to provide the information we need. What opportunities for training and professional development has the project provided?Two graduate students were trained on this project and learned about flying drones and working in forest systems with VERY tall trees. One of the graduate students has attended a scientific conference where she presented her work. Both students are developing their writing and statistical analysis skills. How have the results been disseminated to communities of interest?One of the graduate students has presented her work at the annual Entomological Society of America meeting and the PI has given a talk about this work to the PA Department of Agriculture (PDA) - Hardwoods Development Council. What do you plan to do during the next reporting period to accomplish the goals?We will complete analysis of the data and processing of images collected using drones. Ultimately these results will be submitted for publication and provided to the PA Department of Conservation and Natural Resources (DCNR) Bureau of Forestry and the USDA Forest Service in PA.

Impacts
What was accomplished under these goals? Objective 1 (Hoover, Grozinger & Choi) During the spring of 2020 we selected four sites across Centre Co., PA where at least fivemature black cherry trees were present. At each site, we selected threefocal trees in which we deployed canopy traps. Each canopy trap consisted of a platform base with threebee bowls, one white, one yellow, and one blue, that could be hoisted into the tree canopy. In May, we deployed twocanopy traps per tree for fivedays before bloom, for the duration of bloom, and for fivedays after the end of bloom. We also placed bee bowls on the ground next to each tree as background sampling. Bee bowls were filled with soapy water and collected every 24 hours. All bowl trap contents have been pinned and will be identified to species. During bloom we also implemented visual observations and hand collections in two additional flowering trees at each site. Observations and hand collections lasted for 30 minutes each at 9:30am, 11am, 12:30pm, and 2pm on 1m(2)section of the canopy. We visited one tree per site per day, rotating between sites over subsequent days until both focal trees had been visited at each time point at each site. During visual observations all insects interacting with the flowering racemes were recorded. During the hand collection period all insects interacting with the blooms were collected into individual vials. All hand collected insects will have the pollen washed from their bodies, slide plated, and identified to determine the relative frequency of black cherry pollen. Our 2020 hand collections and visual observations were dominated by Andrenid bees, which is consistent with the 2019 findings (before the grant period started). These results indicate that early springtime bees may be the dominant pollinators of black cherry. In mid-August we deployed fourseed collection buckets per tree to collect the resulting fruit at our focal trees. Dropped fruits were collected in September and will be tested for seed viability using a tetrazolium assay. Objective 2 (Hoover & Grozinger) We identified 17 sites across the Kane Experimental Forest (KEF) with mature black cherry trees. Sites were at least 600m apart to ensure we were sampling separate pollinator communities. At each site, we selected one edge tree and one interior tree, defined as being at least 150 m inside the forest, to serve as our focal trees. Due to the large size of the black cherry trees in the KEF making their canopies inaccessible, it is difficult to monitor when the trees are flowering. To determine when bloom had begun, we used drone imagery to identify open flowers on the racemes. After verifying bloom in June, we deployed two canopy and two ground traps per focal tree. Canopy and ground traps were the same design as those deployed in State College. Traps were filled with soapy water and collected every 24 hours for fivedays. All collected insects will be pinned and identified down to species. To characterize forest composition surrounding each focal tree, we followed a forest inventory plot protocol designed by the USDA Forest Service, with slight modifications. Using the focal tree as a plot center, we deployed four 12.6m lines in each cardinal direction. Within the generated quadrants, we measured the diameter at breast height (DBH) of all snags, as well as the DBH, crown position, and identification of all trees over 12.7 cm DBH. For each focal tree, we recorded the DBH, crown position, percent crown density, percent live crown, and percent fine twig dieback. In November after sap flow, we will return to take two increment cores from our focal trees. Tree cores will be analyzed to determine tree age and vigor. To measure seed production, we deployed four seed collection buckets per tree at the start of August. Bucket contents were gathered twice a month during August and September. Collections will continue until the end of October, and all of the resulting fruit will be tested for viability using a tetrazolium assay. In conjunction with this project, Dr. Dana Choi and her graduate studentused imagery from drones to test and then acquire images of black cherry trees. A smaller drone with an RGB (red, greeen, blue)camera and an optical zoom lens showed better performance for image acquisition of black cherry racemes.A machine vision algorithm was developed to identify the stages of black cherry flowering. The algorithm results were combined with GPS coordinates of the images and used to track locations of the black cherry flowers.A software program was used to reconstruct a digital surfacer model (DSM) and 3D point clouds of trees. The 3D model will be used later to compute flower abundance in racemes on the tress (canopy/flower density).

Publications